Prehistoric Land Degradation in Hungary: Who, How and Why?

Article excerpt

The recent study of Kis-Mohos To lake in Hungary reveals an important sequence of prehistoric landscape changes from the earliest land clearance to the early Middle Ages. The recognition of land degradation, through the application of new analytical methods, forms an important part of the discussion.

Introduction

Land degradation is the physical and chemical depletion of the soil resulting from processes such as erosion, acidification, depletion of plant nutrients and reduction of organic matter content (Barrow 1991). One of the most important triggering mechanisms for land degradation appears to be the removal of the established vegetation cover, and in many regions of the world a close relationship has been demonstrated between anthropogenic activity and land degradation (Bell & Walker 1992; van der Leeuw 1995; Ellis & Mellor 1995). However, land degradation is not just a recent phenomenon and a number of studies suggest that from the early postglacial onwards prehistoric populations could have had a considerable effect upon landscape stability (Davidson 1982; Flenley 1988; Evans 1990; Bell & Walker 1992; Willis 1995a). Population density, farming techniques and grazing pressures are some of the principal current causes of land degradation, but is this also true of prehistoric times? Were there particular activities/technologies that triggered soil instability such as the introduction of the cart and plough (Sherratt 1981) from approximately 3500 Be? Is it possible to identify a critical threshold after which degradation occurs? And what happened after a period of degradation - did the soils and vegetation recover naturally or was the damage permanent?

Impact of different technologies and cultures upon the landscape can be measured in a variety of ways. Ideally a record of degradation is needed that is independent of the archaeological record (Edwards 1991) and has an appropriate spatial scale for human impact upon the landscape (Willis & Bennett 1994). Such a record is provided through various palaeoecological analyses of lake sediment.

The geochemical elements preserved in lake sediment provide a record both of soil erosion within the basin catchment and also mining activity in the wider region. Previous studies have indicated that increases in the abundance of elements such as aluminium, potassium and magnesium are indicative of both physical and chemical weathering associated with soil erosion (Engstrom & Wright 1984; Heathwaite & Burt 1992). These elements represent silicate minerals eroded from catchment soils and rocks. This material is carried by throughflow and overland flow into the lake basin, and can be used as an indicator of local (usually within the hydrological catchment) land degradation. Additionally, prehistoric mining can be detected through geochemical analyses, because these activities release dust into the atmosphere, including metal particles, which are transported into the basin, and their representative catchments are somewhat different. Similar to the transport of pollen and microscopic charcoal (see below), these particles are transported in the atmosphere and deposited in the lake as a 'rain'. Unfortunately there is no work on how far these particles can be transported, but it is assumed that similar to microscopic charcoal, they can represent a local (within the vicinity of the basin) or regional catchment (up to hundreds of kilometres away). Although the spatial resolution might be broad, their temporal resolution is tight, and increased input of metals such as copper can be a useful indicator for the initiation of mining activity (Willis et al. 1995).

Analysis of the pollen contained in lake sediment will identify changes in the vegetation composition about the lake, including clearance, arboriculture and cultivation of crops. Theoretical modelling has suggested that spatial representation of the pollen rain is determined by size of lake basin; that is, the smaller the basin, the more local the pollen source area (Janssen 1966; Jacobson & Bradshaw 1981; Jackson 1990; 1994). …