The Aurignacian is one of the first cultural-technological traditions commonly associated with the expansion of Homo sapiens in Europe. Early Homo sapiens demographics across the continent are therefore typically inferred using the distribution of Aurignacian assemblages. Western Romania has been used as a tie-point to connect the well-researched lithic assemblages from the eastern Mediterranean and Western Europe through its early Homo sapiens fossils. However, Romania’s archeological record remains underexplored thereby hindering our ability to directly connect better understood regions through time and space. Here we report on excavations from the open-air Middle/Upper Paleolithic site of Românești-Dumbrăvița I in southwestern Romania. Three stratified Paleolithic assemblages were extensively excavated within a 1-m-thick eolian-deposited sequence. Spatial, geochemical, raw material, techno-typological, and use-wear analysis of the site reveal patterns of artifact configuration, resource exploitation, fire history, knapping objectives, and functionality. Taken together, Românești-Dumbrăvița I is the first well-contextualized archeological site in close spatiotemporal proximity to many early, well-preserved human fossils and in East-Central Europe.
The dispersal of early modern humans into Europe is one of the main debates in paleoanthropology. A main hypothesis indicates that the Danube was a main axial migration route for modern humans to move into the upper catchment where behavioral developments subsequently took place and spread to other parts of Europe1,2,3,4,5. This scenario is intimated by early archaeological remains in the Upper and Lower Danube basins and corroborated by paleoanthropological remains and ancient mitochondrial DNA6,7,8. Still, a lack of robust archeological evidence from the middle of the continent makes it difficult to verify these models and understand how early Homo sapiens consolidated their presence across Europe9,10.
Western Romania is an important region for testing Homo sapiens dispersal hypotheses as it lies along the Danube Valley and astride some of the earliest known associated Aurignacian sites in Central Europe and the hypothesized Initial Upper Paleolithic progenitor industries in Moravia and Bulgaria11,12,13. It has additionally furnished some of the oldest undisputed Homo sapiens fossils in Europe14,15,16 along with associated fossil footprints and figurative cave art17,18.
Still, Romania’s Aurignacian has been neglected in Pan-European discussions even though assemblages have been known for decades19,20,21. Of particular importance are the sites in Banat region that include Coșava-Cuca, Temerești-Dealul Vinii, and Românești-Dumbrăvița (hereafter Românești)22,23. Further afield but still in Banat are Tincova, Crvenka-At (Serbia), and Tabula Traiana Cave, sites that have additionally contributed to Aurignacian chronology and settlement patterns in East-Central Europe24,25,26,27,28,29,30,31. Banat sites are of particular importance because they are the closest contemporary sites to the early Homo sapiens findspots at the Peștera cu Oase, Peștera Muierilor, and Peștera Cioclovina (Fig. 1b)14,15,16,32.
Interpretations of Banat sites and their role in Homo sapiens dispersals have been hindered by unclear stratigraphic proveniences, few radiometric ages, small assemblage sizes, and a lack of contextual information. Even though Românești has been established as a key early Upper Paleolithic (eUP) site in East-Central Europe33,34,35,36,37, five main factors have limited Românești’s integration into transcontinental narratives:
Broadscale proveniencing, selective artifact recovery and/or the small-scale extent of past excavations have restricted inferring the degree and nature of post-depositional processes and anthropogenic spatial arrangements.
Abundant burned artifacts absent of combustion features have hindered appraisals of on-site anthropogenic fire activity to evaluate settlement dynamics.
Unknown raw material proveniences have impeded assessments of past regional landscape use and the site’s connection to wider mobility patterns.
This paper presents results of new excavations from Românești during 2016–19 describing site-level spatial investigations, sediment geochemistry, fire residue analysis, techno-typology, raw material economy, and use-wear to provide a detailed assessment of occupation and technological activities at an open-air Aurignacian site in a key part of Europe. The results provide the only cultural link to the early, well-preserved fossils from a region in Europe that is otherwise devoid of secure archaeological sites.
The site of Românești is situated on a river terrace outside Românești village near the confluence of the Bega Luncani and Bega Poieni rivers in Banat of Romania (Fig. 1a). The site comprises two localities, Românești I and II (80 m apart) excavated between 1960–1964 and 1967–197222.
Românești I was excavated to an average depth of 1.3 m across 450 m2 (Supplementary Fig. 1). Six Paleolithic levels were reported though they were never found in full succession. The lowest level (I) was attributed to a “Quartzite Paleolithic”, being manufactured exclusively on Quartzite, and typified by Mousterian points, naturally backed knives, and Upper Paleolithic forms (prismatic cores and endscrapers). The subsequent four levels (II–V) were attributed to the Banat Aurignacian. Level VI was typified by Aurignacian elements alongside later Upper Paleolithic artifacts such as backed bladelets, Gravette points and scalene triangles. No faunal remains or features were recovered from the site.
Research resumed following the discovery of the antiquity of the Homo sapiens fossils in Banat14,42 and the re-examination of the nearby Tincova assemblage43,44. Test pits were installed at Românești in 2009–2010 to validate the stratigraphy, evaluate the frequency of smaller artifacts, obtain absolute ages, and frame the site within a wider European context34. Radiometric dates from optically stimulated luminescence (OSL) and thermoluminescence (TL) bracketed the Aurignacian artifacts to between 42.1 and 39.1 ka ago35.
5655 piece-provenienced artifacts were recovered in three discrete find layers associated with the Epigravettian, Aurignacian, and Middle Paleolithic (Fig. 2). For the Aurignacian layer in Geological Horizon (GH) 3 (see “Methods”), lithics show no visible trend in size sorting according to length and weight (Figs. 2, 3).
For GH3, a density plot showed an inhomogeneous spatial distribution of artifacts. A chi-squared (χ2) test indicated that for most excavation units, the number of lithics exceeded the range of expected values of a random pattern with the highest deviation in unit P102 (R = 15) confirming an inhomogeneous distribution (χ2 (0, N = 3737) = 1216.5, p < 0.01; Supplementary Fig. 2). A K function demonstrated that artifacts were clustered at short distances (r < 50 cm) and were randomly distributed thereafter (Supplementary Fig. 3). A scan test uncovered four clusters (CL1–4) with radii ranging between 17 and 40 cm (Fig. 3d). The relative probabilities showed substantial associations between bladelets and cores in CL1, cores and core trimming elements in CL3, and blades and bladelets in CL4 (Supplementary Fig. 2).
A nearest neighbor equality function demonstrated a positive spatial correlation in the distribution between bladelets–cores and partially between blades–bladelets (Supplementary Fig. 3) indicating that for example, between bladelets and cores, bladelets tend to be closer to other bladelets whereas cores tend to be closer to other cores. In the case of blades and bladelets, a positive correlation occurred only at short distances (between 3 and 10 neighboring artifacts). Similarly, blades and bladelets were positively associated but only after six neighboring artifacts.
Black carbon identification
Fire residue input, i.e., black carbon (BC) content, was found in all samples with amounts ranging from 3.7 ± 1.2 to 72.6 ± 7.0 g C kg−1 Corg. The lowest BC contents were detected in GH1 with an average of 10.7 ± 2.1 g C kg−1 Corg, followed by GH2-3 with 16.8 ± 3.1 g C kg−1 Corg (Fig. 4a). Both GH2 and GH3, with dense accumulations of lithics, showed increased BC contents (18.4 ± 3.5 and 33.0 ± 5.7 g C kg−1 Corg, respectively). GH2-3 and GH3 each had one outlier with high BC contents (GH2-3: 65.6 ± 11.3 g C kg−1 Corg; GH3: 72.6 ± 7 g C kg−1 Corg). Fluctuations in BC were low within horizons (GH1, GH2 and GH2-3) apart from GH3, which showed comparatively high heterogeneity in BC values across the profile.
Benzene polycarboxylic acids (BCPA) composition ideally allows the reconstruction of past fire regimes and changes in fire temperature and is expressed by the ratio of pentacarboxylic acid to mellitic acid (B5CA/B6CA)45. The lowest B5CA/B6CA ratio values were found in GH2 with an average of 0.7 ± 0.1, followed by GH1 and GH2-3 (0.8 ± 0.1 and 0.9 ± 0.1); GH3 had the highest ratio with 1.0 ± 0.1 (Fig. 4b).
Raw materials and MAN analysis
The GH3 lithic assemblage is composed of local shear zone related cherts supplied from locations close to faults, local gravels, and creeks as well as quartzite and other hard metamorphic rocks from the nearby gravels. Non-local raw materials notably “black flint” (non-local fault chert) and greyish-bluish chert (other siliceous rocks; Supplementary Fig. 4) are also present.
Six main raw material categories were distinguished based on macroscopic criteria and thin sections (Supplementary Fig. 4)46: Local shear zone related cherts found in the surveyed area (Poieni-Pietroasa fault chert, Căldărilor fault chert, ductile shear zone related cherts), metamorphic hard rocks found in the surveyed area (quartzite, vein quartz, graphitic quartz), non-local fault cherts, diagenetic cherts (very fine-grained diagenetic cherts, bioclastic cherts, non-skeletal cherts, siliceous mudstone/marlstone), obsidian and other siliceous rocks, indeterminate cherts, and (partially) burned rocks. Sr and Zr from an obsidian blade found in GH3 matches the obsidian outcrops of the C1b cluster from Cejkov, Slovakia 300 km away but could also fit within the C1a cluster 7 km away in Vinicky, Slovakia (Fig. 5)47.
2942 lithics were used in a Minimum Analytical Nodules Analysis (MANA) as the lithologies of some small, burned, or indeterminate raw materials could not be confidently assigned to a Minimum Analytical Nodule (MAN; Supplementary Fig. 4). Artifacts were grouped into 194 MANs (Supplementary Fig. 4), further characterized by the number of artifacts contained in a group (Supplementary Fig. 5): single item MANs (1 artifact) and multiple item MANs (two to tens of artifacts). The non-local raw materials from the assemblage are represented by single item MANs or multiple item MANs (< 10 artifacts) introduced as end-products and preformed cores and a few multiple item MANs (> 40 artifacts) introduced as preformed cores supplied from distances ranging from 13–60 km. MANs were assigned to archeological layers based on the GH of most of the artifacts in a MAN (some contained artifacts from other GHs) and/or the presence of culturally specific tool types. The MANs were also confirmed by artifact refits in the same group that were performed during categorization (Figs. 2e,3d). Single item MANs from both assemblages contained tools, debitage, or in some isolated cases, knapping waste products, suggesting that other artifacts of that MAN were outside of the excavation area.
Cores primarily produced blades and bladelets or just bladelets on narrow longitudinal faces with frontal knapping (narrow-fronted) or on broad longitudinal faces with semi-circumferential knapping (semi-rotating; Fig. 6c; Supplementary Table 1). Narrow-fronted cores were prevalently initiated on large flakes, while semi-rotating cores were manufactured on nodules or cobbles. 50% of cores preserved cortex (Supplementary Table 1). Semi-rotating cores preserved cortex on the posterior faces while narrow-fronted cores preserved cortex on the lateral faces. Cores primarily had single and plain unipolar striking platforms where overhang was frequently abraded and knapping angles were moderately acute to acute.
In GH3, debitage is equally divided between flakes and blades/bladelets (Table 1). Debitage platforms were mainly plain, linear, or punctiform (Supplementary Table 2). Bulbs were mostly diffused among blades and bladelets, while flakes were diffused and pronounced. Overhang abrasion was ubiquitous, but dominant among blade and bladelet blanks. Lipping was infrequent though well represented among blade blanks.
Most debitage was non-cortical, especially among blades and bladelet blanks. Extensive patches of cortex were only on initialization flakes. Cortical debitage was primarily semi-cortical found on lateral and distal positions. Negatives, preserving at least undisturbed width, were present on 35% of the debitage and were unidirectional; most showed a final bladelet removal.
Blade and bladelet profiles were primarily straight or twisted (Supplementary Table 3). Twisted profiles were found among asymmetrical blades and bladelets (Fig. 6a,b). Curved profiles were more frequent among blade blanks than among bladelet blanks. Complete and semi-complete debitage were mostly blanks or related to core shaping (Supplementary Table 4). Management debitage was more frequent among flakes and blades than bladelets (Supplementary Fig. 6).
72 artifacts were retouched (Table 2). Burins (dihedral or on truncations) were made on all debitage forms. Endscrapers were rare, manufactured on blades and a single flake. Retouched artifacts were primarily laterally retouched blades and bladelets, Dufour (subtype dufour) bladelets, and laterally backed blades and bladelets.
Lithic use-wear analysis
Lithics showed sporadic use primarily on larger (> 5 cm) unretouched artifacts and evidence for two hafted armature points (Supplementary Fig. 7). While some lithics recovered from GH1 showed post-depositional modification48, lithics from GH3 retained fresh, unpolished surfaces. Of the 209 lithics analyzed, 167 (80%) showed no evidence of use-wear.
43 artifacts from GH3 showed localized use-wear traces (Fig. 7). Of these, 16 had poorly developed, extended generic polishes with features unattributable to a specific use. Of those that were diagnostic, eight exhibited small alternate flake scars on dorsal and ventral surfaces and nonstriated polishes indicative of vegetal cutting. Striated polishes from hide-scraping were observed among six while five exhibited nonstriated, ventral wood cutting polish. Five lithics were related to bone scraping and one was associated with antler processing.
Two laminar points had abrupt, isolated, proximal tip damage. One bore a feather-terminating ‘spin-off fracture’ (Fig. 6d). Abrupt localized flaking without polish was found on corresponding lateral edge locations suggesting that the artifact was hafted during use49,50,51.
Only one Dufour bladelet showed use-wear as an unspecific, nonstriated polish between retouch along the proximate lateral edge. The two manuports showed a single concave groove (c. 1 cm width; 0.5 cm depth) traversing the longitudinal center of a single face (Supplementary Fig. 8). No pitting or percussion marks were observed, and micrographs showed well-developed unidirectional striations and polish possibly consistent with wood polishing. Other areas of the artifact did not exhibit signs of use.
Spatial analysis of artifacts
The vertical distributions of artifact weight and size in GH3 show no indications of post-depositional movement. As artifact weight distribution from GH3 at Românești shows a greater abundance of lighter lithics (50% for < 1 g) and shorter lithics (50% for 14–28 mm) upslope, downslope movements did not substantially alter the artifact distribution52,53.
Rather, the χ2 test and Ripley’s K function demonstrate anthropogenic aspects of the horizontal artifact distribution of GH3. Notably, blades occurred in all four clusters, whereas cores only substantially occurred in CL1 and core trimming elements only substantially occur in CL3 (Fig. 3). The dominance of blades, bladelets, and cores in CL1 and CL3 suggests that these clusters represent knapping areas of similar activities.
Lithic distributions and associations revealed by the nearest neighbor equality function suggest three different anthropogenic patterns. The first pattern relates to the correlated distribution of bladelets, blades, and cores; the second indicates an independent distribution of core trimming elements; the third relates to the change in the distribution of blades at short distances to a separate, different distribution at larger scales.
Black carbon identification
BC in sediment samples were highest in GH3 (Fig. 4a). Overall, BC did not change with depth, thus the profile likely reflects past differences in BC accumulation rather than post-depositional processes. Sediment leaching was unlikely as water solubility of BC is generally low54. Nevertheless, bioturbation could have contributed to BC content heterogeneity in individual GHs55,56,57.
BC at Româneşti was highest in GHs with the highest artifacts concentrations coinciding with human occupation intensity (GH2 and GH3). This is supported by the lower BC contents in the horizons on top and between two find layers which had few lithics (GH1 and GH2-3, respectively; Fig. 4a). However, one sample in GH2-3 showed a comparatively high BC value given the overall low contents within that horizon. As GH2-3 had low artifact density, natural processes were likely the cause as sedimentary cracks may have caused surface water and BC to percolate downwards. In contrast, the distinct BC peak in GH3 where cracks were not observed, suggests an anthropogenic influence.
The increased fire activity in GH2 (indicated by BC content) was accompanied by high fire intensity (i.e., low B5CA/B6CA ratios; Fig. 4b) typical for domestic fires. Wood-fueled anthropogenic fires typically burn substantially hotter than natural fires (e.g., forest ground/grass fires)45,58. Compared to GH2, values for GH3 indicate lower fire temperatures. In GH3, overall artifact richness, increased BC contents, and ratio values could be interpreted as a mixed signal from both natural and domestic fires. However, it could also illustrate a change in fuel source induced by changing environmental conditions or anthropogenic fire behavior.
Fire frequency in grass-dominated environments is typically higher than in dense forests59,60. During the Late Pleistocene, grassland may have been the predominant vegetation type in Banat61, with potential temporary shifts to forest steppes during MIS 362. At Româneşti, higher fire activity could explain the absence of macro-charcoal in the profile as it may have been reburned63. However, the near absence of organic residues at Româneşti and other regional open-air sites suggests that taphonomic processes may also be responsible.
Raw materials and MAN analysis
The GH3 assemblage is composed of many multiple items MANs of local raw materials introduced to the site as slightly prepared, tested, or unprepared blocks from sub-autochthonous sources and nearby gravels. The modes of introduction, quantities, and distances from the sources indicate that non-local raw materials were acquired by direct procurement indicating a complex provisioning strategy46. The obsidian artifact from GH3 was introduced as an isolated end-product that may have been acquired by indirect procurement (exchange) from sites between Românești and Southeastern Slovakia.
Still, raw materials from Românești overall suggest a narrow raw material procurement territory (< 60 km) for the GH3 assemblage primarily supplied from nearby sources (< 10 km). Altogether, local and presumed non-local shear zone related cherts comprise 84% (90% including burned artifacts) of the GH3 assemblage. The same raw materials from Românești were used at other Aurignacian sites from central and northeastern Banat, but little is known about their modes of introduction and exploitation23,30. The MAN data suggest nuanced lithic procurement patterns at Românești: non-local raw materials track the incoming path of Upper Paleolithic hunter-gatherers and outline the scale of group mobility; the raw materials supplied from local sources are potentially connected to the site position and the activities carried out there.
Techno-typology of GH3
GH3 shows on-site knapping aimed at bladelet production attested to by cortical flakes and shaped crests. Cores have single platforms where the flaking surface was installed on a narrow-fronted or broad (semi-rotating) longitudinal face. Though semi-rotating cores are fewer, their larger volume likely yielded comparable amounts of debitage. Narrow-fronted cores usually have one lateral cortical face attesting to their initiation on large decortication flakes. However, at least one pre-core (Fig. 6c; RD_N105-109) indicates that raw materials were also shaped to produce a crest on naturally narrow surfaces. Core platforms are unfaceted corresponding with debitage platforms and tablet dorsal faces. Assemblage knapping angles are mostly acute and striking platform overhangs are frequently abraded, both characteristic of laminar volumetric knapping by direct marginal percussion introduced with the Protoaurignacian64. Core faces tend to fuse into a single, broad, semicircular flaking surface throughout the reduction sequence. As narrow-fronted cores already possessed a narrow, straight flaking surface, it was sufficient to exploit the longitudinal ridges offered by the intersection of the two lateral core faces.
The GH3 assemblage shares several similarities with other contemporaneous eUP assemblages in East-Central Europe namely Kozarnika layer VII65, Siuren I layers H and G66,67,68, Berehove I AH I69,70 and Willendorf II AH 3 in Central Europe12,71. Preferentially retouched bladelets and microbladelets are main components of these inventories. All were manufactured from independent cores using broad, narrow flaking surfaces with semi-circumferential and sub-prismatic morphologies. Most bladelet profiles show flaking surfaces are organized to isolate a central, straight/slightly curved area achieved by knapping the lateral periphery of the flaking surface with off-axis, plunging laminar blanks, or by limiting the main flaking surface through lateral flakes. Striking platforms are usually plain and negatives are unidirectional. As such, these assemblages are recognized as Protoaurignacian, except Willendorf II AH 3, which is ascribed to the Early Aurignacian. The GH3 assemblage can therefore be attributed to the Protoaurignacian37.
Nevertheless, it is unclear to what degree Western European Aurignacian subcategories can be applied to those further east. Disjointed blade/bladelet production and microbladelets, aspects of the Early Aurignacian, are observed in many Southeastern Protoaurignacian assemblages34,66,68,70. This may be because carinated (Early Aurignacian) or prismatic (Protoaurignacian) cores classification can be subjective as they may exhibit similar reduction sequences67,72,73. Therefore, it is challenging to discern the Protoaurignacian from the Early Aurignacian in Central-Southeastern European record either due to the paucity of data or a genuine absence of subphases12,36,74. Nonetheless, the GH3 techno-typological characteristics and radiometric ages match eUP sites in the region currently ascribed to the Protoaurignacian.
Lithic use-wear analysis
Most analyzed artifacts at Românești showed no resharpening, abrasion, rounding, or striations and were therefore unused or only used for short durations on softer materials. The low proportion of retouched artifacts across the assemblage suggests that resharpening was not a main factor in obscuring use-wear and that lithics were generally used ad hoc. The multiple (semi-)complete lithic reduction sequences found in GH3 suggests that lithics were therefore primarily abandoned byproducts characteristic of a workshop.
Among artifacts with use-wear, evidence was primarily generic and disassociated with specific artifact types or retouched edges implying that lithics were unsystematically exploited, perhaps repeatedly for a range of tasks. Domestic tools such as endscrapers, often related to discrete activities, are virtually absent from the site suggesting a functional orientation and focus of planning depth different from other Aurignacian sites40,41,75 at a time when domestic versus hunting tools were becoming increasingly differentiated76.
When attributable to a specific use, cutting and scraping wear show that vegetal and wood processing was the main on-site activity though a range of animal processing activities were also observed. Harder materials (i.e., wood, bone) are less represented than softer materials (i.e., hide, vegetal) suggesting that overprinting of prior use traces does not obscure the overall pattern.
Spin-off fractures on elongated blanks like those found at Românești are often cited as a diagnostic impact fractures (DIFs) indicative of projectile weaponry possibly delivered by composite tools49,77,78,79,80. However, DIFs can also infrequently occur because of trampling and other natural processes77,81,82,83,84. Even combined with the potential hafting traces on a single piece, it is tempting to ascribe the Românești artifacts with DIFs to projectile armatures though such near isolated evidence cannot robustly confirm evidence for projectile damage.
Still, the virtual absence of use-wear on Dufour bladelets may provide insight into projectile manufacture processes. Upper Paleolithic bladelets, particularly Dufour bladelets38,85,86,87 have long been associated with armature inserts or points88,89,90,91. The scarcity of use traces on the Românești Dufour bladelets indicate that they were either unmanipulated in substantial processing tasks or that use-wear traces were later removed by retouch. Here, the lack of bifacial polish on the unifacially retouched lateral edges would demonstrate that if used, they were not used for substantial lengths on hard materials.
This absence of use-wear would be consistent with them either being discarded before intended use or their use as hafted inserts or points. If used as inserts, weak generic polish from the armature, hafting material or even mastic residues would be expected on a lateral edge none of which would be observed. Rather, the absence of polish and intense alternating retouch of the Dufour bladelets may be the result of inserting the bladelet into the tip of the shaft and blunting/fitting/hafting the bladelet while rotating the shaft.
The proximity, comparable sizes, coarse-grained lithology, and dimensions of the two manuports with heavily polished rectilinear grooves implicate similar functions. Both grooves appear to be furrowed, perhaps in tandem, through intense polishing of branches and therefore may have been polishers/straighteners for wooden shafts. These artifacts bear resemblance to standardized forms of shaft-straighteners known from later periods, but which have also been recovered at other eUP sites92,93,94,95,96.
A significant feature that has emerged from Aurignacian research is the potential for multiple chronologically, spatially, and technologically distinct Homo sapiens expansions across Europe68,97,98,99. Validating Trans-European dispersal scenarios, however, has remained difficult due to the bias towards Western European datasets. The new excavations from Româneşti contribute to our understanding of the earliest Homo sapiens in Europe by delivering the largest, well-contextualized Aurignacian lithic assemblage in East-Central Europe within a multi-layered site.
Taken together, Româneşti demonstrates that the onset of the East-Central European Upper Paleolithic, the Aurignacian marks a profound shift towards a large, spatially structured open-air site complex indicative of a sustained occupation or repeated visits100. This pattern, corroborated by neighboring Banat sites21,23,29,30,33, stands in contrast to ephemeral Late Middle Paleolithic assemblages101,102,103. Coupled with the contemporaneous increased anthropogenic fire input, GH3 implicates settlement dynamics at odds with a Southeastern European “pioneering phase” of rapid, unidirectional, westward expansion posited by iterations of the Danube Corridor Hypothesis74,104.
The use of good-quality local, meso-local, and exotic raw materials at the onset of the Aurignacian also demonstrates that Româneşti and the other Banat sites were repeatedly returned to from a range of distances up to 300 km and were persistent places in the landscape. Curated obsidian artifacts indicates that throughout the eUP, Româneşti was repeatedly returned to over a range of distances.
Româneşti’s Aurignacian assemblage also enables insights into the objectives and use of lithics. The narrow focus on bladelet production through a limited array of methods indicates that the site was continually geared towards a prescribed set of technological tasks. The near absence of domestic tools and the low frequency of use-wear almost exclusively on blanks, indicates that tools were generally used extemporaneously and that finished forms were exported. If the association of the site with armature production is valid, one might begin to envisage a retooling locale from which retouched bladelets were attached to armatures and transported offsite105.
The results of the large lithic assemblages and their high-quality contexts from the new excavations at Româneşti indicate profound changes in hominin occupation patterns at the onset of the early Upper Paleolithic from ephemeral to more persistent in relation to landscape dynamics during the Late Pleistocene in East-Central Europe. These data contribute to a clearer understanding of how early Homo sapiens mobility and how they consolidated their presence in Europe. Româneşti provides an important material culture context to the early European Homo sapiens fossil record delivering a key inflection point for continental-wide Aurignacian sites.
In 2016, 17 m2 were excavated adjacent to the 2009 test-trench to provide correlation with previous radiometric dates and geochemical proxies (Supplementary Fig. 1). In 2018, the excavation was extended 8 m2 to the east and in 2019, it was expanded 4 m2 north and units M–P, 100–104 were excavated further until 20 cm of sterile sediments were reached. We conformed to the stratigraphic descriptions of Sitlivy et al.37 which used a four-part division of GH based on sedimentary color and composition (GH1–4) that coincided with lithic technological changes. However, as the transition between GH2 and GH3 were not always clear due to sedimentary interfingering, an intermediary level GH2-3 was designated when attributions were ambiguous.
The excavation surface was gridded in square meters that were subdivided into quadrants (A–D). Quadrants were excavated by removing sediment in 2 cm intervals. Archeological remains ≥ 5 mm in maximum dimension were spatially recorded in three-dimensions by measuring their center point with a Leica TS06 total station using EDMwin mobile software106. Excavated sediments were wet sieved in 5 mm and selectively in 2 mm mesh to facilitate finer artifact recovery.
Finds were predominantly chipped stone artifacts and their associated debris though larger allochthonous stones were also recovered. Organic preservation was modest: Only two fragments (2 cm) of poorly preserved conjoining dental enamel (possibly Mammuthus sp.) were recovered that did not yield collagen for dating. Sediment analysis by the Cologne Archeobotanical Laboratory did not recover microfloral remains from the profile.
Spatial analysis was performed to clarify archeological stratigraphy, investigate post-depositional processes, and identify potential anthropogenic patterns. To ascribe lithics to archeological layers, lithic artifacts that vertically deviated from two neighboring artifacts by at least 1 cm were considered undetermined, as conservative threshold based on the vertical distribution of artifacts. To identify potential post-depositional processes, artifact lengths were classed according to eight categories107 and artifact weights were separated into six classes based on their overall distribution. To identify potential artifact clustering and spatial correlations, the horizontal distribution of lithic artifacts from GH3 was analyzed using a χ2 test with a Monte Carlo simulation (nsim = 39)108. The type of distribution (aggregated, segregated, or random) and the corresponding radii were determined using Ripley’s K function. A Monte Carlo spatial scan test determined cluster locations. Relative spatial distributions and artifact cluster significances were assessed with the relative risk function. Artifact spatial correlations were evaluated with the nearest neighbor equality function109,110. Analyses were performed using the R spatstat package111,112.
Black carbon identification
To evaluate past burning events, excavated sediments were analyzed for BC, the sum of organic fire residues from charcoal to soot113. In the environment, BC sedimentary input is driven by fire activity that is in turn, influenced by climate and/or human activity60,114,115.
At Româneşti, 45 samples were taken from each geological horizon except GH4 at 1 m horizontal intervals following the lithic artifact distributions across the eastern profile. Samples were dried (50 °C), sieved (< 2 mm) and milled. Organic carbon contents were determined with a Soli TOC Cube116. The quantity and quality of BC content was assessed by its oxidation to BCPA117 with modifications by Brodowski et al.118 and Kappenberg et al.. A threshold of 5 mg of organic carbon per sample was maintained and BC quantification was restricted to BPCAs with five and six carboxyl groups (B5CA and B6CA). BPCAs were measured using a Packard 6890 gas chromatograph equipped with a flame ionization detector and an HP-5 capillary column (30 m × 0.32 mm i.d., 0.25 mm film thickness; for oven program see Brodowski et al.118. BPCA yield was corrected for CO2 loss and insufficient conversion of BC to BPCAs using a factor of 2.27 to provide a conservative, minimum estimate of total BC in the sediment. The internal standard (citric acid) was recovered with a mean value of 73%.
The relative proportion of BPCA with 5–6 carboxyl groups depends on the combustion temperature. The hotter a fire burns, the more B6CA is formed, thus the ratio of B5CA to B6CA relates inversely to combustion temperature45,120.
Raw materials and MAN analysis
Lithic artifact raw materials were petrographically characterized and sourced by comparison to geological samples derived from field surveys surrounding the site46. All piece-provenienced artifacts were macroscopically analyzed and assigned to an established raw material category and type. Lithic assemblages were then submitted to a MANA that grouped artifacts with similar macroscopic features and defined the number of MANs for each raw material category determined121,122.
A retouched obsidian blade from GH3 was analyzed with a NITON XL3t portable X-ray fluorescence (pXRF) reader for trace elements at three different locations for 160 s. Uncorrected results were averaged and compared to published elemental values for continental European sources47,123,124.
The lithic analysis of the GH3 assemblage combined a chaîne opératoire approach with an attribute analysis125,126,127. The analysis used nearly all piece-provenienced artifacts from GH3 except those from unit P104 that were archived for future residue analyses.
Lithic artifacts were classified as cores, debitage, and undetermined artifacts. Core types were distinguished according to their reduction methods while debitage was subdivided into flakes (l≈w), blades (l > 2 × w with sub-parallel edges) and bladelets (blades where w < 12 mm)129. Undetermined artifacts were defined as those with no recognizable flaking axis or coherent set of negative scars.
Debitage was measured along the flaking axis and the preservation was defined as either complete, semi-complete (meso-proximal, meso-distal), or fragmented (proximal, mesial, or distal)130. Techno-typological analysis was conducted exclusively on complete and semi-complete debitage. Knapping platforms and bulb morphologies were classified according to Inizan et al.131 and Pelegrin132. Overhang abrasion was noted for cores and debitage. Debitage flaking angles (the intersection of the dorsal face with the platform) were identified as acute (< 70°), or moderately acute (70–89°). Artifacts preserving cortical surfaces were grouped as semi-cortical (< 50%), extensively cortical (51–90%), or entames (91–100%). Negative orientation was determined according to the artifact flaking direction. Blade and bladelet lateral edges were recorded as sub-parallel, convergent, or off-axis. Longitudinal profiles were sorted according to Bon133. Retouch type, position, and extent was described according to Inizan et al.131.
Debitage was additionally grouped into technological categories: cortical (entames and extensively cortical), blank (lithic artifacts flattening core convexities; i.e., blades, bladelets, and flakes), management (lithic artifacts renewing core convexities), tablets (lithic artifacts renewing the core platform), and burin spalls125. Tool typologies were classified according to Inizan et al.131.
Lithic use-wear analysis
289 piece-provenienced lithic artifacts were selected from four units (N103–N106) for use-wear and microfracture analysis. Artifact selection was based on analytical suitability (i.e., raw material, size, edge preservation, burning, patination) independent of techno-typological classification 134. Additionally, 56 (semi-) complete Dufour bladelets and two manuports (all from GH3) were analyzed. Macro and micro use-wear analysis examined the entire perimeter of the dorsal and ventral sides of the artifacts. Use-wear traces were systematized in three main categories: abrasive (e.g., striations), impact (surface macro fractures), and micro polish (i.e., sheen). Macro analysis of artifact edges and surfaces (including fractures, edge scarring, and surface abrasion) was performed with a digital microscope (ZEISS SmartZoom 5, 1.6 × objective). Micro-polished areas were analyzed using a reflected light microscope (ZEISS Axio Scope.A1 MAT, objective EC Epiplan 10 × /0.25 M27 FWD = 11.0 mm and Objective EC Epiplan 20 × /0.4 M27 FWD = 3.2 mm). All photographs were acquired with the dedicated ZEISS Zen Core software employing the image Extended Depth of Focus (EDF) stacking module to generate in-focus images. Polished surfaces were qualitatively categorized and described using common terminology and categories 135,136,137,138,139,140.
Photographs (including overviews, areas of interest and particular macro features) were edited using GIMP (free open-source image editor, available at https://www.gimp.org/, v.2.10.18). All edited images were later combined and processed using Inkscape (free and open-source vector graphics editor, available at https://inkscape.org/, v.1.1).
All data generated or analyzed during this study are included in this published article and its supplementary information files.
Conard, N. J. & Bolus, M. Radiocarbon dating the appearance of modern humans and timing of cultural innovations in Europe: New results and new challenges. J. Hum. Evol. 44, 331–371 (2003).
Tostevin, G. B. Seeing Lithics: A Middle-Range Theory for Testing for Cultural Transmission in the Pleistocene. (Oxbow Books, 2013).
Higham, T. et al. Τesting models for the beginnings of the Aurignacian and the advent of figurative art and music: The radiocarbon chronology of Geißenklösterle. J. Hum. Evol. 62, 664–676 (2012).
Floss, H., Fröhle, S. & Wettengl, S. The Aurignacian along the Danube. Its two-fold role as a transalpine and cisalpine passageway of early Homo sapiens into Europe. In Southeast Europe Before Neolithisation (eds. Krauß, R. & Floss, H.) 13–40 (Universität Tübingen, 2016).
Bertacchi, A., Starkovich, B. M. & Conard, N. J. The zooarchaeology of Sirgenstein Cave: A Middle and Upper Paleolithic site in the Swabian Jura, SW Germany. J. Paleolit. Archaeol. 4, 7 (2021).
Fu, Q. et al. The genetic history of Ice Age Europe. Nature 534, 200–205 (2016).
Posth, C. et al. Pleistocene mitochondrial genomes suggest a single major dispersal of non-africans and a late glacial population turnover in Europe. Curr. Biol. 26, 827–833 (2016).
Vallini, L. et al. Genetics and material culture support repeated expansions into Paleolithic Eurasia from a population hub out of Africa. Genome Biol. Evol. 14, evac045 (2022).
Chu, W. The Danube corridor hypothesis and the Carpathian Basin: Geological, environmental and archaeological approaches to characterizing Aurignacian dynamics. J. World Prehistory 31, 117–178 (2018).
Hajdinjak, M. et al. Initial Upper Palaeolithic humans in Europe had recent Neanderthal ancestry. Nature 592, 253–257 (2021).
Richter, D., Tostevin, G. & Skrdla, P. Bohunician technology and thermoluminescence dating of the type locality of Brno-Bohunice (Czech Republic). J. Hum. Evol. 55, 871–885 (2008).
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. Proc. Natl. Acad. Sci. 111, 14394–14399 (2014).
Hublin, J.-J. et al. Initial Upper Palaeolithic Homo sapiens from Bacho Kiro Cave, Bulgaria. Nature 581, 299–302 (2020).
Soficaru, A., Doboș, A. & Trinkaus, E. Early modern humans from the Peştera Muierii, Baia de Fier, Romania. Proc. Natl. Acad. Sci. 103, 17196–17201 (2006).
Harvati, K., Gunz, P. & Grigorescu, D. Cioclovina (Romania): Affinities of an early modern European. J. Hum. Evol. 53, 732–746 (2007).
Trinkaus, E., Constantin, S. & Zilhão, J. Life and Death at the Peştera Cu Oase: A Setting for Modern Human Emergence in Europe. (Oxford University Press, 2012).
Webb, D. et al. Ancient human footprints in Ciur-Izbuc Cave, Romania. Am. J. Phys. Anthropol. 155, 128–135 (2014).
Gély, B. et al. Peştera Coliboaia (Campani, Bihor) grotte ornée aurignacienne de Roumanie. In XIXe International Rock Art Conference IFRAO 2015 118–146 (2018).
Cârciumaru, M. Le Paléolithique en Roumanie. (Editions Jérôme Millon, 1999).
Dobrescu, R. Aurignacianul din Transilvania. (Renaissance, 2008).
The Prehistory of Banat: The Palaeolithic and Mesolithic. (Publishing House of the Romanian Academy, 2011).
Mogoșanu, F. Paleoliticul din Banat. vol. 32 (Editura Academiei Republicii Socialiste Romānia, 1978).
Chu, W. et al. Geoarchaeology and geochronology of the Upper Palaeolithic site of Temerești Dealu Vinii, Banat, Romania: Site formation processes and human activity of an open-air locality. Quartär 66, 111–134 (2019).
Nicolăescu-Plopșor, C. S. & Stratan, I. Săpăturile de la Tincova. Mater. Şi Cercet. Arheol. 7, 29–31 (1961).
Mihailović, D. Aurignacian flint industry from the site Crvenka-At near Vršac. (Univerzitet u Beogradu, Filozofski fakultet, Centar za arheološka istraživanja, 1992).
Teyssandier, N. The initial dispersion of anatomically modern humans in Europe? The early Aurignacian in Central Europe and its relationship with neighbouring areas. In New Aspects of the Central and Eastern European Upper Palaeolithic (eds. Neugebauer-Maresch, C. & Owen, L. R.), Vol. 72 65–80 (Mitteilungen der Prähistorischen Kommission, 2010).
Chu, W., Hauck, T. & Mihailović, D. Crvenka-At– Preliminary results from a lowland Aurignacian site in the middle Danube catchment. in Palaeolithic and Mesolithic Research in the Central Balkans (ed. Mihailović, D.) 69–75 (Serbian Archaeological Society, 2014).
Chu, W., Mihailović, D., Pantović, I., Hauck, T. & Lehmkuhl, F. Archaeological excavations at the site of At (Vršac, Serbia). Antiquity Project Gallery. 90(352). http://www.antiquity.ac.uk/projgall/chu352 (2016).
Chu, W., Zeeden, C. & Petrescu, S. The early Upper Paleolithic of the Banat and recent research at the Paleolithic site of Tincova. Banatica 26, 51–72 (2016).
Nett, J. J. et al. The early Upper Paleolithic site Crvenka-At, Serbia—the first Aurignacian lowland occupation site in the southern Carpathian Basin. Front. Earth Sci. 9, 56 (2021).
Borić, D. et al. Neanderthals on the Lower Danube: Middle palaeolithic evidence in the Danube Gorges of the Balkans. J. Quat. Sci. 37, 142–180 (2022).
Doboș, A., Soficaru, A. & Trinkaus, E. The Prehistory and Paleontology of the Pestera Muierii (romania). (ERAUL, 2010).
Anghelinu, M., Niţă, L., Sitlivy, V., Uthmeier, T. & Băltean, I. Looking around Peştera Cu Oase: The beginnings of Upper Paleolithic in Romania. Quat. Int. 274, 136–157 (2012).
Sitlivy, V. et al. The earliest Aurignacian in Romania: New investigations at the open air site of Româneşti-Dumbrăviţa I (Banat). Quartär 59, 85–130 (2012).
Schmidt, C. et al. First chronometric dates (TL and OSL) for the Aurignacian open-air site of Româneşti-Dumbrăviţa I, Romania. J. Archaeol. Sci. 40, 3740–3753 (2013).
Sitlivy, V. et al. Placing the Aurignacian from Banat (southwestern Romania) into the European Early Upper Paleolithic Context. in Modes de Contactes et de Deplacements au Paléolithique Eurasiatique (eds. Otte, M. & Brun-Ricalens, F. L.), Vol. 8, 243–277 (ERAUL, 2014).
Sitlivy, V. et al. Preliminary reassessment of the Aurignacian in Banat (South-western Romania). Quat. Int. 351, 193–212 (2014).
Hays, M. A. & Lucas, G. Experimental investigations of Aurignacian Dufour bladelets. In Questioning the Answers: Resolving Fundamental Problems of the Early Upper Paleolithic (eds. Hays, M. A. & Thacker, P. T.), Vol. 1005 109–117 (BAR International Series, 2001).
Tomáŝková, S. What is a burin? Typology, technology, and interregional comparison. J. Archaeol. Method Theory 12, 79–115 (2005).
Hardy, B. L., Bolus, M. & Conard, N. J. Hammer or crescent wrench? Stone-tool form and function in the Aurignacian of southwest Germany. J. Hum. Evol. 54, 648–662 (2008).
Aleo, A., Duches, R., Falcucci, A., Rots, V. & Peresani, M. Scraping hide in the early Upper Paleolithic: Insights into the life and function of the Protoaurignacian endscrapers at Fumane Cave. Archaeol. Anthropol. Sci. 13, 137 (2021).
Trinkaus, E. et al. An early modern human from the Peştera cu Oase, Romania. Proc. Natl. Acad. Sci. 100, 11231–11236 (2003).
Teyssandier, N. Questioning the first Aurignacian: Mono or multi cultural phenomenon during the formation of the Upper Paleolithic in Central Europe and the Balkans. Anthropologie 44, 9–29 (2006).
Teyssandier, N. l’émergence du Paléolithique supérieur en Europe: Mutations culturelles et rythmes d’évolution. PALEO 367–389 (2007).
Wolf, M. et al. Towards reconstruction of past fire regimes from geochemical analysis of charcoal. Org. Geochem. 55, 11–21 (2013).
Ciornei, A., Chu, W., Mariș, I. & Doboș, A. Lithic raw material patterns at the Upper Palaeolithic site of Românești-Dumbrăvița. Dacia NS 64, 67–122 (2020).
Rosania, C. et al. Revisiting Carpathian obsidian. Antiquity 82 (2008).
Levi Sala, I. Use wear and post-depositional surface modification: A word of caution. J. Archaeol. Sci. 13, 229–244 (1986).
Dockall, J. E. Wear traces and projectile impact: A review of the experimental and archaeological evidence. J. Field Archaeol. 24, 321–331 (1997).
Iovita, R., Schönekeß, H., Gaudzinski-Windheuser, S. & Jäger, F. Identifying weapon delivery systems using macrofracture analysis and fracture propagation velocity: A controlled experiment. In Multidisciplinary Approaches to the Study of Stone Age Weaponry (eds. Iovita, R. & Sano, K.) 13–27 (Springer Netherlands, 2016). https://doi.org/10.1007/978-94-017-7602-8_2.
Rots, V., Hayes, E., Cnuts, D., Lepers, C. & Fullagar, R. Making sense of residues on flaked stone artefacts: Learning from blind tests. PLoS ONE 11, e0150437 (2016).
Fanning, P. & Holdaway, S. Stone artifact scatters in western NSW, Australia: Geomorphic controls on artifact size and distribution. Geoarchaeology 16, 667–686 (2001).
Lenoble, A., Bertran, P. & Lacrampe, F. Solifluction-induced modifications of archaeological levels: Simulation based on experimental data from a modern periglacial slope and application to French Palaeolithic sites. J. Archaeol. Sci. 35, 99–110 (2008).
Abiven, S., Hengartner, P., Schneider, M. P. W., Singh, N. & Schmidt, M. W. I. Pyrogenic carbon soluble fraction is larger and more aromatic in aged charcoal than in fresh charcoal. Soil Biol. Biochem. 43, 1615–1617 (2011).
Czimczik, C. I., Preston, C. M., Schmidt, M. W. I., Werner, R. A. & Schulze, E.-D. Effects of charring on mass, organic carbon, and stable carbon isotope composition of wood. Org. Geochem. 33, 1207–1223 (2002).
Schmidt, M. W. I. & Noack, A. G. Black carbon in soils and sediments: Analysis, distribution, implications, and current challenges. Glob. Biogeochem. Cycles 14, 777–793 (2000).
Rodionov, A. et al. Black carbon in grassland ecosystems of the world. Glob. Biogeochem. Cycles 24 (2010).
Braadbaart, F. & Poole, I. Morphological, chemical and physical changes during charcoalification of wood and its relevance to archaeological contexts. J. Archaeol. Sci. 35, 2434–2445 (2008).
Pyne, S. J., Andrews, P. L. & Laven, R. D. Introduction to Wildland Fire. (Wiley, 1996).
Thonicke, K. et al. The influence of vegetation, fire spread and fire behaviour on biomass burning and trace gas emissions: Results from a process-based model. Biogeosciences 7, 1991–2011 (2010).
Marković, S. B. et al. The Crvenka loess-paleosol sequence: A record of continuous grassland domination in the southern Carpathian Basin during the Late Pleistocene. Palaeogeogr. Palaeoclimatol. Palaeoecol. https://doi.org/10.1016/j.palaeo.2018.03.019 (2018).
Kels, H. et al. Genesis of loess-like sediments and soils at the foothills of the Banat Mountains, Romania—Examples from the Paleolithic sites Româneşti and Coşava. Quat. Int. 351, 213–230 (2014).
Czimczik, C. I. & Masiello, C. A. Controls on black carbon storage in soils. Glob. Biogeochem. Cycles 21 (2007).
Pelegrin, J. Sur les débitages laminaires du Paléolithique supérieur. in François Bordes et La Préhistoire (eds. Delpech, F. & Jaubert, J.) 141–152 (Comité des Travaux Historiques et Scientifiques, 2012).
Tsanova, T. Les débuts du Paléolithique supérieur dans l’Est des Balkans: Réflexion à partir de l’étude taphonomique et techno-économique des ensembles lithiques des sites de Bacho Kiro (couche 11), Temnata (couches VI et 4) et Kozarnika (niveau VII). (Archaeopress, 2008).
Zwyns, N. Small laminar blanks at Siuren I rockshelter: Technological & comparative approach. In Siuren I rock-shelter. From Late Middle Paleolithic to Epi-Paleolithic in Crimea. (eds. Demidenko, Y. E., Otte, M. & Noiret, P.) 359–373 (ERAUL, 2012).
Bataille, G. Extracting the “Proto” from the Aurignacian. Distinct production sequences of blades and bladelets in the lower Aurignacian phase of Siuren I, units H and G (Crimea). Mitteilungen Ges. Für Urgesch. 25, 49–85 (2016).
Bataille, G., Tafelmaier, Y. & Weniger, G.-C. Living on the edge – A comparative approach for studying the beginning of the Aurignacian. Quat. Int. 474, 3–29 (2018).
Usik, V. I. et al. New data on the Early Upper Paleolithic of Western Ukraine: Chronology, environment and human behavior at the Aurignacian site of Beregovo I. In The origins of Upper Palaeolithic in Eurasia. Abstract Book. XVII. World UISPP Congress 2014. Burgos, 1–7 September 227–228 (2014).
Demidenko, Y. E., Rácz, B. & Nemergut, A. Proto-aurignacian unique site cluster in Europe logistic settlement. pattern with a base camp and a series of supply chain loci at raw material outcrops in Transcarpathia (Ukraine). Slov. Archeológia 2, 193–218 (2020).
Nigst, P. R. & Haesaerts, P. L’Aurignacien en Basse Autriche : résultats préliminaires de l’analyse technologique de la couche culturelle 3 de Willendorf II et ses implications pour la chronologie du Paléolithique supérieur ancien en Europe centrale. L’Anthropologie 116, 575–608 (2012).
Tafelmaier, Y. Technological variability at the beginning of the Aurignacian in Northern Spain: Implications for the Proto- and Early Aurignacian distinction. (Neanderthal Museum, 2017).
Falcucci, A. & Peresani, M. Protoaurignacian core reduction procedures: Blade and bladelet technologies at Fumane Cave. Lithic Technol. 43, 125–140 (2018).
Davies, W. A very model of a modern human industry: New perspectives on the origins and spread of the Aurignacian in Europe. Proc. Prehist. Soc. 67, 195–217 (2001).
Ríos-Garaizar, J. & Ortega-Cordellat, I. The Aurignacian toolkit. 15 years of traceological analysis on Aurignacian collections from Western Europe (2018).
Tartar, E., Teyssandier, N., Bon, F. & Despina, L. Equipement de chasse, équipement domestique: une distinction efficace? Réflexion sur la notion d’investissement technique dans les industries aurignaciennes. In Normes Techniques et Pratiques Sociales. De la Simplicité des Outillages Pré- Et Protohistoriques (eds. Astruc, L., Bon, F., Léa, V., Milcent, P.-Y. & Phillibert, P.-Y. M. et S.) 107–118 (ADPCA, 2006).
Sano, K. Hunting evidence from stone artefacts from the Magdalenian cave site Bois Laiterie, Belgium: A fracture analysis. Quartär 56, 67–86 (2009).
Shea, J. J. & Sisk, M. L. Complex projectile technology and Homo sapiens dispersal into western Eurasia. PaleoAnthropology 100, 122 (2010).
Eren, M. I. et al. Test, model, and method validation: The role of experimental stone artifact replication in hypothesis-driven archaeology. Ethnoarchaeology 8, 103–136 (2016).
O’Driscoll, C. A. & Thompson, J. C. The origins and early elaboration of projectile technology. Evol. Anthropol. Issues News Rev. 27, 30–45 (2018).
O’Farrell, M. Etude préliminaire des éléments d’armature lithique de l’Aurignacien ancien de Brassempouy. In Productions lamellaires attribuées à l’Aurignacien: Chaînes opératoires et perspectives technoculturelles, Vol. 1 (ArchéoLogiques, 2005).
Pargeter, J. Assessing the macrofracture method for identifying Stone Age hunting weaponry. J. Archaeol. Sci. 38, 2882–2888 (2011).
Pargeter, J. Rock type variability and impact fracture formation: Working towards a more robust macrofracture method. J. Archaeol. Sci. 40, 4056–4065 (2013).
Weitzel, C., Borrazzo, K., Ceraso, A. & Balirán, C. Trampling fragmentation potential of lithic artifacts: An experimental approach. Intersecc. Antropol. 15, 97–110 (2014).
Lucas, G. The origin of Dufour bladelet torsion. In Questioning the Answers: Resolving Fundamental Problems of the Early Upper Paleolithic (eds. Hays, M. A. & Thacker, P. T.) 99–107 (BAR International Series, 2001).
Ollé, A. & Vergès, J. M. The use of sequential experiments and SEM in documenting stone tool microwear. J. Archaeol. Sci. 48, 60–72 (2014).
Giner, P. J., Pion, G. & Hortelano, L. Experimental basis in lithic arrows usage and hafting at the end of the last glaciation in the French Alps. Quat. Int. 427, 193–205 (2017).
Bon, F. Little big tool enquête autour du succès de la lamelle. In Productions lamellaires attribuées à l’Aurignacien: Chaînes opératoires et perspectives technoculturelles (ArchéoLogiques 1, 2005).
Le Brun-Ricalens, F., Bordes, J.-G. & Eizenberg, L. A crossed-glance between southern European and middle near eastern early Upper Palaeolithic technocomplexes: Existing models, new perspectives. In The Mediterranean From 50 000 to 25 000 BP: Turning Points and New Directions (eds. Camps, M. & Szmidt, C.) 11–33 (Oxbow Books, 2009).
Teyssandier, N., Bon, F. & Bordes, J.-G. Within projectile range: Some thoughts on the appearance of the Aurignacian in Europe. J. Anthropol. Res. 66, 209–229 (2010).
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 and Yafteh (Iran). J. Hum. Evol. 65, 39–64 (2013).
Solecki, R. L. & Solecki, R. S. Grooved stones from Zawi Chemi Shanidar, a Protoneolithic site in Northern Iraq. Am. Anthropol. 72, 831–841 (1970).
Belfer-Cohen, A. & Bar-Yosef, O. The Aurignacian at Hayonim Cave. Paléorient 7, 19–42 (1981).
Wright, K. A classification system for ground stone tools from the prehistoric Levant. Paléorient 18, 53–81 (1992).
De Beaune, S. Nonflint stone tools of the Early Upper Paleolithic. In Before Lascaux: The Complex Record of the Early Upper Paleolithic (eds. Knecht, H., Pike Tay, A. & White, R.) 163–191 (CRC Press, 1993).
Caricola, I. et al. An integrated method for understanding the function of macro-lithic tools. Use wear, 3D and spatial analyses of an Early Upper Palaeolithic assemblage from North Eastern Italy. PLOS ONE 13, e0207773 (2018).
Mellars, P. The earliest modern humans in Europe. Nature 479, 483–485 (2011).
Banks, W. E., d’Errico, F. & Zilhão, J. Human–climate interaction during the Early Upper Paleolithic: Testing the hypothesis of an adaptive shift between the Proto-Aurignacian and the Early Aurignacian. J. Hum. Evol. 64, 39–55 (2013).
Higham, T., Wood, R., Moreau, L., Conard, N. & Ramsey, C. B. Comments on ‘Human–climate interaction during the early Upper Paleolithic: Testing the hypothesis of an adaptive shift between the Proto-Aurignacian and the Early Aurignacian’ by Banks et al. J. Hum. Evol. 65, 806–809 (2013).
Bailey, G. & Galanidou, N. Caves, palimpsests and dwelling spaces: Examples from the Upper Palaeolithic of south-east Europe. World Archaeol. 41, 215–241 (2009).
Doboș, A. & Chu, W. Between the woods and the water: The early Upper Paleolithic from the Romanian karst. Analele Banat. 26, 17–34 (2019).
Alex, B., Mihailović, D., Milošević, S. & Boaretto, E. Radiocarbon chronology of Middle and Upper Paleolithic sites in Serbia, central Balkans. J. Archaeol. Sci. Rep. 25, 266–279 (2019).
Anghelinu, M. & Boroneanţ, A. The archaeology of caves in Romania. in Cave and karst systems of Romania (eds. Ponta, G. M. L. & Onac, B. P.) 501–517 (Springer, 2019). https://doi.org/10.1007/978-3-319-90747-5_57.
Davies, W. Re-evaluating the Aurignacian as an expression of modern human mobility and dispersal. In Rethinking the Human Revolution: New Behavioural and Biological Perspectives on the Origin and Dispersal of Modern Humans. (eds. Mellars, P., Boyle, K. V., Bar-Yosef, O. & Stringer, C.) 263–274 (McDonald Institute for Archaeological Research, 2007).
Caspar, J.-P. & De Bie, M. Preparing for the hunt in the Late Paleolithic camp at Rekem, Belgium. J. Field Archaeol. 23, 437–460 (1996).
Dibble, H. L. & McPherron, S. P. EDMwin (2010).
Bertran, P., Lenoble, A., Todisco, D., Desrosiers, P. M. & Sørensen, M. Particle size distribution of lithic assemblages and taphonomy of Palaeolithic sites. J. Archaeol. Sci. 39, 3148–3166 (2012).
Robert, C. & Casella, G. Monte Carlo Statistical Methods. (Springer, 2004). https://doi.org/10.1007/978-1-4757-4145-2.
Cobo Sánchez, L. Taphonomic and spatial study of the archeological site DS from Bed I in Olduvai Gorge (Tanzania). (Universidad Complutense de Madrid, 2020).
Diez-Martín, F. et al. Tracing the spatial imprint of Oldowan technological behaviors: A view from DS (Bed I, Olduvai Gorge, Tanzania). PLoS ONE 16, e0254603 (2021).
Baddeley, A., Rubak, E. & Turner, R. Spatial Point Patterns: Methodology and Applications with R. (CRC Press, 2015).
R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing. (2020).
Goldberg, E. D. Black Carbon in the Environment: Properties and Distribution (Wiley, New York, 1985).
Marlon, J. R. et al. Global biomass burning: A synthesis and review of Holocene paleofire records and their controls. Quat. Sci. Rev. 65, 5–25 (2013).
Lehndorff, E., Wolf, M., Litt, T., Brauer, A. & Amelung, W. 15,000 years of black carbon deposition—A post-glacial fire record from maar lake sediments (Germany). Quat. Sci. Rev. 110, 15–22 (2015).
Mörchen, R. et al. Carbon accrual in the Atacama Desert. Glob. Planet. Change 181, 102993 (2019).
Glaser, B., Haumaier, L., Guggenberger, G. & Zech, W. Black carbon in soils: The use of benzenecarboxylic acids as specific markers. Org. Geochem. 29, 811–819 (1998).
Brodowski, S., Rodionov, A., Haumaier, L., Glaser, B. & Amelung, W. Revised black carbon assessment using benzene polycarboxylic acids. Org. Geochem. 36, 1299–1310 (2005).
Kappenberg, A., Bläsing, M., Lehndorff, E. & Amelung, W. Black carbon assessment using benzene polycarboxylic acids: Limitations for organic-rich matrices. Org. Geochem. 94, 47–51 (2016).
Schneider, M. P. W., Hilf, M., Vogt, U. F. & Schmidt, M. W. I. The benzene polycarboxylic acid (BPCA) pattern of wood pyrolyzed between 200°C and 1000°C. Org. Geochem. 41, 1082–1088 (2010).
Larson, M. L. & Kornfeld, M. Chipped stone nodules: Theory, method, and examples. Lithic Technol. 22, 4–18 (1997).
Miller, G. L. MANA from heaven: Testing the utility of minimum analytical nodule analysis at large, repeatedly reoccupied ceremonial sites. J. Archaeol. Sci. Rep. 8, 1–10 (2016).
Oddone, M., Márton, P., Bigazzi, G. & Biró, K. T. Chemical characterisations of Carpathian obsidian sources by instrumental and epithermal neutron activation analysis. J. Radioanal. Nucl. Chem. 240, 147–153 (1999).
Kohút, M., Westgate, J. A., Pearce, N. J. G. & Bačo, P. The Carpathian obsidians—Contribution to their FT dating and provenance (Zemplín, Slovakia). J. Archaeol. Sci. Rep. 37, 102861 (2021).
Soressi, M. & Geneste, J. M. The history and efficacy of the chaîne opératoire approach to lithic analysis: Studying techniques to reveal past societies in an evolutionary perspective. PaleoAnthropology 334, 334–350 (2011).
Tostevin, G. B. Levels of theory and social practice in the reduction sequence and chaîne opératoire methods of lithic analysis. PaleoAnthropology 351, 351–375 (2011).
Scerri, E. M. L., Gravina, B., Blinkhorn, J. & Delagnes, A. Can lithic attribute analyses identify discrete reduction trajectories? A quantitative study using refitted lithic sets. J. Archaeol. Method Theory 23, 669–691 (2016).
Conard, N. J., Soressi, M., Parkington, J. E., Wurz, S. & Yates, R. A unified lithic taxonomy based on patterns of core reduction. S. Afr. Archaeol. Bull. 59, 12 (2004).
Tixier, J. Typologie de l’épipaléolithique du Maghreb. (Arts et Métiers Graphiques, 1963).
Andrefsky, W. Lithics Macroscopic Approaches to Analysis. (Cambridge University Press, 2005).
Inizan, M.-L., Reduron-Ballinger, M., Roche, H. & Tixier, J. Technology and terminology of knapped stone. (CREP, 1999).
Pelegrin, J. Les techniques de débitage laminaire au Tardiglaciaire: critères de diagnose et quelques réflexions. In L’Europe centrale et septentrionale au Tardiglaciaire: confrontation des modèles régionaux de peuplement. Actes de la table-ronde internationale de Nemours, 14–16 mai 1997 (eds. Valentin, B., Bodu, P. & Christensen, M.), Vol. 7 73–86 (Association pour la promotion de la recherche archéologique en Ile-de-France, 2000).
Bon, F. L’aurignacien Entre Mer et Océan: Réflexion Sur L’unité Des Phases Anciennes De L’aurignacien dans le Sud De La France, Vol. 29 (Société préhistorique française, 2002).
Marreiros, J. et al. Rethinking use-wear analysis and experimentation as applied to the study of past hominin tool use. J. Paleolit. Archaeol. 3, 475–502 (2020).
Tringham, R., Cooper, G., Odell, G., Voytek, B. & Whitman, A. Experimentation in the formation of edge damage: A new approach to lithic analysis. J. Field Archaeol. 1, 171–196 (1974).
Keeley, L. H. Experimental Determination of Stone Tool Uses: A Microwear Analysis. (University of Chicago Press, 1980).
Odell, G. H. Toward a more behavioral approach to archaeological lithic concentrations. Am. Antiq. 45, 404–431 (1980).
Geneste, J.-M. & Plisson, H. Analyse technologique des pointes à cran solutréennes du Placard (Charente), du Fourneau du Diable, du Pech de la Boissiere et de Combe-Saunière (Dordogne). PALEO 1, 65–106 (1989).
Lemorini, C., Estevez, J. J. I., Hayden, B., Sawada, A. & Gallart, F. Traces et Fonction: Les Gestes Retrouvés, Vol. 2 (ERAUL, 1993).
Rots, V. Prehension and Hafting Traces on Flint Tools: A Methodology. (Leuven University Press, 2010). https://doi.org/10.2307/j.ctt9qf05s.
Thank you to Alexandru Szentmiklosi†, the Museum of Banat, the commune of Tomești, Sorin-Marius Petrescu, and our field excavation team members. Thank you to Astrid Röpke for checking sediments for pollen and phytolith remains and to Peter Abbot and Hendrik Vogel for analyzing them for tephra. Our gratitude to Dirk Leder and Gyuri Lengyel for their help recording artifacts and to Ellery Frahm, Stephan Opitz and Dominik Hezel for their advice and assistance measuring the obsidian artifact. Thorsten Uthmeier, Valéry Sitlivy, Mircea Anghelinu and Jürgen Richter contributed to thoughtful discussions on the site. Lucia Cobo-Sánchez gave important advice on the spatial analyses. Anja Rüschmann produced lithic drawings. We are grateful to Marie Soressi and Alexander Verpoorte for reading and commenting on an earlier draft.
Funding was provided by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation)—Project Number 57444011 - SFB 806 and a Dutch Research Council grant (VI.C.191.O70).
The authors declare no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Chu, W., McLin, S., Wöstehoff, L. et al. Aurignacian dynamics in Southeastern Europe based on spatial analysis, sediment geochemistry, raw materials, lithic analysis, and use-wear from Românești-Dumbrăvița. Sci Rep 12, 14152 (2022). https://doi.org/10.1038/s41598-022-15544-5