Dealul Guran: Evidence for Lower Palaeolithic (MIS 11) Occupation of the Lower Danube Loess Steppe

Article excerpt



South-eastern Europe features consistently in models of hominin colonisation of the continent. It is the geographic point at which northward migration routes out of Africa through the Near East meet eastern ones arriving from Asia, via the Caucasus and Ukrainian steppe (e.g. Carbonell et al. 1999; Harvati et al. 2008; Muttoni et al. 2010). However, despite its geographically important position as a potential crossroads for population movements, the quality and quantity of data from the Balkans and eastern European steppe have so far been insufficient to reliably evaluate such models. Questions relating to the timing of the first arrival of hominins into Europe, to the geographic extent and timing of Mode 2 technologies (sensu Clark 1969), to the process of Neandertalisation (Stringer 1998; White & Ashton 2003), and to the association of hominins to migrating mammalian guilds (Turner 1992; Muttoni et al. 2010; Bar-Yosef & Belmaker 2011), among others, are difficult to address with only the handful of presently known sites covering the two easternmost thirds of the European continent (Figure 1).


Several explanations have been proposed for this discrepancy. Most of these attribute it to shorter and less intensive research histories in eastern Europe during the Communist period (Tabaczynski 2007; see Romanowska 2012 for a review). Others invoke behavioural responses to the climate and seasonality of eastern Europe, as well as the hominins' own difficulties in adapting to these conditions (e.g. Hopkinson 2007). However, even accounting for the lack of communication caused by the physical separation of research communities in former Communist countries from their Western counterparts, and their relative lack of finances to carry out large projects, the bulk of the problem of identifying sites is more likely to be geological in nature (Romanowska 2012).

Much of the region from Hungary eastward to China is blanketed by at least 5m of loess (Haase et al. 2007; Figure 1; also see the online supplement for more detailed information). Loess deposition in the middle and lower Danube basin of eastern Europe started more than one million years ago, providing potentially long and complete terrestrial palaeoenvironmental records (Markovic et al. 2008). The widespread loess deposition across eastern Europe leads to the likelihood that the Lower Palaeolithic (LP) of the region is deeply buried and largely inaccessible, despite isolated finds (Ranov 1995). The loess cover, however, is not always uniform, particularly in the upper parts of valley slopes where underlying karstified limestone can be exposed at the surface. Positions in the landscape such as these are particularly well suited for archaeological exploration, and form the focus of this study.

Further compounding the problem of deep burial is the perceived lack of distinguishing features of the simple core-and-flake or Mode 1 (sensu Clark 1969) lithic technology found at those few securely dated LP sites in eastern Europe. Combined with the fact that faconnaoee (form-shaping, or Mode 2)--the technique of lithic reduction commonly associated with the production of handaxes in the western European LP--is a common feature of the Middle Palaeolithic in this region (Bosinski 1967; Joris 2006), this results in substantial uncertainty even in establishing the relative chronostratigraphic position of artefacts found on the surface during field surveys. This has led to over-reliance on the 'primitive' character of an industry in order to assign it to the LP. It also increases the likelihood of classifying geofacts or gelifracted flints as artefacts, the so-called 'possibiliths' (Doronichev 2008). Such dubious attributions of lithic assemblages--without chronostratigraphic control--have led to the rejection of claims for any confirmable existence of the LP within Romania (Dobos 2008). …