Search tips
Search criteria 


Logo of scirepAboutEditorial BoardFor AuthorsScientific Reports
Sci Rep. 2017; 7: 43460.
Published online 2017 March 2. doi:  10.1038/srep43460
PMCID: PMC5333163

Understanding the emergence of modern humans and the disappearance of Neanderthals: Insights from Kaldar Cave (Khorramabad Valley, Western Iran)


Kaldar Cave is a key archaeological site that provides evidence of the Middle to Upper Palaeolithic transition in Iran. Excavations at the site in 2014–2015 led to the discovery of cultural remains generally associated with anatomically modern humans (AMHs) and evidence of a probable Neanderthal-made industry in the basal layers. Attempts have been made to establish a chronology for the site. These include four thermoluminescence (TL) dates for Layer 4, ranging from 23,100 ± 3300 to 29,400 ± 2300 BP, and three AMS radiocarbon dates from charcoal samples belonging to the lower part of the same layer, yielding ages of 38,650–36,750 cal BP, 44,200–42,350 cal BP, and 54,400–46,050 cal BP (all at the 95.4% confidence level). Kaldar Cave is the first well-stratified Late Palaeolithic locality to be excavated in the Zagros which is one of the earliest sites with cultural materials attributed to early AMHs in western Asia. It also offers an opportunity to study the technological differences between the Mousterian and the first Upper Palaeolithic lithic technologies as well as the human behaviour in the region. In this study, we present a detailed description of the newly excavated stratigraphy, quantified results from the lithic assemblages, preliminary faunal remains analyses, geochronologic data, taphonomic aspects, and an interpretation of the regional paleoenvironment.

Understanding the initial spread of anatomically modern humans (AMHs) out of Africa is a key goal for palaeoanthropologists. AMHs originated in Africa and spread across the Middle East into Eurasia and towards Australia and the Americas. These AMHs were the first humans to occupy the latter two continents, but they replaced other populations in Eurasia. Because few well-dated human remains are available to study this dispersal process, the spread of AMHs is best documented by the appearance of the early phase of their culture, inferred to be the Aurignacian (but see ref. 1). In western Eurasia, this technocomplex replaced the Mousterian, associated in this region with Neanderthals. This transition may have occurred at approximately 50 to 40 ka (see refs 2, 3, 4, 5, 6, 7).

One key area relevant to the dispersal process is Iran and Iraq, particularly the Zagros Mountains. Since the first survey of the Zagros by D. Garrod in 1930, Palaeolithic deposits and surface finds have been reported from a large number of caves, rockshelters, and open-air sites, but few of them have been fully excavated. Locally, the early phase of the technocomplex associated with early AMHs is known as the Baradostian. The early Upper Palaeolithic assemblages are also known as the Rostamian, which is defined as a bladelet-based technocomplex8,9,10,11. Although Conard et al. view the Rostamian as an industry distinct from the Baradostian11, both terms refer to the early Upper Palaeolithic in the Zagros region.

Many of the researchers who study materials from the Zagros agree that the lithic assemblages from this region share some features with assemblages from central Europe and the Levant. These include typo-technological characteristics of the Aurignacian tradition as well as inter-assemblage variability12,13,14. Olszewski and Dibble12, for example, proposed changing the name of the Baradostian to the ‘Zagros Aurignacian’ in light of the perceived similarities with Aurignacian material.

There is disagreement, however, regarding whether the Upper Palaeolithic evolved from earlier Mousterian industries in the region15. Some authors have proposed that the Baradostian might have developed locally from the Mousterian5,12,13,14,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32. On the one hand, recent work on two stratified assemblages from Warwasi and Yafteh support an in situ evolution of the Upper Palaeolithic from the local Mousterian15. That study, however, focused on only two assemblages; thus, the conclusions might not be fully applicable to the entire Zagros region. On the other hand, Tsanova15 raised doubts about whether the Iranian Zagros was the source of bladelet technology. However, the discovery of over 90 sites in the southern Zagros mostly associated with bladelet-based technologies—one of which dates to 40,000 cal. BP—suggests that the technology in the region featured a high degree of complexity5,8,9,10,11.

Additionally, the Zagros is more than 2,000 km long from northwest to southeast and up to several hundred kilometres wide from east to west27. Due to the lack of extensive surveys and archaeological excavations in the region, many aspects remain poorly understood. The latest typo-technological analyses on the lithic assemblage from the site of Ghar-e-Khar, for example, indicate the presence of multiple sites containing both Middle and Upper Palaeolithic sequences in the Zagros region. These findings confirm the potential for continued research into the Middle to Upper Palaeolithic transition33. However, only a few excavated sites contain uninterrupted archaeological sequences that include both Middle and Upper Palaeolithic deposits34. Some well-excavated sites, e.g., Yafteh, do not have Middle Palaeolithic occupation levels. Besides the reported sites in the Gilvaran and Ghamari caves35, Warwasi and Ghar-e-Khar are the only sites in the Iranian Zagros containing cultural remains belonging to both the Middle and Upper Palaeolithic (Fig. 1). To date, however, neither has been dated. Warwasi and Ghar-e-Khar were coarsely excavated (20 cm spits in Warwasi and 10–30 cm spits in Ghar-e-Khar). Chronometric control of the sites has been hampered by the poor preservation of organic material extracted from the archaeological sites and by political challenges and instability, which have made excavation work virtually impossible for more than 20 years. Here, we present the recently excavated and dated well-stratified sequence of Kaldar Cave, which documents the transition from the Middle to the Upper Palaeolithic.

Figure 1
The excavated sites containing Middle Palaeolithic and early Upper Palaeolithic sequences in the Zagros.


Site stratigraphy

Kaldar Cave is situated in the northern Khorramabad Valley at 48° 17′35″E longitude, 33°33′25″N latitude, and an elevation of 1,290 m above sea level. It is 16 m long, 17 m wide, and 7 m high. The potential of this site for excavation was first realized during a survey in 2010, when we started our regional study of the Khorramabad Valley as a goal-oriented research project. The first excavation35 was conducted in 2011–12.

The 2014–15 excavation focused on gaining a better understanding of the stratigraphy and obtaining samples for dating. We opened a 3 × 3 m trench near the entrance and kept a 50 cm bulk sample from the previous test pit (squares E5, E6, E7, F5, F6, F7, G5, G6 and G7) (Supplementary Fig. S1). The excavation was conducted using spits of 5 cm within each archaeostratigraphic unit, as well as 3D recording of all findings.

The excavated trench exposed an approximately 2-m (195-cm) section of the sedimentary deposit, which is characterized by five main layers. During fieldwork, distinctions within the layers were made according to minor sedimentological differences. Ongoing microstratigraphic research will provide a proper characterization of the sub-layers.

Layers 1 to 3 (including sub-layers 4 and 4II) consist of ashy sediment with a blackish-green colour containing both thick and thin angular limestone clasts. These layers varied in thickness from 60 to 90 cm and contained many phases dating to the Holocene: the Islamic and historical eras, Iron Age, Bronze Age, Chalcolithic, and Neolithic. However, due to the presence of some bioturbation in these layers, the phases were recognized only by a preliminary study of the potsherds, metal artefacts and some diagnostic lithic artefacts from the lower layers.

Layer 4 (including sub-layers 5, 5II, 6 and 6II) consists a silty but compact dark-brown sediment with cultural remains from the Upper and early Upper Palaeolithic. In the uppermost parts of this layer, two fireplaces made of clay were recovered and dated through thermoluminescence, yielding ages that ranged from 23100 ± 3300 to 29400 ± 2300 BP (Table 1). The dates obtained show that these fireplaces were made or re-used from existing older sediment from the upper part of this layer in the later stages of the Upper Palaeolithic. AMS radiocarbon dates of 38650–36750 cal BP, 44200–42350 cal BP, and 54400–46050 cal BP have been obtained from charcoal material located below this layer (Table 2).

Table 1
List of thermoluminescence dating results from Kaldar Cave.
Table 2
Radiocarbon results for charcoal samples from Kaldar Cave.

Layer 5 (including sub-layers 7 and 7II) consists of an extremely cemented reddish-brown sediment with some small angular limestone blocks and Middle Palaeolithic artefacts (Figs 2a,b and and3).3). To date, no radiometric data are available for this layer.

Figure 2
(a) North-south longitudinal projection of the materials from squares in line E. (b) West-east transversal projection of the materials from squares in line 6. Materials from Layers 1 to 3 have been projected together (blue). Projected separately are the ...
Figure 3
(Above) Stratigraphy (eastern section) along with transparent north-south longitudinal projection of the materials from squares in line E.

Bioturbation or disturbance was plotted, and sediment associated with the disturbance was removed without coordinating the finds, which were recorded as general finds with their approximate depths. Isolated evidence for intrusion below the Holocene layers was identified in a deep pit in square E7 in the upper part of the junction of sub-layers 5 and 5II. In the remainder of the site’s sequence, these layers are extremely hard and contain no evidence of bioturbation or disturbance. Heavy hammers and chisels were necessary to excavate these deposits (Supplementary Fig. S2). Consequently, we reached bedrock only in squares E6, E7, F6 and F7 (Supplementary Fig. S3).

Faunal and floral remains

Bioarchaeological remains recovered to date allow us to make some initial environmental inferences and to correlate the faunal and the lithic records to reconstruct human subsistence activities.

A small portion of the faunal assemblage from Kaldar Cave was previously described35, but the recent excavations have yielded new material—some of which is described in the supplemental information (Supplementary Information, Supplementary Fig. S11). The preliminary study of the small vertebrates from Kaldar Cave has identified 218 remains coming from Layer 4 (sub-layer 5II) and Layer 5 (sub-layer 7II), comprising rodents, squamate reptiles, and amphibians. The updated faunal list is given in Table 3. There is no indication of a faunal change coincident with the cultural change from Layer 5 to Layer 4.

Table 3
Distribution of the faunal taxa identified in Kaldar Cave, Layers 4 and 5.

Most of the amphibians and reptiles (Agamidae, Eryx and Elapidae) live in savannah, steppe and desert environments and feature lifestyles linked to warm arid areas in rocky or sandy environments. Pseudopus lives in dry and bushy environments, sometimes in open woodlands, but avoids dense forest areas. The two most abundant rodent species in both Layers 4 and 5 are Microtus gr. socialis and Meriones spp., indicating that the environment surrounding the cave was composed mainly of dry open areas, with some vegetation cover, as indicated by the presence of Gliridae and Murinae taxa in both layers. Cervus elaphus, Sus scrofa and Capreolus may have preferred the more closed and humid environments in the valley near the river, whereas Equus may have favoured more open environments somewhat farther away on the flood plain (which could not have been very wide). Additionally, Capra may have lived in the higher areas. The region surrounding the cave was likely relatively humid close to the river and drier farther away, i.e., more or less similar to the modern conditions.

The charcoal assemblage shows the presence of Prunus (Layers 4 and 5) and Salix (Layer 5). This would suggest the presence of tree cover composed of willows near the river and open woodland possibly composed of several species including plum trees farther away. The presence of these taxa support the interpretation of an open woodland under mild climatic conditions inferred from the other proxies.

The animal species present in Kaldar Cave originated long before the Late Pleistocene. Some of the species show changes during this period, but the material from Kaldar Cave is not yet sufficient to assess the evolutionary level of these species. Thus, from this perspective, the fauna has limited biochronological value at the scale needed here.

A preliminary taphonomic analysis of the small mammal assemblage has shown a high number of digested elements, suggesting predation activity. According to the different degrees of digestion observed in the remains (light, moderate and some heavy), a category 3 predator, such as the tawny owl (Strix aluco) or the Eurasian eagle owl (Bubo bubo36), might be responsible. Both species are compatible with the inferred habitat and are present in the area today37. Additionally, both have opportunistic hunting habits and are sedentary; therefore, their prey well represents the local ecosystem.

The large vertebrates in Layers 4 and 5 are represented by highly fractured bones and teeth. Only seven complete remains (8.2% in Layer 4 and 7.1% in Layer 5) were recovered (1 unciform of Capra, 1 coracoid of Testudo, 1 tarso-metatarso of Aves from Layer 4; 2 teeth, 1 sesamoid of Capra and 1 caudal vertebra of a small mammal from Layer 5). Remarkably, approximately half of both sets (44.7% and 50%, respectively) are shaft fragments. An analysis of the fracture edges (according to Villa and Mahieu38) shows that breakage of the bones occurred when they were fresh because most fracture delineations are curved or v-shaped (59.4%) and longitudinal (36.8%), with oblique angles (85.8%). Despite the high degree of fracturing of the large vertebrate bones and teeth in Layers 4 and 5, the assemblages on the whole appear to be well preserved. Post-depositional modifications were generally scarce in the Kaldar assemblage, except for black stains from manganese oxide deposits, which were found on 24.1% of the Layer 4 remains and on 30.9% of the Layer 5 remains, and the cemented sediment attached to surfaces, which were found on 18.2% of the Layer 5 remains. These modifications suggest alternating damp and dry periods in the cave during the formation of Layer 5. Furthermore, sub-aerial weathering (stage 1 according to Behrensmeyer39) has been identified in just one specimen in each of the layers.

Evidence of anthropogenic activity appears in three ways: cut marks, bone fracturing, and cremations (Supplementary Fig. S4). Cut marks were observed on thirteen specimens: five from Layer 4 and eight from Layer 5. The remains from Layer 4 are long bones (humerus of Caprini, two tibia fragments and one femur of Cervidae and a long bone shaft of an indeterminate mammal). The cut marks appear in the form of slicing and scraping marks, and all instances are located on the shaft portions, indicating the defleshing of the carcasses. In Layer 5, the elements with cut marks comprise one radius and one tibia of Capra, one rib and one caudal vertebra of indeterminate taxa and four indeterminate long bone fragments. The incisions on the rib fragment were located on the neck of the bone and were associated with disarticulation activities. The incisions on the caudal vertebrae were located in the central part of the bone. The positions of the marks suggest that they are related to skinning tasks. Other bones have cut marks in midshaft positions, indicating defleshing of the carcasses. The Capra radius with cut marks also had an impact point produced by the anthropogenic breakage of the bone.

Among the anthropogenic modifications of the bones in Kaldar Cave, the most important are the changes in coloration due to cremation, which is present in all layers. Fully 23.2% of the remains of the assemblage are burned (24.1% in Layer 4 and 21.8% in Layer 5). These remains include charred (black coloured, 34.4%) and rubefacted (brown and red coloured, 9.4%) bones. Bones with multiple colours are also common (53.2% of the burned bones). The most common combination is rubefacted (brown) and charred (black) colours (46.9%) on the same bone, although partially calcined (grey-blue-white colours) specimens are also present. The presence of multiple colours on the surface of the bones has been associated with meat cooking40. The distribution of the colours is homogeneous on the surface and affects the fracture edge and the cortical and medullar faces, suggesting that the bones were burned after they had been broken. According to several experimental studies40,41,42, the presence of multiple colours indicates that the bones (regardless of the size) experienced cremation damage when they were fresh and unburied. The origin of this modification may be related to cooking but may also be related to their use as fuel for the maintenance of fires, cleaning of the living floor, or accidentally building a fire near the location where the bones had been deposited.

Little carnivore activity is recorded by the assemblage. Three bones in Layer 4 (1 tibia of Cervidae and 2 indeterminate long bones) showed carnivore tooth marks (Supplementary Fig. S4e). It is difficult to determine the size of the carnivore because only a few tooth marks are recorded. However, the low frequency of these modifications suggests carnivores played a limited role in the formation and/or modification of assemblage.

The zooarchaeological results suggest that not only were the early AMHs that occupied Kaldar Cave among the first to come into contact with large Palaearctic mammals but that they also quickly adapted to exploiting them as a resource.

Lithic industry

The technological analysis of the archaeological samples associated with the Mousterian assemblage from Kaldar Cave (Layer 5 - sub-layers 7 and 7II) indicates that by-products (fragments and flake fragments) are the most common elements (12%) followed by retouched tools (10.8%), Levallois flakes (8.5%), cortical pieces (5.8%), Levallois blades (4%), Levallois points (2.4%), Levallois cores (0.8), other types of cores (0.8%) and hammerstones (0.4%). A large amount of debris (54.5%) is also present in the assemblage. The flakes are dominated by Levallois and cortical pieces, mostly with elongated morphologies and predetermined pointed shapes. Among the 82 flakes counted, 24.4% are cortical pieces, 24.4% are retouched, 20.7% have pointed shapes, 15.8% are broken, and 15.8% show enough major characteristics to be defined as a Levallois flake. Among the blade group, 37.1% are pointed in shape, 25.9% do not fit within a standard category, 14.8% are cortical, 11.1% are broken, 7.4% are retouched, and just one core (3.7%) was found. The retouched artefacts are dominated by marginal and broken retouched flakes (37%), Mousterian points (26%), different types of scrapers (24.1%), retouched points (5.6%), retouched blades (3.7%), Tayac points (1.8%), and limace (1.8%). The points (including Mousterian, Levallois, retouched and Tayac), along with pointed flakes and blades, comprise 11.4% of the entire assemblage in this layer. Among the material other than debris, the points and pointed elements comprise 25.1% of the assemblage in this layer. Mousterian points, Levallois points and retouched points comprise 2.8%, 2.4% and 0.6% of the assemblage, respectively. Not counting the debris, the Mousterian points and Levallois points comprise 6.2% and 5.3% the assemblage, respectively (Table 4).

Table 4
Quantified results of the lithics attributed to the Middle Palaeolithic Layer 5 of the 2014–2015 excavation season at Kaldar Cave.

The low number of cores (all exhausted) among the Mousterian assemblage from Kaldar Cave could be meaningful. This observation is in agreement with the techno-typological results from the Mousterian assemblage of the nearby Kunji Cave43. Given the notable scarcity of cores, the absence of refittable pieces, the large differences between the size of the tools and the size of the cores and their negative scars, and the condition of the cores that are exhausted, the chaîne opératoire is incomplete. Therefore, many of the artefacts were likely carried in from elsewhere (Fig. 4 and Supplementary Figs SI, S5, S9A & B, S10A & B).

Figure 4
Selection of Middle Palaeolithic Levallois pieces from Kaldar Cave (Layer 5).

In the Upper Palaeolithic lithic assemblages of Layer 4 (sub-layers 5 & 5II), bladelets dominate (13%), followed by blades (12.5%), retouched tools (5.1%), cortical pieces (4.4%), by-products (3.5%), bladelet cores (1.6%), undetermined cores (1.4%; including a centripetal core), pointed flakes, blanks, and other types of tools (a borer and point; all less than 1%), a blade core (0.2%) and finally a considerable amount of debris (56.4%). Within the bladelet categories, there is a good representation of twisted bladelets (14.3%). Among the retouched tools, Arjeneh points are abundant, but pointed pieces (including Tanged, retouched points, pointed blades and bladelets and Arjeneh points) are more numerous (54.5%) compared to other types of retouched tools (Figs 5, ,66 and and7,7, and Supplementary Figs S6, S7 and S8). Excluding the debris in this layer, the points and pointed elements comprise 11.2% of the entire assemblage. The next most abundant tools among the retouched pieces are the scrapers (including side-scraper, end scraper and nosed scraper), representing 18.2% of the tools. The number of flakes in this layer is very low (4.6% of the assemblage), and among the flakes, 3.7% are cortical flakes, 0.7% are pointed flakes and 0.2% are retouched flake (Table 5).

Figure 5
Selection of Upper Palaeolithic retouched pieces from Kaldar Cave (Layer 4).
Figure 6
Selection of Upper Palaeolithic blades and bladelets from Kaldar Cave (Layer 4).
Figure 7
Selection of Upper Palaeolithic cores from Kaldar Cave (Layer 4).
Table 5
Quantified results of the lithics attributed to the Upper Palaeolithic Layer 4 of the 2014–2015 excavation season at Kaldar Cave.

Despite the small size of the assemblage, a quick examination of the assemblage data from both Layer 5 (sub-layers 7 & 7II) and Layer 4 (sub-layers 5, 5II, 6 & 6II) shows a significant technological change from flake technology towards the production of blades and bladelets. However, to be more precise, a preliminary comparative analysis between the two layers was performed (Supplementary diagrams S1 to S5). In this analysis, we compared the weights and average values of metric measurements of various characteristics and attributes. The comparison of comparable categories was performed to provide meaningful results and to aid our interpretation of these two layers. Therefore, we compared Levallois cores vs. blade/bladelet cores, pointed blades vs. pointed blade/bladelets, and the retouched points, cortical pieces and cortical flakes (within the cortical pieces) from both the layers. Interestingly, our comparative analysis shows that significant differences are present among all the elements from the Middle and Upper Palaeolithic industries of Kaldar Cave. The weights and sizes of all the compared elements tend to be greater in Layer 5 than in Layer 4. The only exception was found within the retouched points. In this case, the average length and thickness are slightly greater for Layer 4 than for Layer 5.


Considerable efforts have been made to address fundamental questions concerning the cultural remains associated with AMHs. These attempts have been based mainly on Upper Palaeolithic lithic assemblages and their potential places of origin.

Based on technological comparisons between the lithic assemblages found in Europe and those found in the Zagros, some authors report close typological similarities between the two and further propose the latter region as the most probable source of the technology, with an east-to-west diffusion into Europe. Consequently, the Upper Palaeolithic tradition of the Zagros has been termed the ‘Zagros Aurignacian’12,13,14,19,20,21,23,24,25,26,27,44. Based on the techno-typological analysis of material from Warwasi, some also claim that the Baradostian (or Zagros Aurignacian) technology evolved from a local Mousterian foundation in the area.

In conflict with the statement by Tsanova15 that the Iranian Zagros cannot be the source of bladelet technology and cultural modernity as the Warwasi rockshelter lacks both radiocarbon dates and evidence of antecedent blade technology, strong evidence indicates that the Zagros assemblages are not merely blade-based. Over 90 sites contain evidence of clear blade(let) production (defined as the Rostamian tradition), and these tools are all similar to and associated with those from the well-stratified Ghare-Boof locality8,9,10,11.

Based on the detailed techno-typological analysis of the industries from Yafteh, some authors claim that the Baradostian technology of the Zagros is an Early Ahmarian-like technology45,46 and conclude that, on the basis of the available data, continuity from the Zagros Mousterian to the Zagros Aurignacian cannot be confirmed. However, based on the gradual transition from the Middle Palaeolithic to the Upper Palaeolithic at Warwasi and the technological and chronological analogies between the lower Baradostian at Yafteh and the Early Ahmarian, the Zagros region remains a potential candidate for the origin of the Aurignacian5,32. In a very recent typo-technological study on the Ghar-e-Khar lithic assemblages33, the authors estimated the potential of the area for future research. Nevertheless, in addition to the small sizes of the studied assemblages, methodological problems (e.g., using 10- to 30-cm arbitrary levels) during the excavation and the lack of absolute chronometric data might raise concerns similar to those for the material from Warwasi and cast doubt on the results from Ghar-e-Khar, which are not compelling. Thus, the hypothesis of Middle-to-Upper Palaeolithic continuity in Zagros and the possibility of a gradual transition are hard to assess due to the current state of the technological data33.

Similar to Yafteh, the Üçağizli sequence in Turkey provides some evidence of evolution from the Initial Upper Palaeolithic (IUP) into the early Upper Palaeolithic “Early Ahmarian”. Given the absence of Middle Palaeolithic underlying the IUP layers in Üçağizli, however, the site offers little to the discussion of the appearance of the IUP in the region47 (see also Shidrang32). Additionally, an IUP assemblage has also been discovered in Manot Cave, to the north of Mount Carmel48. The presence of both Mousterian and Baradostian cultural remains in Kaldar Cave and the recent chronometric data can be used to address many of the stated uncertainties associated with the transition process.

In regard to the terms “IUP”, “Aurignacian”, “Baradostian” and “Zagros Aurignacian”, our data from Kaldar Cave and other excavated localities35 support the arguments advanced by Kuhn and Zwyns49 with respect to the technological diversity within the assemblages and the long duration of the Upper Palaeolithic in Kaldar. We therefore avoid using the term “IUP” for this assemblage. On the other hand, we cannot simply assign the term “Aurignacian” to the assemblage based on certain similarities with assemblages from European sites. However, our observations and technological analysis of the Kaldar assemblage are in agreement with that of Olszewski12,13,14,19,20,21,22,23,24,44: certain similarities do exist, yet the Zagros industry differs from the purely European Aurignacian. Therefore, to us, the terms “Baradostian” or “Zagros Aurignacian” are more appropriate.

Notably, based on our earlier technological work35, the recent TL dates are older than we anticipated for the lithic assemblages of the uppermost part of Layer 4. These dates have led us to abandon the Epipalaeolithic designation we previously applied to these bladelet assemblages.

The AMHs in Kaldar Cave may have been among the first of their kind to interact with Palaearctic fauna. Thus, many of the species were new to them. In this part of Eurasia, the Palaearctic had an east-west-oriented southern border with the newly defined Saharan-Arabian biogeographic realm. The Zagros Mountains acted as an extension of the Palaearctic into the more southern realm50. However, it is not known whether the boundary between these realms occupied the same location during the Late Pleistocene. The presence of large mammals is indicative of the seasonality of the Palaearctic, but most of the reptiles have Saharan-Arabian affinities, and the rodents yield a mixed signal.

The fauna present in the mid-latitude Palearctic represent “interglacial” fauna, and similar faunas (albeit generally richer in species) occupied the area during previous interglacials. During glacial periods, these species survived in southern refugia, while cold-adapted species occupied the mid-latitudes. Iran may have acted as one of these refugia. Up to now, no typically glacial species has been recorded in Iran or other areas at similar latitudes. The Palaearctic mammal species recorded in Iran, and in particular in Kaldar Cave, are “interglacial”, suggesting the presence of at least temperate conditions. In contrast, the herpetofauna clearly indicates warm conditions. This combination is consistent with a position at the limit of the two biogeographic realms during climatic conditions similar to those of today. Furthermore, the study period is thought to correspond to MIS3, which had conditions similar to the modern climate. Because there is no indication of faunal change between layers 4 and 5, the available evidence suggests that the cultural change was not related to climatic or environmental changes.


Radiocarbon dating

Radiocarbon dating of the five charcoal samples (listed in Table 4) was performed at the Oxford Radiocarbon Accelerator Unit (ORAU). The samples were chemically cleaned using the acid-base-wet oxidation-stepped combustion (ABOx-SC) protocol (after Brock and Higham51, also see ref. 52) or a modification of the same. The ABOx-SC method was employed as it has been shown to remove contaminants from Palaeolithic-aged charcoal more efficiently than the routine acid-base-acid (ABA) protocol, often yielding significantly older dates (e.g. refs 51, 53, 54, 55, 56, 57, 58, 59). The analytical data obtained are shown in Table 4, and no data fall outside the expected ranges for well-preserved charcoal. The calibration of all the resulting AMS radiocarbon determinations was performed using the OxCal 4.2 software60,61 and the IntCal13 calibration curve62.

Among the seven charcoal samples submitted, five yielded enough material for AMS radiocarbon dating after chemical preparation (see Table 6). Only three of these, however, yielded reliable radiocarbon dates following a congruent age-depth pattern; the two others were substantially younger. This is almost certainly due to taphonomic influences. While it would be useful to incorporate the Palaeolithic-aged results into a Bayesian model, we cannot as we have too few results at this time. More dating work is currently underway, and we hope to be able to report new results in the future. In Fig. 8, we show the calibrated results for the Palaeolithic specimens (see Table 2 for the data).

Figure 8
Calibrated results of the Palaeolithic specimens.
Table 6
Analytical and sample data for the charcoal dated from Kaldar Cave.

TL dating

Thermoluminescence dating was performed on five heated samples (four heated sediments from two fire places in the upper most part of Layer 4 and one burnt flint from Layer 5) at the Research Centre for Conservation & Restoration of Cultural Relics of the Research Institute of Iranian Cultural Heritage (RICHT). At present, the samples from Layer 4 have successfully been dated. Three of the dated samples come from a fireplace within squares E6/7 and one from square E5 (Table 1).

For the sample preparation and instrumentation, the outer surface (3 mm) of the samples was removed. To account for the alpha radiation contribution to the natural dose measurements, the fine grain technique is used (ibid). Alpha radiation travels an extremely short distance in heated objects (approximately 25 μm63). Thus, we used grains less than 10 μm in size. The samples were crushed and treated with 10% HCl to remove carbonates and organic material. Then, all samples were washed with distilled water and then with acetone. Finally, the grains were suspended in acetone and deposited on aluminium discs that were 10 mm in diameter and 0.5 mm in thickness.

The TL measurements were performed using an ELSEC7188 instrument. The potassium contents of the samples were determined by flame photometry. To determine the contributions from U and Th, the “pairs” technique was used; thus, the dose rate was measured using a 7286 low-level alpha counter64. External dose rates were measured by in situ dosimetry65. The CaF2 TL-Dosimeter was located in site for 36 days. These values were calculated for different levels, up-level: 0.787 mGy/a, down-level: 0.660 mGy/a. Measurements of the water content and fading test for all samples were considered (Table 1).


The newly excavated sequence in Kaldar Cave provides evidence for the replacement of the Mousterian industry, usually associated with Neanderthals, by the Baradostian industry, similar to the Aurignacian, which is unique to anatomically modern humans. Radiocarbon dates suggest that this may have occurred prior to 49,200 ± 1800 BP, probably during the relatively warm MIS3. The faunal evidence is consistent with the replacement occurring during MIS3 and does not support a coincident climatic change. Kaldar Cave is situated in the southernmost part of the Palaearctic biogeographic realm. Evidence from Kaldar Cave is among the oldest to show that AMHs were capable of exploiting the Palaearctic fauna and were thus well adapted to this new environment, which they colonized shortly after the period of time recorded in the cave.

Excavations at Kaldar Cave have yielded evidence for Baradostian (Layer 4) and Mousterian assemblages (Layer 5) in stratigraphic superposition. This is an exceptional find in the Zagros. The preliminary technological analysis on the lithic industry from both layers indicates a clear shift from flake production to blade and bladelet technology. Furthermore, despite the small size of the lithic assemblage so far, the quantitative comparative analysis shows a significant difference between elements within the Middle and Upper Palaeolithic layers. The homogeneity of the differences between all the compared elements—that is to say, the greater weight and size of the items in the Mousterian assemblage compared to those of the Upper Palaeolithic assemblage—could be a reliable foundation for interpretation and understanding the two industries.

We have obtained new chronometric data from the site. Four TL dates from the uppermost Layer 4 revealed ages that ranged from 23100 ± 3300 to 29400 ± 2300 BP.

The three 14C dates from Layer 4 and sub-layers 5 and 5II produced results in the ranges of 38650–36750 cal BP, 44200–42350 cal BP, and 54400–46050 cal BP, respectively (all at 95.4% probability). The wide chronometric ranges and few dates do not allow us to make a confident and precise assessment of the age of the transition to the Upper Palaeolithic. Further work is needed to refine the chronology.

In addition to the presence of a clear Mousterian industry in the >0.5-m-thick Layer 5 and despite the need for more chronometric data, the obtained dates from the lower part of the Upper Palaeolithic sequence in Kaldar Cave are some of the earlier dates attributed to a lithic industry produced by AMHs in western Asia. Although we do not intend to challenge the Levantine dispersal theory, previous work has noted that the Aurignacian may not have originated in only one area22 (see also Groucutt66). It has been suggested that the ages of the so-called “transitional” or Initial Upper Palaeolithic layers at Ksar Akil may represent that the transition from the Middle to Upper Palaeolithic in this area (and possibly in the wider northern Levant) occurred later than previously estimated. This finding would cast doubt on the assumed singular role of the region as an origin for human dispersal into Europe67.

Another important clue derived from the preliminary quantified results of the Mousterian and Upper Palaeolithic lithic industries in Kaldar Cave is the high percentage of points and pointed elements in both the layers. This abundance may indicate that the site functioned as an important hunting camp in the Zagros Mountains during both the Middle and Upper Palaeolithic times. This hypothesis appears to be supported by the zooarchaeological evidence. Hence, Kaldar Cave provides one of the oldest examples of modern human existence in this part of the world and provides data on how these populations coped with the Palearctic climatic and environmental situations, which were new to them.

To reach a consensus regarding the Middle to Upper Palaeolithic transition/continuity, several lines of evidence are required. Indeed, accurate information and maximum control of the context, including careful sampling for chronometric dating from well-stratified sites and detailed techno-typological analysis, are crucial factors. Our understanding of the behavioural dimension of the transitional phenomenon would also benefit from more excavations using multidisciplinary methods, including spatial analysis and functional aspects.

Additional Information

How to cite this article: Bazgir, B. et al. Understanding the emergence of modern humans and the disappearance of Neanderthals: Insights from Kaldar Cave (Khorramabad Valley, Western Iran). Sci. Rep. 7, 43460; doi: 10.1038/srep43460 (2017).

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Material

Supplementary Information:


We would like to thank the directors of RICHT and ICAR (Seyed Mohamad Beheshti and Hamide Choubak) for their support and for issuing us the necessary permission to conduct this excavation season. We also thank Kamran Ahmadi (director of Research Centre for Conservation & Restoration of Cultural Relics of RICHT) for supporting us with the TL dating. We thank A. Mohajeri Nezhad, former director of Lorestan Cultural Heritage Organization for all his supports during the field-work. The unsparing efforts by M. Jayez, N. Hashemi, I. Beheshti, S. Azadbakht, S. Horshid, M. Hashemi, F. Bakhtiari, Z. Feiz, R. Khazaeli and A. Baharvand, who participated in a period or full season of the excavation, are greatly appreciated. This research is conducted in the framework of a signed scientific agreement between RICHT and IPHES. This project is funded by IPHES and the Department of Prehistory at the University of Liege. Radiocarbon dating was undertaken through the ERC-funded PalaeoChron Project, and the TL dating was performed in conjunction with the RICHT. This research was supported by the projects CGL2015-65387-C3-1-P (MINECO/FEDER), SGR 2014-899 (AGAUR, Generalitat de Catalunya), 2014 PFRURV-B2-17 and 2015 PFR-URV-B2-17 (Universitat Rovira i Virgili), and developed within the framework of CERCA Programme/Generalitat de Catalunya. L. Tumung and M. Fernández-García are beneficiaries of PhD scholarships founded under the Erasmus Mundus Programme – International Doctorate in Quaternary and Prehistory. A. Picin is the beneficiary of a Humboldt Postdoctoral Research Fellowship from the Alexander von Humboldt Foundation. B. Bazgir is a beneficiary of a predoctoral scholarship from the Fundación Atapuerca. J. van der Made contributed to the project CGL20l5 65387-C3-3-P (MINECO/FEDER).


The authors declare no competing financial interests.

Author Contributions B. Bazgir, A. Ollé, and L. Tumung investigated the site in detail, studied the lithic industry, and monitored the data recovered from the stratigraphy. L. Becerra-Valdivia, K. Douka and T. Higham performed the 14C dating and helped revise the draft. J. van der Made and D. Arceredillo studied the macrofauna. A. Picin collaborated on the early phases of writing the draft and contributed to some excavation days. P. Saladié performed the zooarchaeological studies. J.M. López-García, M. Fernández-García and I. Rey-Rodríguez studied the small mammals. H.A. Blain studied the amphibians and reptiles. E. Allué studied the charcoal assemblage. F. Bahrololoumi and M. Azimi performed the TL dating. M. Otte and E. Carbonell provided the research strategy and acted as the senior archaeological advisors. B. Bazgir wrote the paper and all the co-authors contributed to the final version of the manuscript.


  • Benazzi S. et al. . Early dispersal of modern humans in Europe and implications for Neanderthal behaviour. Nature 479, 525–528 (2011). [PubMed]
  • Bischoff J. L. et al. . Dating of the Basal Aurignacian Sandwich at Abric Romaní (Catalunya, Spain) by Radiocarbon and Uranium-Series. J. Arch. Sci. 21, 541–551 (1994).
  • Higham T. et al. . The timing and spatiotemporal patterning of Neanderthal disappearance. Nature 512, 306–309 (2014). [PubMed]
  • Nigst P. R. et al. . Early modern human settlement of Europe north of the Alps occurred 43,500 years ago in a cold steppe-type environment. PNAS. 111, 14394–14399 (2014). [PubMed]
  • Otte M., Shidrang S., Zwyns N. & Flas D. New radiocarbon dates for the Zagros Aurignacian from Yafteh cave. Iran. J. Hum. Evol. 61, 340–346 (2011). [PubMed]
  • Talamo S., Soressi M., Roussel M., Richards M. & Hublin J. J. A radiocarbon chronology for the complete Middle to Upper Paleolithic transitional sequence of Les Cottés (France). J. Arch. Sci. 39, 175–183 (2012).
  • Wood R. E. The chronology of the earliest Upper Paleolithic in northern Iberia: New insights from L’Arbreda, Labeko Koba and La Viña. J. Hum. Evol. 69, 91–109 (2014). [PubMed]
  • Ghasidian E. Azadi A., Heydari S. & Conard N. Late Paleolithic cultural traditions in the Basht region of the southern Zagros of Iran. In Iran Paleolithic (ed. Otte M., Biglari F., Jaubert J.) 125–140 BAR S1968, Archaeopress (2009).
  • Ghasidian E. Small River Pebbles and Late Pleistocene Craftsman: The Dawn of Bladelet Production in the Southern Zagros Mountains of Iran. Hugo Obermaier Gesellschaft, Leipzig. Abstract Book (2010).
  • Conard N. J. & Ghasidian E. The Rostamian cultural group and the taxonomy of the Iranian Upper Paleolithic. In Between sand and sea: The archaeology and human ecology of southwestern Asia, Festschrift in honor of Hans-Peter Upermann (ed. Conard N. J. Drechsler P. & Morales A.) 33–52 Tübingen, Kerns Verlag (2011).
  • Conard N. J., Ghasidian E. & Heydari S. The Paleolithic of Iran. In Ancient Iran. (ed. Potts D. T.) 29–48 Oxford press (2013).
  • Olszewski D. & Dibble H. L. The Zagros Aurignacian. Curr. Anthropol, 35, 68–75 (1994).
  • Olszewski D. The Early Upper Paleolithic in the Zagros Mountains. In Dorothy Garrod and the Progress of the Paleolithic: Studies in the Prehistoric Archaeology of the Near East and Europe (ed. Davies W. & Charles R.) 167–180 Oxbow Books (1999).
  • Olszewski D. Ruminations on the Early Upper Paleolithic and a Consideration of the Zagros Aurignacian. In Questioning the Answers: Resolving Fundamental Problems of the Early Upper Paleolithic (ed. Hays M. & Thacker P.) 79–90BAR IS1005, Archaeopress (2001).
  • Tsanova T. The beginning of the Upper Paleolithic in the Iranian Zagros. A taphonomic approach and techno-economic comparison of Early Baradostian assemblages from Warwasi end Yafteh (Iran). J. Hum. Evol. 65, 39–64 (2013). [PubMed]
  • Hole F. Archaeological survey and excavation in Iran, 1961, Science. 137, 524–526 (1962). [PubMed]
  • Hole F. The Paleolithic Culture Sequence in Western Iran. In Actes du VII Congrès de l’UISPP, Bratislava, 1967, 286–293 (1970).
  • Hole F. & Flannery K. V. The prehistory of Southwestern Iran: a preliminary report. Proc. Prehist. Soc. 22, 147–206 (1967).
  • Olszewski D. The Late Baradostian Occupation at Warwasi Rockshelter, Iran. In The Paleolithic Prehistory of the Zagros-Taurus (eds. Olszewski ,. D. I. & Dibble H. L.) 187–205. Philadelphia: University Museum Press (1993).
  • Olszewski D. Carinated tools, cores and mobility: The Zagros Aurignacian. In Tools versus Cores Alternative Approaches to Stone Tool Analysis (ed. McPherron S. P.) 92–106 Cambridge scholars publishing (2007a).
  • Olszewski D. Issues in the Development of the Early Upper Paleolithic and a “Transitional” Industry from the Zagros Region. In New Approaches to the Study of Early Upper Paleolithic “Transitional” Industries in Western Eurasia: Transitions Great and Small (ed. Riel-Salvatore J. & Clark G. A.) 131–142 BAR IS1620, Archaeopress (2007b).
  • Olszewski D. Whither the Aurignacian in the Middle East? Assessing the Zagros Upper Paleolithic. In Iran Paleolithic (ed. Otte M., Biglari F., Jaubert J.) 39–45 BAR S1968, Archaeopress (2009).
  • Olszewski D., Dibble H. L. (eds.). The Paleolithic prehistory of the ZagrosTaurus. The University Museum, University of Pennsylvania, University Museum Monograph 83 (1993).
  • Olszewski D. & Dibble H. L. To be or not to be Aurignacian: the Zagros Upper Paleolithic. In Towards a definition of the Aurignacian (ed. Bar-Yosef O., Zilhao J.) 355–373 (2006).
  • Otte M. & Kozlowski J. Constitution of the aurignacian through Eurasia, In The chronology of the Aurignacian and of the transitional technocomplexes: dating, stratigraphies, cultural implications (ed. Zilhao J. & d’Errico F.) 19–28 Portugese Institute of Archaeology PIA (2003).
  • Otte M. & Kozlowski J. The significance of the Baradostian in the origin of the upper Paleolithic of Eurasia. L’Anthrop 107, 395–406 (2004).
  • Otte M. & Kozlowski J. Origins du paleolithique superieur en Asieoccidentale. In Iran Paleolithic (ed. Otte M., Biglari F., Jaubert J.) 57–72 BAR S1968, Archaeopress (2009).
  • Otte M. et al. . The Aurignacian in the Zagros region: new research at Yafteh Cave, Lorestan, Iran antiquity 81, 82–96 (2007).
  • Otte M., Shidrang S. & Flas D. The Aurignacian of Yafteh Cave and Its Context (2005–2008 excavations). ERAUL 132 (2012).
  • Shidrang S. The Early Upper Paleolithic ornamental objects from Yafteh Cave and Pa Sangar Rockshelter, Lurestan. Iranian J. Arch. Hist. 41, 38–44. (In Farsi, with an English abstract.) (2007a).
  • Shidrang S. The early Upper Paleolithic lithic assemblages from F 15 test pit (2005), Yafteh Cave, Iran: a typo-technological study. Master thesis, University degli studi di Ferrara, Italy (2007b).
  • Shidrang S. Middle East Middle to Upper Paleolithic Transition Industries. Springer Science+business media New York. 4894–4907 (2013).
  • Shidrang S. & Biglari F. Continuity and Change in the Late Pleistocene Lithic Industries of the Central Zagros: A Typo-Technological Analysis of Lithic Assemblages from Ghar-e Khar Cave, Bisotun, Iran. Arch. Ethnol. Anthropol. Eurasia 44, 27–38 (2016).
  • Vahdati Nasab H. Paleolithic Archaeology in Iran. Intern. J. Humanities 18, 63–87 (2011).
  • Bazgir B. et al. . Test excavations and initial results at the Middle and Upper Paleolithic sites of Gilvaran, Kaldar, Ghamari caves and Gar ArjeneRockshelter, Khorramabad Valley, western Iran. CR Palevol. 13, 511–525 (2014).
  • Andrews P. Owls, caves and fossils: predation, preservation and accumulation of small mammal bones in caves with an analysis of the Pleistocene cave faunas from Westbury-Sub-Mendip, Somerset, U.K. University of Chicago Press (1990).
  • Firouz E. The complete fauna of Iran I.B.Tauris (2005).
  • Villa P. & Mahieu E., Breakage patterns of human long bones. J. Hum. Evol. 20, 1–22 (1991).
  • Behrensmeyer A. K. Taphonomic and Ecologic Information from Bone Weathering. Paleobiology, 4, 150–162 (1978).
  • Bennett J. L. The thermal alteration of burnt bones. J. Arch. Sci. 26, 1–8 (1999).
  • Stiner M., Houze R. A. & Yuter S. E. Climatological characterization of three-dimensional storm structure from operational radar and rain gauge data. J. Appl. Meteorol 34, 1978–2007(1995).
  • Yravedra J. & Uzquiano P. Burnt bones assemblages from El Esquilleu cave (Cantabria, Northern Spain): deliberate use for fuel or systematic disposal of organic waste? Quat. Sci. Rev. 68, 175–190 (2009).
  • Baumler M. F. & Speth J. D. A Middle Paleolithic assemblage from kunji cave, Iran. In The Paleolithic prehistory of the Zagros-Taurus (ed. Olszewski D. & Dibble H.) 1–74 The University Museum of the University of Pennsylvania (1993).
  • Otte M. Central Asia as a Core Area: Iran as an Origin for the European Aurignacian. Int. J. Society of Iranian Archaeologists 1, 27–32 (2014).
  • Shidrang S. A typo-technological study of an Upper Paleolithic collection from Sefid-Ab, Central Iran. In Iran Paleolithic (ed. Otte , M., Biglari F., Jaubert J.) 73–84 BAR S1968, Archaeopress (2009).
  • Bordes J. G. & Shidrang S. The Baradostian sequence of the Yafteh Cave. A typo-technological lithic analysis based on the Hole and Flannery collection. In The Aurignacian of Yafteh cave and it’s context (2005–2008 excavations). 27–39 ERAUL 132 (2012).
  • Kuhn S. L. et al. . The early Upper Paleolithic occupations at Üçağizli Cave (Hatay, Turkey). J. Hum. Evol. 56, 87–113 (2009). [PubMed]
  • Hershkovitz I. et al. . Levantine cranium from Manot Cave (Israel) foreshadows the first European modern humans. Nature 520, 216–219 (2015). [PubMed]
  • Kuhn S. L. & Zwyns N. Rethinking the initial Upper Paleolithic. Quat. Intern. 347, 29–38 (2014).
  • Holt B. G. et al. . An update of Wallace’s zoogeographic regions of the world. Science, 339, 74–78 (2013). [PubMed]
  • Brock F. & Higham T. AMS radiocarbon dating of Paleolithic-aged charcoal from Europe and the Mediterranean Rim using ABOx-SC. Radiocarbon, 51, 839–846 (2009).
  • Bird M. et al. . Radiocarbon dating of “old” charcoal using a wet oxidation, stepped-combustion procedure. Radiocarbon, 41, 127–140 (1999).
  • Bird M. et al. . Radiocarbon dating from 40 to 60 kaBP at Border Cave, South Africa. Quat. Sci. Rev. 22, 943–947 (2003).
  • Douka K., Higham T. & Sinitsyn A. The influence of pretreatment chemistry on the radiocarbon dating of Campanian Ignimbrite-aged charcoal from Kostenki 14 (Russia). Quatern. Res. 73, 583–587 (2010).
  • Higham T. et al. . Radiocarbon dating of charcoal from tropical sequences: results from the Niah Great Cave, Sarawak, and their broader implications. J. Quatern. Sci. 24, 189–197 (2009a).
  • Higham T. et al. . Problems with radiocarbon dating the Middle to Upper Paleolithic transition in Italy. Quat. Sci. Rev. 28, 1257–1267 (2009b).
  • Santos G. et al. . A revised chronology of the lowest occupation layer of Pedra Furada Rock Shelter, Piauí, Brazil: the Pleistocene peopling of the Americas. Quat. Sci. Rev. 22, 2303–2310 (2003).
  • Turney C. S. et al. . Early human occupation at Devil’s Lair, southwestern Australia 50,000 years ago. Quatern. Res. 55, 3–13 (2001).
  • Wood R. et al. . Testing the ABOx-SC method: dating known-age charcoals associated with the Campanian Ignimbrite. Quatern. Geochronol. 9, 16–26 (2012).
  • Bronk Ramsey C. Bayesian analysis of radiocarbon dates. Radiocarbon 51, 337–360 (2009a).
  • Bronk Ramsey C. Dealing with outliers and offsets in radiocarbon dating. Radiocarbon 51, 1023–1045 (2009b).
  • Reimer P. J. et al. . IntCal13 and Marine13 radiocarbon agecalibration curves 0-50,000 years cal BP. Radiocarbon 55, 1869–1887 (2013).
  • Aitken M. J. An Introduction to Optical Dating. The Dating of Quaternary Sediments by the Use of Photon-Stimulated Luminescence Oxford University Press (1998).
  • Aitken M. J. Thermoluminescence dating Academic (1985).
  • Martini M., Piccinini G. & Spinolo G. A new dosimetric system for g-ray dose-rate determination in soils. PACT J. 9, 87–94 (1983).
  • Groucutt S. et al. . Stone tool assemblages and models for the dispersal of Homo sapiens out of Africa. Quat. Intern. 382, 8–30 (2015).
  • Douka K., Bergman C. A., Hedges R. E. M., Wesselingh F. P. & Higham T. F. G. Chronology of KsarAkil (Lebanon) and Implications for the Colonization of Europe by Anatomically Modern Humans. PloS One. 8, e72931 doi: (2013).10.1371/journal.pone.0072931 [PMC free article] [PubMed] [Cross Ref]

Articles from Scientific Reports are provided here courtesy of Nature Publishing Group