Untitled Document

Fluvial Terraces as Geo-archives of the Holocene Floodplain Genesis of the Upper Dnister Valley, Ukraine[*]

Mirko Huhmann & Helmut Brückner

Department of Geography, Philipps-University of Marburg

The floodplain of the Dnister valley contains plenty of information about the younger history of its evolution. Special attention was given to the palaeosols within and the recent soils on the terraces. The stage of maturity of different soils allows to discriminate the Holocene river terraces. Moreover, is it possible to draw conclusions about palaeoecology. With the help of floodplain soils it was for the first time possible to differentiate seven Holocene Dnister terraces and to put them into a stratigraphic order. Thus, the fluvial morphodynamic history of the Upper Dnister Valley was reconstructed (Huhmann & Brückner 1999: 97)

Research was carried out in the districts of Mykolayiv and Halych at the border between Subcarpathia in the South and the Podilia Plate in the North. This area is especially characterised by comperatively insignificant human impact, such as changes due to water management and intensive agriculture. Therefore, in the Upper Dnister Valley typical elements of a natural floodplain landscape within levees, old branches, marginal peat bogs etc. and the normal flood dynamics are still existing.

Every Holocene river terrace has its own fluvial series consisting of basal facies, channel sediment, floodplain channel sediment, floodplain sediment and floodplain soil, which have been aggregated laterally during the evolution of the inner banks in meandering streams. This principle is also valid for the Holocene terraces of the Dnister. They were distinguished by geomorphologic means and mapped according to the ”seamchannel mapping method” (Schirmer 1983: 26). Every terrace is covered with a characteristic sequence of sediments, generally topped by meadow loam, out of which its soil is formed. The older the terraces, the higher the degree of soil evolution; therefore, nearly each terrace is characterised by its own stage of pedogenesis (from para-brownearth on the Late Pleistocene terraces to virgin soil on the youngest Holocene terrace). The mapping of the specific soil types leads to a relative stratigraphic classification. The chronological framework is rendered by absolute datings (mostly ¹⁴C). Together with the sedimentological investigation of the terrace bodies and soil properties which are dependent on soil age (among others: contents of carbonate and phosphate, relation of pedogenic iron and manganese to their total amount, soil colour, soil structure), it is possible to decipher the genesis and palaeoecology of the river valley.

Near Mykolayiv we identified and mapped two Late Pleistocene (Nt1-Nt2) and seven Holocene terraces (H1-H7). In their sequence from old to young they developed - beside other features - characteristic climax soils which were used as benchmark soils.

While the older Late Pleistocene terrace (Nt1) shows a characteristic topographic difference to the younger terraces, we found the younger Late Pleistocene terrace (Nt2) - unlike in other valleys - already as a so-called row terrace in floodplain level, often directly connected with the Preboreal terrace (H1). These three terraces are represented by more or less strong pedogenesis (floodplain para-brownearth) so that their obvious differentiation in the field without a simultaneous investigation of their sedimentological setup (Nt1 and Nt2 with vertical aggregation of a braided Dnister, H1 with lateral aggregation of a meandering Dnister) is delicate.

Since the accumulation of the Atlantic terrace (H2), the time for the development of a floodplain para-brownearth has been too short. Therefore, the following terraces (H2-H5) are characterised by different formations of a floodplain brownearth (vega). Distinguishing features are first of all the decalcification depth and the differences in the iron mobilisation. Because of a later covering by meadow loam, the soils of terraces H2 and H5 are in several cases buried which offers a supplementary criterion to differentiate the older Holocene terraces. From terrace H4 (Iron Age) onwards, the increasing sedimentation rate of the meadow loam proves an increase in soil erosion due to human impact (deforestation) in the Dnister catchment area.

The modern time H6 terrace (16^th-18^th centuries) is too young for the development of a floodplain brownearth. Therefore, its characteristic soil is the floodplain para-rendzina of shallow depth with no decalcification phenomenon. The youngest terrace (H7) from the 19^th/20^th century occupies the recent inner banks of the river; because of frequent flooding it only bears a floodplain virgin soil.

For further information see also http://www.dnister.de

HUHMANN, M. & H. BRÜCKNER (1999): Fluvial Terraces as Geo-archives of the Holocene Floodplain Genesis of the Upper Dnister Valley. - In: Ministry of Education of Ukraine, Lviv Ivan Franko State University & Lviv Department of the Ukrainian Society of Soil Scientists and Agrochemists (Hrsg.): Genesis, Geography and Ecology of Soils. Collection of the scientific works of the International Conference, Lviv 16-18. Sep.1999, S. 97-101; Lviv.

SCHIRMER, W. (1983): Die Talentwicklung an Main und Regnitz seit dem Hochwürm. - In: Geol. Jb., vol. A 71: 11-43; Hannover.

* This study was carried out under the umbrella of the German-Ukrainian research project “Ecological Analysis and Nature Conservation Evaluation of the Upper Dnister Region as a Model for the Development of a Floodplain Landscape in Eastern Europe“, funded by the German Ministry for Education, Research, Science and Technology (BMBF No. 0339699A3), coordinated by UNESCO (Paris). Additional promotion was granted by a scholarship from DAAD (HSP III) which is gratefully acknowledged.