Early Upper Palaeolithic occupation at Gelimgoush cave, Kermanshah; West-Central Zagros mountains of Iran

Abstract

The timing and dispersal routes of Homo sapiens (H. sapiens) into the Iranian Plateau have always been a matter of debate in the recent years. Current studies on the Upper Palaeolithic period of the Zagros mountains demonstrated the later colonisation of West-Central Zagros by H. sapiens based on techno-typological and radiocarbon dating. The Kermanshah region is one of the main concentrations of Palaeolithic sites in the West-Central Zagros mountains. Despite presenting rich Palaeolithic sequence records, it suffers from the lack of stratified data associated with chronological control. This issue, until now, has prevented us from evaluating and knowing how these archaeological records relate to the patterns of H. sapiens dispersal and colonisation into the West-Central Zagros mountains. Here, we present the first excavated and dated Palaeolithic site in Kermanshah. Eshkaft-e Gelimgoush Cave yielded a classic Upper Palaeolithic assemblage, representing the Lorestan and Kermanshah (LaK) cultural group documented in the West-Central Zagros. Radiocarbon dates, associated with the material culture from this cave, provide the first dated stratified Upper Palaeolithic evidence in Kermanshah. The data from Eshkaft-e Gelimgoush are consistent with the cultural diversity model among the Upper Palaeolithic populations in the Zagros and confirm later colonisation of the West-Central than northern and southern Zagros mountains.

Introduction

During the last decade, the origins, distribution and behaviour of late Pleistocene Homo sapiens (H. sapiens) have been the subject of critical examination especially since they directly related to the demise of Neanderthals and the broad establishment of H. sapiens across Eurasia (Alex et al., 2017, Bae et al., 2017). The Iranian Plateau located at the crossroads between Africa, Asia and Europe. Despite this important geographical location, we know little of the presence of H. sapiens on this region. Recent research revolving around Late Pleistocene hominin behaviour in Iran has documented variability among the lithic assemblages, namely different cultural groups in the Middle Palaeolithic (MP) and Upper Palaeolithic (UP) in different parts of the Iranian Plateau including the Zagros mountains (Heydari-Guran and Ghasidian, 2017, Ghasidian et al., 2019, Ghasidian, 2019, Heydari-Guran and Ghasidian, 2020). However, one of the most important questions in this regard is the hominin species responsible for these cultural facies during MIS 3. The timing of H. sapiens arrival on the Iranian Plateau is directly influenced by the lack of physical remains. What is, so far, accepted – especially in the Zagros mountains, is that the MP Zagros Mousterian was manufactured by Neanderthals based on several fossils and cultural remains from Shanidar Cave (Solecki, 1963, Pomeroy et al., 2017, Pomeroy et al., 2020) and Bawa Yawan Rocshelter (Heydari-Guran et al., in press) whereas the laminar UP techno-complexes were linked to H. sapiens (Heydari-Guran and Ghasidian, 2020).

The timing of when H. sapiens entered the Iranian Plateau remains a key question. Bayesian modelling combining new and previously obtained radiocarbon dates and their stratigraphic contexts, suggested that the beginning of the UP dates to around 45 ka cal BP both in the Northern and Southern Zagros mountains (Becerra-Valdivia et al., 2017). There are two possibilities: the colonisation of the Zagros by modern humans occurred from two sources/directions, i.e. from the south and the north, independently and simultaneously (Ghasidian et al., 2019), or that it was fast and hence the direction and tempo undetectable using radiocarbon. Recent research on the UP of the Zagros documented cultural diversity among the different groups responsible for manufacturing UP techno-complexes throughout the Zagros. The UP in the Zagros has been identified in at least three varieties: the Baradostian of the Northern, the UP of the West-Central, mostly from Lurestan and Kermanshah regions (LaK: Ghasidian, 2019) and the Rostamian of the Southern Zagros mountains. These three technocomplexes have been documented in excavated and stratified, mostly rock-shelter sites, of which only few have been directly dated.

In the Southern Zagros, the Rostamian technocomplex was documented in the Dasht-e Rostam region (Ghasidian, 2014, Ghasidian, 2019, Heydari-Guran, 2014) and its start was dated at Ghār-e Boof Cave between 45950 and 39350 cal BP (95.4% probability; Becerra-Valdivia et al., 2017).

In the Northern Zagros, the number of archaeological sites is limited. The UP there is mainly documented at Shanidar Cave which yielded stratified UP deposits starting from 48,800 to 39400 cal BP (95.4% probability; Becerra-Valdivia et al., 2017, Solecki, 1958).

In the West-Central Zagros mountains, in the Khorramabad region, two sites -Kaldar and Yafteh- provided stratified material culture associated with absolute dates (Otte et al., 2011, Bazgir et al., 2017). Bayesian age modelling shows the arrival of the UP populations into West-Central Zagros mountains occurred slightly later, starting at Yafteh between 39,250 and 37,600 (95.4% probability, Becerra-Valdivia et al., 2017). These sites are located in the Khorramabad plain, regarded by some as one of the Palaeolithic hot spots in the West-Central Zagros (Hole and Flannery, 1967).

Yet an adjacent region, the Kermanshah plain, is equally promising for holding important information about the progression of the Palaeolithic period, in particular the beginning of the UP in the Zagros. Its high potential during different Palaeolithic periods has been confirmed by recent work (Heydari-Guran and Ghasidian, 2017, Heydari-Guran and Ghasidian, 2020, Biglari and Shidrang, 2016). Despite the discovery of numerous Palaeolithic sites, however, the Kermanshah plain suffers from lack of stratified deposits associated with absolute dates. So far only 2 UP sites have been excavated and published from the region. These are the Warwasi rockshelter and Ghar-e Khar Cave (Braidwood et al., 1961, Smith, 1986). Both sites yielded complete Palaeolithic sequences, spanning from the MP to the Epipalaeolithic period (Dibble and Holdaway, 1993, Olszewski, 1993, Shidrang et al., 2016). However, they lack absolute dates and it remains unclear when each period started and ended. This also limits meaningful comparisons of any lithic techno-typological analyses between neighbouring areas, especially with the Khorramabad plain.

To better understand the chronology and Palaeolithic subsistence strategies in the Kermanshah region of West-Central Zagros and its role in hominin dispersals into the Iranian Plateau, the project “Human Evolution in the Zagros Mountains” (HEZM) started by conducting field surveys to identify new Palaeolithic sites with excavations being conducted on the key sites (Heydari-Guran and Ghasidian, 2017, Heydari-Guran and Ghasidian, 2020). The newly discovered cave site of Eshkaft-e Gelimgoush in the Nawdarwan Valley has been targeted for recovering its Palaeolithic sequence and detailed investigation of the early UP in Kermanshah. The cave is located in one of the richest Palaeolithic areas of the Kermanshah plain, provided the first dated and documented stratified early UP artefacts in Kermanshah. Through the analysis of the findings from this site, we wanted to addresses two issues. The first revolves around the timing of H. sapiens arrival into Kermanshah region, and the second aims at understanding how the new site’s lithic assemblage compare to known UP assemblages from the Zagros. especially from the West-Central region including Warwasi, Yafteh, Kaldar and the other survey sites there.