A revised radiocarbon chronology for the mammoth bone structures and associated features at Mezhyrich, Ukraine

RESEARCH ARTICLE A revised radiocarbon chronology for the mammoth bone structures and associated features at Mezhyrich, Ukraine [version 1; peer review: 1 approved, 2 approved with reservations] Wei Chu 1, Pavlo Shydlovskyi 2, Andreas Maier 3 1World Archaeology, Universiteit Leiden, Leiden, South Holland, 2333CC, The Netherlands 2Department of Archaeology and Museum Studies, Taras Shevchenko National University of Kyiv, Kyiv, 01033, Ukraine 3Institute of Prehistoric Archaeology, University of Cologne, Cologne, North Rhine-Westphalia, 50969, Germany First published: 25 Jul 2025, 5:198 https://doi.org/10.12688/openreseurope.20112.1 Latest published: 25 Jul 2025, 5:198 https://doi.org/10.12688/openreseurope.20112.1 v1 Abstract Mammoth Bone Structures are a distinct archaeological phenomenon typically ascribed to the Late Pleistocene in East-Central Europe though their chronology and use remain debated. By reviewing the history of research and excavation at the Mezhyrich site, this article presents new 14C results from Mezhyrich (Cherkasy Oblast, Ukraine), revising the chronology of an Upper Palaeolithic Mammoth Bone Structure. Using meso-mammal remains from cultural layers, the results provide a more constrained timeline than previous models based on mammoth bones. The findings date Mammoth Bone Structure 4 and its context to c. 18,248–17,764 years cal BP with a site duration lasting between 0–429 years. This shorter chronology is consistent with a single occupation model but cannot exclude repeated occupation within maximally a few centuries at around 18 ka cal BP contributing to the hypothesis that the structure served as a dwelling, offering insights into Late Upper Palaeolithic life patterns in Eastern Europe. Plain Language Abstract Here we present new radiocarbon dates for a mammoth bone structure at Mezhyrich, Ukraine, refining its timeline. By dating small mammal remains instead of mammoth bones, we establish a more precise age of around 18,248–17,764 years ago. The structure appears to have been used for up to 429 years, supporting the idea that it served as a dwelling. These findings help us better understand how people lived in Eastern Europe during the Late Upper Paleolithic. Keywords Late Upper Palaeolithic, Mezhyrich, Mammoth Bone Structures, radiocarbon dating, isotopic analysis, human-environment interactions, settlement chronology Open Peer Review Approval Status 123 version 1 25 Jul 2025 view view view Timothy J Heaton , University of Leeds, Leeds, UK 1. William Chase Murphree , Universidade do Algarve, Faro, Portugal 2. Andrzej Wiśniewski , University of Wrocław, Wrocław, Poland 3. Any reports and responses or comments on the article can be found at the end of the article. Open Research Europe  Page 1 of 23 Open Research Europe 2025, 5:198 Last updated: 21 NOV 2025 Corresponding author: Wei Chu ([email protected]) Author roles: Chu W: Conceptualization, Funding Acquisition, Visualization, Writing – Original Draft Preparation, Writing – Review & Editing; Shydlovskyi P: Conceptualization, Data Curation, Methodology, Visualization, Writing – Original Draft Preparation, Writing – Review & Editing; Maier A: Conceptualization, Writing – Original Draft Preparation, Writing – Review & Editing Competing interests: No competing interests were disclosed. Grant information: This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (ERC, HOME-101124398). This project has also received funding from the Stichting Nederlands Museum voor Anthropologie en Praehistorie. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Copyright: © 2025 Chu W et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. How to cite this article: Chu W, Shydlovskyi P and Maier A. A revised radiocarbon chronology for the mammoth bone structures and associated features at Mezhyrich, Ukraine [version 1; peer review: 1 approved, 2 approved with reservations] Open Research Europe 2025, 5:198 https://doi.org/10.12688/openreseurope.20112.1 First published: 25 Jul 2025, 5:198 https://doi.org/10.12688/openreseurope.20112.1 This article is included in the Horizon 2020 gateway. This article is included in the European Research Council (ERC) gateway. Open Research Europe  Page 2 of 23 Open Research Europe 2025, 5:198 Last updated: 21 NOV 2025 Introduction Open-air Upper Palaeolithic sites in East-Central Europe offer valuable insights into human activity and biogeography during the Late Pleniglacial, a period of intense environmental change (Verpoorte, 2009). Many of these sites, characterised by deep loessic sedimentary sequences, contain abundant lithic, faunal, and osseous assemblages with embedded environmental proxies to help determine contemporaneous environmental factors (Chu & Nett, 2021). Notably in this region, some of these sites are found in direct association with mammoth bone accumulations posited to be among the earliest evidence of built structures that exhibit spatial and seasonal organisation (Iakovleva & Djindjian, 2005; Iakovleva et al., 2012; Shipman, 2015; Shydlovskyi et al., 2022). Many of these constructions have been previously interpreted as the remains of domestic structures (Aurenche et al., 2013; Gladkih et al., 1984; Iakovleva, 2021; Iakovleva & Djindjian, 2018; Klein, 1974; Pidoplichko & Allsworth-Jones, 1998; Soffer, 1985; Soffer et al., 1997). However, since most were excavated decades ago, the question has reemerged as to whether they functioned primarily as dwellings or if they represent sites of other functions, such as bone beds, food caches, burials, religious traditions, or ritualised middens that may have served as early monuments (Djindjian, 2015; Gavrilov, 2024; Graeber & Wengrow, 2021, p. 73; Pryor et al., 2020; Sablin et al., 2023; Zheltova, 2024). A key site to this debate is Mezhyrich (Межиріч), situated in the Middle Dnieper Basin of Ukraine known for its exceptional preservation of four Mammoth Bone Structures (MBSs) between 12–24 m2 in diameter (Figure 1). These structures are associated with peripheral features and artefacts including artefact-filled pits, hunting weapons, ivory, and bone ornaments, as well as delimited activity areas with butchered animal remains and workshops with dense cultural layers, each delimited into “economical settlement units” with the MBS as the focal point (e.g. Units 1–4; Shydlovskyi et al., 2019). Still, two key challenges hinder thorough interpretations of Mezhyrich: 1. There is still a coarse understanding of the temporal scale of use of MBSs (Gaudzinski-Windheuser, 2011; Iakovleva et al., 2012; Shydlovskyi et al., 2023). While the Mezhyrich MBSs were repeatedly used, radiometric ages for the specific occupation durations are scarce. Previous ages for Mezhyrich’s MBS 4 have largely been based on disparate radiocarbon methods using mammoth remains that have likely been primarily obtained through scavenging. As a result, ages may not coincide with the cultural deposition and may contribute to overestimates of the age of the anthropogenic activities at the site (van der Plicht & Palstra, 2016). It is still unclear how often Mezhyrich MBS 4 was occupied and for how long these occupations lasted. 2. There is still a lack of reliable ages to establish correlations between the cultural layers found within MBS 4 and associated peripheral features, such as artefact dense areas (e.g., Toptalishche areas) and nearby pits (e.g., Pit 6). Sediment accumulation within MBS 4 may be independent of external loess deposition and more closely related to the frequency, duration, and mode of human use. To address these research gaps, this study presents new AMS radiocarbon ages and isotopic data obtained from mediumsized mammal remains directly recovered from archaeological Figure 1. Map of Mezhyrich’s location within East Central Europe. Page 3 of 23 Open Research Europe 2025, 5:198 Last updated: 21 NOV 2025 layers within Mezhyrich. This study focuses on the dating and analysis of MBS 4 (Unit 4), the only one of these structures that contains well-provenienced artefacts within cultural layers from which to obtain direct ages, as well as Pit 6 (Unit 2) and the external saturated cultural layer (Toptalishche; Unit 1; Figure 2). The primary objective is to evaluate the occupation history of MBS 4 and its implications for the broader settlement history of the site during the Late Pleistocene. Based on this refined chronology, we contribute to the interpretation of MBS 4 as a multi-phasic dwelling structure and propose that such structures played significant roles in the East Central European life patterns during the Late Upper Palaeolithic. This adds to a better understanding of the function and chronology the MBS, providing insights into the past human behaviours and adaptations in this region. Background In 1930, M.Ia. Rudynskyi’s excavation at the Mizyn (Mezin) settlement uncovered a cluster he interpreted as a “small residential building” with “curtain walls” (Rudynskyi, 1954, pp. 35–37). Around the same time, Yefimenko emphasised studying Upper Palaeolithic residential structures to understand the social relationships and worldview of prehistoric communities hypothesising dwellings at Eastern European sites like Kostenki I and II, Hintsi (Gontsy), and Suponievo (Yefimenko, 1953). The study of Palaeolithic dwellings subsequently became a focus of Eastern European archaeology, aided by large-area excavation techniques that enabled recording various settlement features simultaneously (Shovkoplias, 1965, p. 260). Significant contributions came from the Desna expedition at sites near Novhorod-Siverskyi, Chulatovo, and Pushkari where at Pushkari I, Boriskovskii proposed a residential structure reconstruction (Boriskovskii, 1940). Further excavations at the Pushkari Cape at sites like Pushkari I, Pogon (aka Pushkari IX, layer 3 and Pushkari VIII), and Bugorok (aka Pushkari IX), revealed substantial evidence of residential structures (Voevodskii, 1950). At the Chulatovo II settlement (1936–1938), Voevodskyi uncovered remains of a small structure, delimited by vertically placed mammoth tusks and long bones (Voevodskii, 1952, pp. 105–106). Initial excavations at Mezhyrich (1966–1974), yielded Palaeolithic artefacts and three discrete MBSs, each c. 6 m in diameter (Pidoplichko, 1976). A lack of detailed provenience recording led to a significant loss of archaeological material and information concerning cultural layers and spatial arrangements (Gladkikh & Kornietz, 1979b). The subsequent discovery of a fourth structure (Dwelling 4; MBS 4) in 1976 has since exposed a complex of surrounding pits and activity areas that exhibit diverse artefacts embedded within distinct cultural layers (Figure 3; Gladkih et al., 1984; Soffer et al., 1997). MBS 4 provides evidence of past domestic activities through the accumulation of at least three internal cultural layers and the presence of butchered remains of juvenile and adolescent mammoths within the surrounding pits (Tsvirkun et al., 2021; Figure 4). The mammoth bones used in the construction of MBS 4 were sourced from a minimum of 37 individuals, predominantly from natural accumulations, though some were obtained in anatomical position as evidenced by at least one freshly deceased mammoth carcass (Gladkih et al., 1984). Subsequent studies have provided environmental context for the Mezhyrich site (Table 1) notably demonstrating a local dynamic environment that may have been warmer and more humid than other contemporary periglacial areas of Europe. Long-term trends in stable isotope analyses revealed that the initially strongly diversified herbivore niches of the mammoth steppe collapsed during the Last Glacial Maximum (Reiss et al., 2023) in Central Europe. While early post-LGM isotope analysis on faunal remains from Barmaky (19 ka; north-western Ukraine) still show a strongly diversified niche spectrum for herbivores (Reiss et al., In Review), measurements on mammoth remains from younger specimens in Eastern Europe give mixed signals, but may point to a disappearance of their optimal habitat and increasing competition with other large herbivores, particularly at the southern fringes of their distribution in the East European Plain (Drucker et al., 2018). The associated eponymous Epigravettian technological tradition, known as the Mezhyrich industry, is believed to have abruptly ceased around c. 13 ka ago, coinciding with the local disappearance of mammoths (Gladkikh, 1977; Nuzhnyi, 2008; Shydlovskyi et al., 2020). Stratigraphy Initial geomorphological data suggested a Late Pleistocene chronology for the site dating the cultural layers to the Late Glacial period no earlier than 17–16 ka BP during cold and arid climatic conditions, though less severe than those during the Last Glacial Maximum (LGM; Kornietz et al., 1981; Velichko et al., 1993a; Velichko et al., 1993b; Velichko et al., 1994. During excavations south of MBS 4 in 1978, two distinct cultural layers separated by a sterile layer (10–20 cm thick) were identified; the lower layer was linked to the construction phase of the MBS, while the upper layer was associated with the settlement’s final occupation phase. These two discrete Figure 2. Photograph of Dwelling 4 during excavation. Page 4 of 23 Open Research Europe 2025, 5:198 Last updated: 21 NOV 2025 Figure 3. Plan of Mezhyrich site: 1 – limits of mammoth bone structures; 2 – pits; 3 – hearths; 4 – limits of dense anthropogenic layers (“Toptalishche”); 5 – plots from which samples were obtained. Page 5 of 23 Open Research Europe 2025, 5:198 Last updated: 21 NOV 2025 cultural layers were interpreted as episodic and intermittent occupations of the site by prehistoric groups, indicative of transient hunting camps (Gladkikh & Kornietz, 1979a, pp. 11–13). Further geoarchaeological research has since corroborated the multi-layered stratigraphy of the site with microstratigraphic analyses of the lithostratigraphy, around the excavation perimeter, Figure 4. Site stratigraphy of MBS4 and with three archaeological layers (after Tsvirkun et al., 2021). Table 1. Previous contextual studies on the Mezhyrich site. Study Findings Reference Stable isotopes • Low 15N due to collapse of ecological niche c. 18–17 ka and focus on mature grasses. • changes in regional flora and climatic conditions may have altered the food resources available to mammoths, potentially leading to direct competition with horse populations. (Drucker et al., 2014; Drucker et al., 2018) Microstratigraphy • The activities of Epigravettian hunter-gatherers between 18.3–17.4 ka took place during a cold period characterised by strong climatic and environmental contrasts. (Haesaerts et al., 2015) Anthracology • The major charcoal signal is preserved within the microscopic part of the archaeological sediments underlying the intensity of taphonomic processes. • Birch and willow were located along the riverbanks. (Marquer et al., 2012) Micromammals • Environment was cold but not as dry as in a typical periglacial area, warmer and wet, a cold mesophilic forest-steppe. (Rekovets et al., 2014) Mammals • The results exhibit a usage of the pit as a dump area of food and technical remains from mainly mammoth, hare, and fox processing, and bone fuels. • The information about mammoth procurement by the last Palaeolithic hunter–gatherers in Eastern Europe allows to document the hunting activities on the mammoth populations, which were probably already weakened at the end of the Pleistocene. (Péan, 2015) Palynology • Plant cover during Upper Palaeolithic had a mosaic structure. (Komar, 2015) Malacology • The molluscan fauna from the Toptalishche is characteristic of drier and milder local conditions. • The molluscan fauna from below the Toptalishche indicates stadial loessic deposits. (Prisiazhniuk, 2015) Page 6 of 23 Open Research Europe 2025, 5:198 Last updated: 21 NOV 2025 targeting the anthropogenic layers of the 1st, 2nd, and 4th MBSs. This research revealed three cultural layers intercalated by sterile layers, thereby elucidating the spatial and temporal relationships among different Units at the site (Haesaerts et al., 2015; Shydlovskyi et al., 2019). Chronology 34 previously reported radiocarbon ages from Mezhyrich have suggested an occupation of the site between c. 24–14 ka calibrated Before Present (cal BP; Table 2). However, these ages vary in 14C-measurement methodologies, sample materials, Table 2. Previous radiocarbon ages from Mezhyrich. Unit, context Method Dated material taxa Lab-ID 14C Age (yrBP) ± 1sigma Calibrated age (95.4% probability) Source DW 4, cultural layer Conventional Burned bone (Mammuthus) KI-1054 17850 950 24110–19484 (Gladkikh & Kornietz, 1979b, p. 16) Cultural layer (1978) Conventional Burned tooth (Mammuthus) KI-1055 18020 600 23278–20408 (Gladkikh & Kornietz, 1979b, p. 16) Cultural layer (1976) Conventional Burned bone (Mammuthus) KI-1056 18470 550 23741–21050 (Gladkikh & Kornietz, 1979b, p. 16) DW 1, cultural layer (1966) Conventional Burned bone (Mammuthus) KI-1057 19100 500 24263–22050 (Gladkikh & Kornietz, 1979b, p. 16) DW 1, cultural layer (1966) Conventional Bone (Mammuthus) KI-1058 19280 600 24893–22090 (Gladkikh & Kornietz, 1979b, p. 16) DW 4, cultural layer Conventional Burned bone (Mammuthus) QC-900 В 15245 1080 21743–15980 (Gladkikh & Kornietz, 1982, p. 18) DW 1, cultural layer Conventional Burned bone (Mammuthus) QC-897 14320 270 18211–16771 (Soffer, 1985, p. 143) DW 3, cultural layer Conventional Tooth (Mammuthus) GIN-2593 14700 500 19056–16611 (Soffer, 1985, p. 143) DW 2, cultural layer Conventional Burned bone (Mammuthus) GIN-2595 14530 300 18619–16982 (Soffer, 1985, p. 143) DW 4, cultural layer Conventional Burned bone (Mammuthus) GIN-2596 14300 300 18221–16641 (Soffer, 1985, p. 143) DW 1 AMS Tooth (collagen) (Mammuthus) OxA-709 12900 200 16070–14591 (Soffer, 1985, p. 26) DW 2 AMS Tooth (collagen) (Mammuthus) OxA-712 14400 250 18225–16988 (Soffer, 1985, p. 26) DW 3 AMS Tooth (Mammuthus) AA-1317 14420 190 18138–17097 (Soffer, 1985, p. 26) DW 1 AMS Femur (Canis lupus) GrA-22501 14450 90 17923–17352 (Haesaerts, 2006, p. 38) DW 2 AMS Femur (Canis lupus) OxA-13044 14380 60 17815–17337 (Haesaerts, 2006, p. 38) DW 2: Idem OxA-13044 AMS Femur (Canis lupus) GrA-22094 14600 110 18181–17461 (Haesaerts, 2006, p. 38) DW 4: OS 07-3 (1984) (Toptalishche) AMS Femur (Canis lupus) SacA-14981 15210 130 18763–18252 (Haesaerts et al., 2015, p. 385) DW 1: OS 07-6 (1966) AMS Tibia (Vulpes lagopus) SacA-14982 14400 90 17886–17312 (Haesaerts et al., 2015, p. 385) DW 4: (To-2) KB-323, bord nord HA-4 (2005–2008) Conventional Femur (Mammuthus) GrN-29876 14550 70 18059–17435 (Haesaerts, 2006, p. 38) Page 7 of 23 Open Research Europe 2025, 5:198 Last updated: 21 NOV 2025 and selection methods resulting in a wide distribution (Gladkikh & Kornietz, 1979b; Gladkikh & Kornietz, 1982; Haesaerts, 2006; Haesaerts et al., 2015; Shydlovskyi et al., 2023; Soffer, 1985; Soffer, 1993). The initial age estimates (those with lab codes KI) were obtained using conventional radiocarbon methods and range from 24.1–14.6 ka cal BP. However, these broad ages are inconsistent with stratigraphic evidence suggesting the initial cultural layer predates the last Late Pleistocene loess accumulation cycle. Additionally, an outlier age of 16.1–14.6 ka cal BP (OXA-709), derived from a mammoth tooth excavated nearly 30 years before analysis, indicates a younger age. These ages contrast with more reliable ages from accelerator mass spectrometer (AMS) measurements that place the initial cultural layer within a narrower range of 15.4–14.3 ka uncal BP, corresponding to a calibrated age range of approximately 2000 years. Analysing the calibrated AMS ages, Haesaerts identified three settlement phases corresponding to three cultural layers observed in the stratigraphic trenches of the 1st, 2nd, and 4th Units (Haesaerts et al., 2015). Several layers within settlement features, such as pits and MBS 4, were associated with three distinct cultural phases identified across the entire settlement area: 1. The first phase (To1, Z1) spanned from 15,050–14,750 BP (18,300–17,900 cal BP; 400 years). 2. The second phase (To2, Z2) ranged from 14,900–14,500 BP (18,150–17,600 cal BP). Unit, context Method Dated material taxa Lab-ID 14C Age (yrBP) ± 1sigma Calibrated age (95.4% probability) Source DW 4: (To-2) KB-344, bord sud HA-4 (2005–2008) Conventional Femur (Mammuthus) GrN-29877 14560 70 18083–17450 (Haesaerts, 2006, p. 38) DW 4: Idem GrN-29877 AMS Femur (Mammuthus) OxA-15587 14790 60 18245–17928 (Haesaerts, 2006, p. 38) DW 2, Pit 6: Z-2a upper (To-2) (2005–2008) MZH-5 AMS Charcoal SacA-11487 14600 60 18149–17540 (Haesaerts et al., 2015, p. 385) DW 2, Pit 6: Z-2a upper (To-2) (2005–2008) MZH-2 AMS Charcoal SacA-11486 14610 60 18162–17571 (Haesaerts et al., 2015, p. 385) DW 2, Pit 6: Z-2a middle (To-2) (2005–2008) OS 07-9 AMS Bone (Mammuthus) GrA-38810 14750 50 18222–17907 (Haesaerts et al., 2015, p. 385) DW 2, Pit 6: Idem GrA-38810 AMS Bone (Mammuthus) SacA-11177 14810 90 18266–17884 (Haesaerts et al., 2015, p. 385) DW 2, Pit 6: Z-2a middle (To-2) (2005–2008) OS 08-01 AMS Metacarpal (Canis lupus) SacA-12040 15320 90 18820–18290 (Haesaerts et al., 2015, p. 385) DW 1, Pit 7: Z-2/3, OS 07-7 AMS Femur (Mammuthus) GrA-38787 14590 60 18130–17515 (Haesaerts et al., 2015, p. 386) DW 1, Pit 7: Idem GrA-38787 AMS Femur (Mammuthus) SacA-11176 15030 90 18647–18191 (Haesaerts et al., 2015, p. 386) DW 1, Pit 7: Idem SacA-11176 AMS Femur (Mammuthus) SacA-14986 15430 90 18898–18337 (Haesaerts et al., 2015, p. 386) DW 1, Pit 8: Z-1 (To-1) OS 08-02 AMS Rib (Mammuthus) SacA-12041 14830 90 18277–17891 (Haesaerts et al., 2015, p. 390) DW 1, Pit 8: Idem SacA-12041 AMS Rib (Mammuthus) SacA-14984 14920 90 18620–17985 (Haesaerts et al., 2015, p. 390) DW 1, Pit 8: Idem SacA-12041 AMS Rib (Mammuthus) SacA-12259 14970 90 18642–18106 (Haesaerts et al., 2015, p. 390) DW 1, upper cultural layer: TO3A (2015) AMS Bone (Mammuthus) GrA-66076 14660 70 18212–17746 (Shydlovskyi et al., 2017) DW 1, upper cultural layer: TO3B (2015) AMS Bone (Mammuthus) GrA-66078 14730 70 18221–17858 (Shydlovskyi et al., 2017) Page 8 of 23 Open Research Europe 2025, 5:198 Last updated: 21 NOV 2025 3. The third phase covers the period from 14,500–14,300 BP (17,750–17,400 cal BP). These settlement phases indicate hiatuses of c. 200 years between each phase, with only the second and third phases associated with the presence of MBSs (Haesaerts et al., 2015). The identification of three cultural layers within MBSs 1, 2, and 4 suggests that each phase corresponded to a distinct Unit and MBS, implying at least three distinct episodes of habitation spanning c. 900 years (Chabai et al., 2020). Despite some rejected ages, these ages suggest synchronicity between Units 2 and 4 (Haesaerts et al., 2015). However, such an interpretation relies primarily on mammoth bone ages that show significant temporal variation. Reliable mammoth bone ages range by millennia and may thus indicate a sustained use of scavenged mammoth remains following the LGM. Under such a scenario, the youngest mammoth bone ages mark the actual settlement period, possibly indicative of active hunting. This is corroborated by other non-mammoth samples, like charcoal and meso-mammal bones which indicate a significantly shorter period, forming two distinct groups: 18.2–17.3 ka cal BP and 18.8–18.3 ka cal BP. Minimal age differences are observed between MBSs 1, 2, and 4, as well as between wolf and mammoth bones, with ages aligning with earlier mammoth material (18.2–17.3 ka cal BP), likely representing the main occupation period. Microstratigraphy comparisons between MBS 4 and Pit 6 show distinct differences but reveal concurrent use, as the ash layer age in Pit 6 matches the establishment of MBS 4. Multiple layers within MBS 4 suggest several habitation episodes. In short, while identifying three settlement phases advanced understanding of the site, uncertainties remain about the duration of occupation. The correlation of lower pit layers with early settlement phases remains tenuous, and careful selection of samples is warranted as the stratigraphy of artificial objects may not align with the overall settlement sequence. The correlation of cultural layers in pits and MBS 4 with broader settlement layers raises concerns, as these features may not represent the overall stratigraphy. With that in mind, this study seeks an independent analysis of the stratigraphy of MBS 4 and Pit 6 and their immediate surroundings and radiocarbon data for more accurate comparisons. Materials and methods In the absence of substantial charcoal samples, twelve meso-faunal osseous samples from three species (Vulpes sp., Lepus sp., Lupus sp.) were selected from recent excavations with well-provenienced finds within cultural layers. Larger, well-preserved, compact bones (e.g. pelvis, long bones) were prioritised to minimise potential vertical displacement within the stratigraphy and maximise collagen extraction. One sample exhibited cutmarks, underscoring its anthropogenic accumulation. The twelve samples were chosen from different archaeological contexts within the site. Four samples originated from archaeological layers within MBS 4, highlighting different aspects of its occupation. Another four samples were derived from Pit 6, to clarify its cultural accumulation and activity. The remaining four samples were selected from saturated cultural layer surrounding MBS 1 (i.e., Toptalishche), providing a broader understanding of the wider site use. AMS dating was employed on twelve samples after pre-treatment with an acid-base-acid (ABA) protocol following the standard protocols of the Centre for Isotope Research at Groningen University (Dee et al., 2020). High-precision measurements of δ13C and δ15N were also conducted prior to AMS measurement to assess sample suitability. Chronological modelling was undertaken using OxCal v4.4 using the IntCal20 calibration dataset to refine estimates for the site’s chronology (Ramsey, 2017; Reimer et al., 2020). A Kernel Density Estimation (KDE) in combination with Bayesian start/end date modelling was used to summarise the distributions of each occurrence based on the available chronological data incorporating 5 previously obtained AMS ages from Pit 6 for a total of 16 ages (Haesaerts et al., 2015). Radiocarbon age determinations inherently contain a degree of uncertainty expressed as a distribution of radiocarbon ages. KDE accounts for this uncertainty by sampling individual radiocarbon age ranges to generate a set of probable calibrated age ranges for each event within a given dataset. The algorithm then applies a KDE to these sampled ages, producing a smoothed estimate of the temporal event density over 10,000 iterations and averaging the results. This approach enables the segregation of phases, represented here by the different localities from which the samples were obtained (i.e., MBS 4, Toptalishche, Pit 6) and subphases, denoted by the cultural layers within these localities (i.e., Z2 and Z3 within Pit 6 and To3/2 and T3 within the Toptalishche). Results 11 new AMS radiocarbon age determinations were obtained (Table 3), along with the associated analytical data for each of the dated bone samples. Of the 12 samples analysed, 11 produced usable collagen, with elemental values (%C, %N) falling within the accepted ranges for well-preserved collagen (Talamo et al., 2021). The one failed sample, from Unit 1, To-3, V20#108, contained insufficient collagen to meet the minimum reliable threshold and was therefore excluded from further analysis. Among the remaining 11, the atomic C/N ratios for all samples ranged from 3.1–3.3, within the acceptable limits of 2.9–3.5 (Higham et al., 2011). %C yields were high, ranging from 43.6%–39.8%. The results of the KDE Bayesian Start-End model (SI1) show that the ages from small mammals from the selected site features can be constrained from 18,504–18,238 years cal BP to 17,865–17,431 years cal BP with a site duration lasting between 396–1027 years. This model had an agreement of A=51.3%, below the standard accepted agreement index of A’c= 60.0%. This poor Page 9 of 23 Open Research Europe 2025, 5:198 Last updated: 21 NOV 2025 Open Peer Review Current Peer Review Status: Version 1 Reviewer Report21 November 2025 https://doi.org/10.21956/openreseurope.21756.r62139 © 2025 Wiśniewski A. This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Andrzej Wiśniewski University of Wrocław, Wrocław, Poland The article addresses several very important issues: 1. assessment of the age of MBS4 and assessment of the chronological relationship between cultural layers from MBS and peripheral structures.The authors answer the questions posed in accordance with current methods, providing very important results that clarify the chronology of the analysed objects and the time of their use. I have no substantive comments. I rate the article as very excellent. I have only a few grammatical comments regarding the spelling of names in references, and in one case the full title of the publication needs to be provided. See the list below. 1.Dee MW, Palstra SWL, Aerts-Bijma AT, et al.: Radiocarbon dating at groningen 2.Gaudzinski-Windheuser S: An introduction to living structures and history of occupation at the Late upper paleolithic site of Oelknitz 3.Pidoplichko IG, Allsworth-Jones P: Upper palaeolithic 4.Rekovets L, Nowakowski D, Lech K: Analyse des micromammifères du site épigravettien de Mezhyrich (Ukraine). L’Anthropologie. 2014 -Complete the title of the publication. I would suggest correcting Figure 3. The hearts should be more prominent. It looks as if the figure was drawn by hand, especially the pits and Toptalishche structures are strangely marked. Is the work clearly and accurately presented and does it cite the current literature? Yes Is the study design appropriate and does the work have academic merit? Yes Are sufficient details of methods and analysis provided to allow replication by others? Yes If applicable, is the statistical analysis and its interpretation appropriate? Open Research Europe  Page 16 of 23 Open Research Europe 2025, 5:198 Last updated: 21 NOV 2025 Yes Are all the source data underlying the results available to ensure full reproducibility? Yes Are the conclusions drawn adequately supported by the results? Yes Competing Interests: No competing interests were disclosed. Reviewer Expertise: Palaeolithic archaeology, Geoarchaeology, Technology, Settlement strategies I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard. Reviewer Report21 November 2025 https://doi.org/10.21956/openreseurope.21756.r62138 © 2025 Murphree W. This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. William Chase Murphree Universidade do Algarve, Faro, Faro District, Portugal This paper is focused on providing a revised chronology for the late Pleistocene Mammoth bone structure (MBS) 4 fromMezhyrich (Cherkasy Oblast, Ukraine). The goals of this paper were to better understand the occupational history of MBS 4. Using new14C results, the authors have provided a new short chronological framework for the occupations of MBS 4 and it's surrounding features to a short window of around 429 years. Generally, I thought the paper is well written, very interesting, and has a solid structure. However, I do have some specific comments and questions regarding the stratigraphic context of these samples. I also have some minor comments as well. I hope my comments and suggestions are constructive for the authors and fixable. Main comments: Overall, I found the descriptions of the context of the samples and stratigraphic descriptions are lacking and hard to follow throughout the text, especially in regards to MBS 4. Considering how important context is to the premise of this paper and its conclusions, this issue is problematic but fixable. Starting in the background chapter, I would suggest a full revision of Figure 4. It is unclear what the section photograph on the right is referring to, the current version is missing key points of overlap with the section drawing. I would suggest adding labels on the photograph indicating the layers to improve the readability of the figure as well as adding markers to show the square ID, Open Research Europe  Page 17 of 23 Open Research Europe 2025, 5:198 Last updated: 21 NOV 2025 especially if the photo is from the same location as the drawing. If the two images are not from the same location then they should be labelled as such. Otherwise both described in detail in the caption. Something that I feel that is relatively ambiguous throughout the discussion and tables is the stratigraphic context for the samples from MBS 4 (Unit/ DW 4). This information is not included in Table 3 and is glossed over in the discussion. The authors only provide a relative height from surface in Table 3. The other samples from Pit 6 and the toptalishche have layer designations. This comes off as confusing to me as the authors repeatedly refer to three distinct layers within MBS 4 (Unit/DW 4) and the importance of understanding the occupation history throughout the text but then do not state from where the samples are coming from. I think it would help improve the discussion or conclusion if they would provide more detail on layers or relative position the samples are taken from in Table 3 and in Figure 4 or if this data is not available state this more explicitly in the text. It's understandable that samples maybe have been procured from an earlier excavation or isn't available but I believe that should be explicit in the text.Additionally, in the discussion section for MBS 4, I don't quite understand how occupations can be both contemporaneous and yet be in stratigraphically distinct horizons. I think the phrasing of this section should be clarified. Reproducibility: Currently, there is no code available with this paper. The authors also mention SI2 (pg. 11) in the results, but it is currently unavailable online. Minor comments: 1. Within the text and tables the authors occasionally alternate between using MBS and DW without giving a clear definition for each phrase. It a bit difficult to follow through the text. I would suggest these terms be defined simply in the background section to improve the readability throughout the rest of the text. 2. Section on how units are defined should be in the background as well. I understand the authors choice to include this section in the discussion but prior to this section it was unclear if unit referred to the MBS/DW or excavation area in which they were found. Possibly they could include a sentence or two in the background section to clarify. 3. In the Chronology section, the three cultural layers/ phases is a bit confusing with how it's worded. All three phases have some overlap in terms of dating so the hiatuses between phases isn't very clear in the current text. 4. Following up on the three phases, It also is not clear what the authors meant by three cultural layers within the MBSs. Does this suggest that within the last two phases there are multiple sublayers within the individual MBSs? I think this section should be revised for greater clarity. Is the work clearly and accurately presented and does it cite the current literature? Yes Is the study design appropriate and does the work have academic merit? Yes Are sufficient details of methods and analysis provided to allow replication by others? Partly Open Research Europe  Page 18 of 23 Open Research Europe 2025, 5:198 Last updated: 21 NOV 2025 If applicable, is the statistical analysis and its interpretation appropriate? I cannot comment. A qualified statistician is required. Are all the source data underlying the results available to ensure full reproducibility? Partly Are the conclusions drawn adequately supported by the results? No Competing Interests: No competing interests were disclosed. Reviewer Expertise: Geoarchaeology; Microstratigraphic analysis I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above. Reviewer Report28 October 2025 https://doi.org/10.21956/openreseurope.21756.r61253 © 2025 Heaton T. This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Timothy J Heaton University of Leeds, Leeds, England, UK A revised radiocarbon chronology for the mammoth bone structures and associated features at Mezhyrich, Ukraine 19th October 2025 Summary The paper provides new 14C dates and a re-analysis of an interesting archaeological site in Mezhyrich (Cherkasy Oblast, Ukraine) where mammoth bone structures from the Late Pleistocene have been discovered. Specifically, the authors have: Provided a set of 11 new AMS 14C dates on samples from the site (Units 1, 2 and 4)1. Combined these with existing AMS dates from the same units obtained by others2. Fitted an updated chronological model using OxCal and a KDE modelling approach 3. I felt the paper is very nicely written. It provided a very engaging and readable story. I enjoyed reviewing it, and it certainly taught me new things.  I do however have some specific questions about the modelling approach (and the need for a KDE/non-parametric approach). I hope/think that these queries can be addressed by the authors. I have written a fair bit below (but this is mainly for explanation). I hope my actual suggestions for edits to address these queries (given in the appropriate section) are relatively simple/quick. Open Research Europe  Page 19 of 23 Open Research Europe 2025, 5:198 Last updated: 21 NOV 2025 I hope my comments/suggestions are useful and constructive.  Main Review Comments on Modelling My specific queries related to the modelling are provided below: Do we need a KDE approach? The authors say that they use a KDE modelling approach in their analysis. However, it is unclear what this corresponds to and critically what it means. The model presented in Fig 5 appears to have some internal phases/groupings (MBS 4, Toptalische, Pit 6) some of which have further phases/subgroupings. Within each grouping there are then the individual 14C samples. What dates exactly are being modelled non-parametrically (all the dates; the dates within each group; the start/end of each group/sub-group)? And why is this needed? To me, incorporating a KDE model into a begin/end phase approach makes it very unclear both what the prior information actually incorporated into the model is, and more importantly what effect it will have on the resultant inference. In my personal opinion, the KDE model here seems rather unnecessary – especially when you have a lot of genuine prior information and not that many dates. Is there a reason why a much more traditional uniform phase type model isn’t more useful and appropriate – not least in terms of clarity of interpretation and explanation? Outliers in a KDE modelling approach Related to the above, I do not understand what an OxCal agreement score would mean (or how it should be interpreted) in a context where you have specifically aimed to place no assumption on the form of the calendar age distribution between the start/end dates (by using a KDE). How can you tell if something is an outlier in such a context? This score really just appears to flag those two samples which determine the start/end of the (nonparametric) phase. By removing these you are substantially shrinking the range of the model (i.e., and hence your interpretation). However, under a KDE model I am not sure you really have solid grounds to use agreement to identify such outliers. I do not believe you can get reliable/quantitative information on which samples are outliers without making some assumption about the form of the calendar distribution (and which don’t fit well). Personally, I think you would be better fitting a standard uniform phase type model as that seems to effectively be what you have used as an argument to eliminate the two 14C samples with the oldest/youngest dates anyway. What is actually being modelled? At some points in the paper you refer to your motivation being to perform an independent Open Research Europe  Page 20 of 23 Open Research Europe 2025, 5:198 Last updated: 21 NOV 2025 analysis of MBS 4. However, this doesn’t appear to be the actual model fitted as you have used samples from other units too. This is fine, but I am not sure that you can say that you are doing an independent analysis of MBS4 when most of the dates/samples that you are using for inference come from different Units entirely (4/16 from Unit 1; and 9/16 from Unit 2 (Pit 6)? In particular when the most recent date of your phase analysis (pre-outlier removal) seems primarily determined by your 1st sample in Table 3 which comes from Unit 1: Unit 1, To-3/2, V19, #532 which has a 14C age of 14295 14C yr BP (almost 500 14C yrs more recent than all other samples) while the last is determined by SacA-12040 which comes from Unit 2, Pit 6 which has a 14C age of 15320 14C yr BP. Failure to combine measurements on the same sample in model It seems (from Fig 5 and Table 2) that the model does not combine measurements on the same sample (if this is what idem denotes) but treats them as though they are independent samples with potentially different calendar ages. For example, the model in Fig seems to consider GrA38810 and SacA-11177 as though they are separate samples – whereas the Table suggests they are repeat measurements of the same sample and so must have the same calendar age. These repeated samples need to be combined (into a single sample) before being introduced into the model – not allowed to have two separate calendar ages. Suggestion that there are distinct clusters in Pit 6 and Toptalishche On pg 12 you suggest there are distinct clusters in these two units. Is this based solely on Fig 5 (or from other information). Personally, I think it is hard to say confidently (based on Fig 5) that there are two distinct clusters. For example, in Pit 6 Z2 there seem only 6 samples that have been analysed and they seem to all overlap in age. I would therefore remove this comment. Equally based on Fig 5 there isn’t much evidence that there are two distinct dating clusters in To3 based purely on 14C – there are only three samples analysed here. Or is this statement based on other evidence not presented here? Reproducibility/Availability of Code: Currently there is no code or dataset attached to the paper (or at least in the version that I have access to). There is a suggestion of an SI but it’s not on the journal portal.  Specific Suggestions for Revisions/Edits: To address the above comments, I would ideally like to see: Model re-run ensuring idem measurements are modelled as having the same calendar age (if these relate to same actual object/sample). This is important. 1. Personally I would entirely ditch the KDE modelling component (as I am not sure what it actually means when included as part of a larger model) and just fit a traditional phase model that is much more easily explainable (both in terms of what the model is, and how it influences the analysis). At the least I would like to know that the results are robust to the KDE assumption. A more traditional parametric model may also allow you identify outliers in a more robust/rigorous manner. 2. Code and data (as e.g. csv file) should made available (ideally with a DOI via Zenodo). Data 3. Open Research Europe  Page 21 of 23 Open Research Europe 2025, 5:198 Last updated: 21 NOV 2025 should not just be in a table in the manuscript. Even if there is an SI, I would suggest that storing it in a more accessible online repository is important for reproducibility. Some relatively minor rewording about (and further explanation) about why you have chosen to include some units but not others, and rephrasing to ensure you do not give the idea you have done an independent analysis of MBS4 (when much of the analysis seems influence by the other units). Currently it’s not entirely clear why you have chosen BMS 4 and Pit 6 (but not included, e.g., AMS dates from other pits)  4. Minor Points/Comments: Introduction - Please can you give specific/approximate calendar dates for the late pleniglacial period? This would aid readers who might not know this nomenclature/specific period to compare with other work more easily. This is especially critical as it only seems to be used as a term once – you later change to late Pleistocene and/or Paleolithic. 1. I found it slightly confusing when MBS 4 was first introduced (pg 3) as to what this was – I wasn’t sure if it was all four structures or just the 4th one. I had to infer from the abstract it was just one of them. I would suggest that to clear this up, when you first introduce MBS in the paragraph before that you might name all four, i.e., “A key site to this debate is Mezhyrich ( Межиріч), situated in the Middle Dnieper Basin of Ukraine known for its exceptional preservation of four Mammoth Bone Structures (denoted in our manuscript as MBS 1, 2, 3 and 4 respectively) between 12–24 m2 in diameter (Figure 1).”You might then need to be consistent with referring to MBSs (multiple structures) or a specific MBS. For example when referring to their first discovery (pg4) you could specify the initial excavations revealed MBS 13; … 2. Pg 4 – Move up LGM abbreviation to 1st usage of term last glacial maximum3. Pg 8/9 – The three phases as described in the manuscript have overlapping calendar ages so I am confused why you say there are hiatuses between them – please can you explain?  4. I do not understand the sub-analysis of phase timings on pg11 – are these obtained by rerunning the analysis again just on these samples? Or are they somehow obtained from the single initial model (i.e. you have start/end phases for all these other bits) 5. Figure 5 – Suggest make clear from caption that this is the model after the outliers have been removed. 6. Conclusion – you have changed from using the term provenience to provenance. Is there a reason for this? Suggest you should be consistent. 7. Is the work clearly and accurately presented and does it cite the current literature? Yes Is the study design appropriate and does the work have academic merit? Yes Are sufficient details of methods and analysis provided to allow replication by others? Partly If applicable, is the statistical analysis and its interpretation appropriate? Partly Are all the source data underlying the results available to ensure full reproducibility? Partly Open Research Europe  Page 22 of 23 Open Research Europe 2025, 5:198 Last updated: 21 NOV 2025 Are the conclusions drawn adequately supported by the results? Partly Competing Interests: No competing interests were disclosed. Reviewer Expertise: Radiocarbon; Statistics I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above. Open Research Europe  Page 23 of 23 Open Research Europe 2025, 5:198 Last updated: 21 NOV 2025