Bone Surface Modifications Research Articles

Artificial intelligence for the identification of taphonomic bio-accumulator agents: an actualistic test of potential faunal accumulation agency applied to Tritons Cave (Lleida, Spain)

Studying bone surface modifications (BSMs) in neotaphonomic research is an important aid to reconstruct agency in the archaeological and palaeontological record. The significance of correctly identifying BSMs has led to extensive debates about adequate methodological and interpretive frameworks to identify taphonomic agents and their bone-modifying patterns in dynamic taphonomic processes; especially those involving inter-agency interactions. Recent analytical innovations in the field have given rise to more updated and less biasing methods, including those rooted in multivariate statistics and artificial intelligence algorithms. Among the latter, convolutional neural networks have demonstrated a substantial potential in precision in the identification and characterization of BSMs. In this study, we present a successful application of these methods to reconstructing taphonomic agency in an archaeological context, specifically in Unit 2 of Tritons Cave (Lleida, Spain). Previous interpretations of this palaeontological layer suggested that it was a natural palimpsest created by carnivores, namely leopards. Through the present research, we objectively tested the original felid-accumulation hypothesis associated with the site’s original taphonomic analysis. Our findings provide additional evidence of the primary leopard agency in the formation of the assemblage and underscores the significant potential of transfer learning for identifying taxon-specific carnivore agency.

Experimental replication of early human behaviour in bird preparation: a pilot-study focusing on bone surface modification and breakage patterns

The origin of complex behaviour amongst early humans is a subject of heated debate within the scientific community, and the study of small prey remains has become a significant aspect when examining such modern behaviour. Nonetheless, the consumption of small prey by human populations poses analytical difficulties due to the often negligible, or entirely absent, traces on bone surfaces. To address this difficulty, an experimental study focusing on terrestrial avifauna has been prepared, and here we present a preliminary phase of this research. The aim is to distinguish potential modifications on bird bone surfaces and fracture patterns that might facilitate the recognition of human manipulation of avian skeletal remains. Building upon the challenges encountered in the study of archaeological findings recovered from recent excavations in Iberian Middle Palaeolithic sites, the experimental protocol was formulated to encompass the processing of two uncooked and three roasted birds; and the lithic use-wear analysis of the flint flake used in the processing of raw birds. The results showcase distinct patterns of bone surface modifications and breakage between cooked and uncooked birds. Higher numbers of cut marks and manual disarticulation breaks are found on raw animals, whereas roasted animals show no cut marks, local-specific burns and higher bone loss. This pilot-study provides a baseline for future research to further explore the role of avifauna in Neanderthal subsistence and food processing, which may help highlight cultural choices.

Anthropic cut marks in extinct megafauna bones from the Pampean region (Argentina) at the last glacial maximum.

The initial peopling of South America is a topic of intense archaeological debate. Among the most contentious issues remain the nature of the human-megafauna interaction and the possible role of humans, along with climatic change, in the extinction of several megamammal genera at the end of the Pleistocene. In this study, we present the analysis of fossil remains with cutmarks belonging to a specimen of Neosclerocalyptus (Xenarthra, Glyptodontidae), found on the banks of the Reconquista River, northeast of the Pampean region (Argentina), whose AMS 14C dating corresponds to the Last Glacial Maximum (21,090-20,811 cal YBP). Paleoenvironmental reconstructions, stratigraphic descriptions, absolute chronological dating of bone materials, and deposits suggest a relatively rapid burial event of the bone assemblage in a semi-dry climate during a wet season. Quantitative and qualitative analyses of the cut marks, reconstruction of butchering sequences, and assessments of the possible agents involved in the observed bone surface modifications indicate anthropic activities. Our results provide new elements for discussing the earliest peopling of southern South America and specifically for the interaction between humans and local megafauna in the Pampean region during the Last Glacial Maximum.

Identification and quantification of projectile impact marks on bone: new experimental insights using osseous points

Shifts in projectile technology potentially document human evolutionary milestones, such as adaptations for different environments and settlement dynamics. A relatively direct proxy for projectile technology is projectile impact marks (PIM) on archaeological bones. Increasing awareness and publication of experimental data sets have recently led to more identifications of PIM in various contexts, but diagnosing PIM from other types of bone-surface modifications, quantifying them, and inferring point size and material from the bone lesions need more substantiation. Here, we focus on PIM created by osseous projectiles, asking whether these could be effectively identified and separated from lithic-tipped weapons. We further discuss the basic question raised by recent PIM research in zooarchaeology: why PIM evidence is so rare in archaeofaunal assemblages (compared to other human-induced marks), even when they are explicitly sought. We present the experimental results of shooting two ungulate carcasses with bone and antler points, replicating those used in the early Upper Paleolithic of western Eurasia. Half of our hits resulted in PIM, confirming that this modification may have been originally abundant. However, we found that the probability of a skeletal element to be modified with PIM negatively correlates with its preservation potential, and that much of the produced bone damage would not be identifiable in a typical Paleolithic faunal assemblage. This quantification problem still leaves room for an insightful qualitative study of PIM. We complement previous research in presenting several diagnostic marks that retain preservation potential and may be used to suggest osseous, rather than lithic, projectile technology.

Taphonomic study of the cave bears (Ursus cf. deningeri and U. spelaeus) from the Sima I of the El Polvorín cave (Northern Iberian Peninsula)

Cave bears used caves for hibernation and giving birth. Caves maintain stable conditions of temperature and humidity, which facilitates the preservation of fossil accumulations. Causes of mortality in caves are related to starvation during hibernation, which affects mainly juveniles and old adults, predation and accidents, affecting bears of all ages. The Iberian Peninsula is located in one of the extremes of the geographical range of this species and has yielded an abundant cave bear fossil record, particularly in karstic systems in the northern fringe. The origin of the accumulation of bear fossil sites in Iberia has been investigated mostly based on selected paleobiological variables (sex-ratio and age-at-death profiles) and qualitative taphonomic assessment. The objective of this paper is to delve into the origin of a bear accumulation with a large proportion of hyaena remains, the Sima I of El Polvorín cave (Biscay, Northern Iberian fringe, Spain). For that end a combination of quantitative paleobiological (age-at-death and sex) and taphonomic (bone fracture patterns and surface modifications) data is used. This analysis has ruled out humans and hyenas as the origin of the accumulation but, due to the limited number of fossil remains available it is not possible to ascertain whether the bear accumulation was the result of a natural trap within the cave, natural mortality during hibernation, or a combination of both causes. This study underscores the necessity of integrated methodological approaches and the potential of old paleontological collections to provide new important paleobiological and taphonomic information under these approaches.

Cumulative effect of high‐resolution silicone moulds on the morphology of cut marks

AbstractThe use of high‐resolution silicone moulds for documenting bone surface modifications, such as cut marks, is common. However, it has not been evaluated whether moulding can affect the originals. In this work, the modification level derived from several moulding–demoulding processes on an experimental sample of cut marks has been characterised using geometric Morphometrics. It has been shown that moulds influence the morphology of cut marks, reducing their variability, and making the sample more homogeneous. These modifications do not affect the identification of cut marks, but if not considered, may have an effect on more specialised studies.

An ethnoarchaeological comparison of observed butchery actions to bone surface modifications of the Dobe !Kung San

ABSTRACT The Dobe !Kung were a group of foragers living in northern Botswana in the Kalahari Desert. In the second half of the twentieth century, their butchery actions on animal prey were recorded (1968–1975) and their abandoned camps from 1944–1976 were excavated to retrieve faunal remains from butchered prey animals. Here we test the hypothesis that bone surface modifications on these excavated faunal remains accurately reflect these butchery observations. We find that despite a few exceptions, the observed bone surface modifications follow the expectations derived from observations of butchery of different sized animals. This supports the idea that past butchery actions, based on the location and presence of bone surface modifications on modern bones, can be confidently linked to past butchery behaviours. We also suggest that bone surface modifications can provide information about past hunter-gatherer processing behaviours, cooking styles, and use of hides.

Evidence of diverse animal exploitation during the Middle Paleolithic at Ghar-e Boof (southern Zagros)

Although Middle Paleolithic (MP) hominin diets consisted mainly of ungulates, increasing evidence demonstrates that hominins at least occasionally consumed tortoises, birds, leporids, fish, and carnivores. Until now, the MP zooarchaeological record in the Zagros Mountains has been almost exclusively restricted to ungulates. The narrow range of hominin prey may reflect socioeconomic decisions and/or environmental constraints, but could also result from a research bias favoring the study of large prey, since archaeologists have undertaken no systematic taphonomic analyses of small game or carnivores in the region. Here, we report on the first comprehensive taphonomic analysis of an MP faunal assemblage from Ghar-e Boof (∼ 81–45 kyr), a Late Pleistocene site in the southern Zagros of Iran. Anthropogenic bone surface modifications point to hominins as the main agent of accumulation. Hominins preyed primarily on ungulates, particularly wild goat. However, we also found evidence for MP hominin exploitation of carnivores and tortoises at the site. Although small game represents only a minor portion of the diet, our results suggest that the hunting behavior of MP hominins in the Zagros was more diverse than previously thought, similar to what we find elsewhere in Eurasia.

Identifying non-anthropogenic accumulation in zooarchaeological assemblages using naive Bayesian classifier: A trace-oriented actualistic taphonomic approach in the hyperarid coasts of the Atacama desert

The taphonomic evaluation of zooarchaeological assemblages is paramount for compelling interpretations of past subsistence strategies, palaeoecological interactions and overall biasing of the anthropogenic bone deposition and modification by non-anthropogenic agents. A recent trend in this field, has focused on bone surface modifications (BSM) to successfully generate reference frameworks for critical evaluation of the interpretive potential of selected zooarchaeological samples by examining BSM patterns produced by known agents in experimental settings. These approaches compare anthropogenic and non-human generated patterns using different machine learning (ML) algorithms, classifying and assigning sets of traces to specific processes such as gnawing, scoring or intentional fracture. We propose a simplification in two key aspects to make this new arising area of ML assisted taphonomic interpretation more accessible for zooarchaeologists. First, we encourage a trace oriented multivariate data recording protocol for taphonomic analysis including the most commonly recorded BSM plus others such as shape modification, penetrative modification and element or tissue loss modifications thus expanding comparative potential. Second, we offer a ML approach based on Naïve Bayes Classifier (NBC) that discriminates each given observation by comparing to actualistic taphonomic patterns to determine the average percentage of non-anthropogenic modification within zooarchaeological samples. To illustrate the potential value of these proposals for taphonomical interpretation, we analyzed seven different Mid to Late Holocene assemblages containing avian and mammal remains from the coastal platform in the hyperarid core of the Atacama Desert. We compared their alteration with a local sample of actualistic data recovered to characterize regional taphonomic patterns. A NBC was trained with a 75% random selection of the total archaeological and actualistic samples. Afterwards we tested the remaining sample, achieving over 90% accuracy in determining whether a random specimen from any sample corresponded to an actualistic or an archaeological context implying key differences in taphonomic trace patterns.

Prey body size generates bias for human and avian agents: Cautions for interpreting small game assemblages

Identifying the accumulating agent(s) of small animal assemblages at archaeological sites is challenging given the range of behaviors, skeletal plans, and body sizes of prey species as well as the range of potential predators. Although mammalian carnivores, raptors, and humans are considered typical predators of small game, there are many contexts—especially in island ecosystems—where humans and raptors are the primary agents for small animal assemblages in the absence of mammalian carnivores. Moreover, smaller animals tend to range in body sizes in these geographically restricted ecosystems. This study aims to broadly explore: 1) how small animal body size affects the taphonomic signature(s) of human and avian predation; and 2) how human and avian-generated small game assemblages compare with one another. To investigate these questions, this study compared the resulting taphonomic patterns of small animal predation by raptors and humans using data collected during a controlled-feeding experiment and an observational study, respectively. In the controlled-feeding experiment, deceased laboratory rats (Rattus norvegicus domestica) (50–500 g) were fed to two Milky Eagle Owls (Ketupa lacteus), one King Vulture (Sarcoramphus papa), and two Lappet-Faced Vultures (Torgos tracheliotos). In the observational study, we explored how humans processed various species of small prey ranging in body mass (80–5000 g) by analyzing anthropogenic modifications (i.e., cutmarks, tooth marks, burning). The resulting small animal assemblages were analyzed post-consumption for skeletal element representation, fragmentation and breakage patterns, and bone surface modifications. Results show that small prey body size significantly affects pre-depositional patterning in terms of skeletal element survival, butchering intensities, and cutmark frequencies. Furthermore, our experimentally generated assemblages show that breakage, fragmentation, and element relative abundance patterns can be used to distinguish between human and avian-accumulated assemblages in the archaeological record.

Experiments suggest rockfall an improbable cause for bone surface modification on 24,000-year-old bone at Bluefish Caves, Canada

At Bluefish Caves in the Yukon Territory of Canada, researchers argue that 24,000-year-old faunal remains bear anthropogenic cutmarks, and provide evidence of a pre-Last Glacial Maximum (23-18 ka) human occupation of North America's arctic. However, non-anthropogenic agents can also leave marks on bones that might resemble bone surface modifications (BSMs) produced by stone tools. Here we test the hypothesis that rockfall traces can mimic stone tool cutmarks by experimentally replicating the spalling limestone shelter at Bluefish Caves, dropping variably sized limestone blocks (12–2322 g) from 3.5 m onto Bovidae remains. These same bones were subsequently sliced with stone tools to produce cutmarks for comparison. The resulting marks were analyzed with a Dino-Lite microscope and a Sensofar S Neox 3-D optical profilometer to allow quantitative comparisons of their morphologies. Results show that cutmarks can be distinguished from rockfall damage, and do not support the hypothesis that the BSM previously identified as cutmarks at Bluefish Caves are instead a product of rockfall.

In front of the retreating ice-sheet: Fauna complex of central-western Poland in MIS 3–2 (Krosinko site)

A new locality in Central Europe (Krosinko, Poland) yields a rich fossil material of great taxonomical variety. Its location in the Warsaw–Berlin ice-marginal valley, where it forms a 1-m horizon at a the depth of 6–8 m, is suited to a taphonomic analysis aimed at reconstructing the paleoevents and depositional history of the late Pleistocene (MIS 3–2).Our analysis revealed that the material accumulated as the result of high-energy fluvial transport. This transport disturbed the original bone associations, as indicated by the lack of articulation, low refitting, and bone surface modification. During water movement, the bones were polished, abraded, pitted, and fragmented. Prior to that, however, taphonomic signatures were left behind by predators, humans, and plant roots. This is the first time that unquestionable burning marks have been found on woolly rhino remains from Poland.The secondary context of the faunal assemblage was confirmed by neodymium isotope studies of the faunal fragments and sediment. Given the results of sedimentological analysis from previous studies, these sediments can be associated with a high-energy braided river. The remains were moved along with paleobotanical material and flints in a westerly direction; this is suggested by the direction of water flows in the valley and the river waters supplying it.The deposition occurred within a single geologic event around 26 ka BP, as indicated by the youngest radiocarbon date of the remains, or ca. 19 ka BP, considering the geochronological constraints for the local LGM ice margin position in central and eastern Poland ∼22–18 ka, and the specified accumulation time of the sediments in which they were found. Secondary mineralization of manganese and iron on the bones was the final taphonomical process that impacted the remains.The combined taphonomic, palaeozoological, isotopic and geological results show for the first time for a Polish paleontological site how data integration improves interpretation and highlights the advantages of a multiproxy approach.

A deep learning-based taphonomical approach to distinguish the modifying agent in the Late Pleistocene site of Toll Cave (Barcelona, Spain)

ABSTRACT One of the most widely used methods to associate lithic tools and bone assemblage in archaeological sites is the identification of cut-marks. However, the identification of these marks is still problematic in some localities on account of the similarities with the modifications generated by non-human processes, including biostratinomic and post-depositional bone surface modifications. Toll Cave (Barcelona, Spain), with chronologies between 47.310 BP and 69.800 BP, is one of the case studies where the cut-marks could easily be confused with abundant grooves generated by the dragging of sedimentary particles (e.g. trampling), but also with the scores produced by carnivores. In this work, we present the results obtained from applying Deep Learning (DL) models to the taphonomic analysis of the site. This methodological approach has allowed us to distinguish the bone surface modifications with 97.5% reliability. We show the usefulness of this technique to help solve many taphonomic equifinality problems in the archaeological assemblages, as well as the need to implement new approaches to eliminate subjectivity in the descriptions of bone damage and make more accurate inferences about the past.

Big questions for small animals. Taphonomic analysis of small vertebrates in Matja Kuru 2 (Timor-Leste) during the late pleistocene

The analysis of subsistence practices in the Lesser Sunda Islands (from Bali in the west to Wetar and Timor to the east) provides data to interpret anatomically modern humans’ subsistence adaptation in insular environments. Most previous analyses of small vertebrate assemblages in these islands have focused on sites with evidence of intensive coastal subsistence, where the fish and other marine faunal remains can be assigned unequivocally to human predation. Inland sites, however, contain large quantities of murids, reptiles and bats of varying sizes, which could have been accumulated by different predators.Here we present the first detailed taphonomic analysis of the faunal assemblage recovered from the Pleistocene levels of the inland cave site Matja Kuru 2 (Timor-Leste), dominated by small murids, followed by large and giant murids, reptiles, birds, bats and fish bones. Through the study of skeletal representation, fragmentation and bone surface modifications in the large and giant murids, turtle and fish remains, we aim to determine whether humans were responsible for their presence in the older units of the site, whether they comprise natural deaths, or were prey of non-human predators such as raptors.Our analyses indicate differences in predator actions according to age of prey and taxa, with adult large and giant murids, freshwater turtles and fish showing evidence of human predation. The results presented here contribute significantly to understanding human behaviour in insular environments and deposition dynamics in karstic taphosystems in Island Southeast Asia, as well as providing a framework for future taphonomic studies in the region.

ASSESSING PALEOECOLOGICAL DATA RETENTION AMONG DISPARATE FIELD COLLECTION REGIMES: A CASE STUDY AT THE MYGATT-MOORE QUARRY (MORRISON FORMATION)

ABSTRACT Field work protocols in the recovery of vertebrate fossils can vary between sites, and also within sites, due to differing researcher goals. Disparate researcher priorities can affect the resulting collections in terms of species richness, size distribution, specimen completeness, taphonomic condition, and aesthetic value. We examined paleoecological data, in the form of bone surface modifications (e.g., abrasion, feeding traces, etc.), from a single site worked by multiple collectors to determine the sensitivity of this type of data to collector bias. We examined 2,368 fossils from the Mygatt-Moore Quarry and divided them into two cohorts: 2016–2019 (bulk collection under a single collector) and pre-2016 (mixed collectors and priorities). Frequencies of modified bone surfaces were then calculated in each cohort among the recovered specimens. However, the specimens within the cohorts were of unequal size, completeness, and amount of preserved surface area, making inferences of modified bone surface frequencies difficult. To correct for unequal morphologies and preservation, we estimated the percentage of altered surface area among specimens by overlaying photos with a 4.0 cm2 digital grid to create a digital set of equal sized fragments. With such a large dataset, we took a random 10% subsample of specimens from each cohort for the grid study. We estimated the sample size needed to accurately reflect the frequency of bone surface modifications by specimen and surface area for each cohort. Results show the recovery of modified bone surfaces between the two cohorts was highly disparate, and potentially sensitive to the effects of collector bias when using specimen-level data. However, frequencies based on estimates of surface area were much more consistent and appeared to equalize data between cohorts and showed little influence of collector bias on data recovery. Thus, the traditional method of calculating frequencies using specimen-level data may create an illusion of bias that is removed when frequencies are calculated from estimated bone surface area. We posit that the digital fragmentation method is more informative when comparing paleoecological traces between datasets and should be applied to fossil assemblages going forward, especially when collection protocols between assemblages are significantly different or unknown.

Human-bird interaction in last Interglacial Iberia: A combined approach using skeletal part analysis, bone surface modification, bird ethology and ethnography

The procurement of small prey by Palaeolithic hunter-gatherers has been much discussed over the past decades, particularly with regards to birds. Birds provide information about the different ecological niches available to exploitation and contribute to the clarification of basic aspects of human behaviour and subsistence. Here, we report on the Middle Palaeolithic bird bone assemblages from reliably dated (by U-series and OSL) MIS-5 deposits of two cave sites in central Portugal with abundant stone tool assemblages and Neanderthal skeletal remains: Gruta da Figueira Brava and Gruta da Oliveira. The associated vertebrate bone assemblages document human processing of both large and small animals, including tortoises; carnivore activity is scant. Avifaunal remains can be naturally deposited in caves, or accumulated by hominin and/or carnivore/raptor activity. The collection from Gruta da Oliveira is almost exclusively composed of terrestrial species, with a predominance of corvids that are known to nest in caves and may have integrated the archaeological deposit due to natural deaths; a few, however, bear anthropogenic modification marks that, in light of the evidence from skeletal part analysis and other archaeological and ethnographic examples, suggest that they may have been targeted by Neanderthals. The assemblage from Gruta da Figueira Brava comprises marine and terrestrial taxa, including edible species, as well as diurnal and nocturnal raptors. The marine assemblage is anthropogenic and reflects the exploitation of rocky cliff and marshy/estuarine environments ∼2000 m away.

Taphonomic skeletal disturbances in the Sima de los Huesos postcranial remains.

The postcranial skeleton of fossil hominins is crucial for reconstructing the processes that occurred between the time of death and the recovery of the bones. Thousands of postcranial skeletal fragments from at least 29 hominin individuals have been recovered from the Sima de los Huesos Middle Pleistocene site in Spain. This study's primary objective is to address the main taphonomic features of the postcranial remains from the Sima de los Huesos sample, including antemortem, perimortem, and postmortem skeletal disturbances. We present an updated assessment of the bone surface modification analysis, the fracture pattern analysis, and the skeletal part representation to facilitate interpretation of the biostratinomic and fossil-diagenetic processes in this large paleoanthropological collection. We conclude that carnivores (probably bears) had limited access to the hominin bones and complete bodies were probably placed in the site.

An experimental approach to the analysis of altered cut marks in archaeological contexts from Geometrics Morphometrics

Geometric Morphometrics (GMMs) is a useful tool for the experimental characterization of different taphonomic modifications on bone surfaces, including anthropogenic cut marks. Until now, most research has been conducted evaluating experimental and non-altered cut marks, in order to obtain a better control of the studied variables. However, in archaeological contexts, bones are frequently subject to different taphonomic alterations and, in most cases, bone surfaces and the associated taphonomic signals show different grades of alteration. In this study, we present an experimental approach in which experimental cut marks were produced, mechanically altered, digitized, and analyzed using GMMs, in order to provide an in-depth characterization of how overlying taphonomic traces can alter cut marks. In the experiment, 40 cut marks were subjected to abrasion by a mixture of distilled water, sands, and gravels, using a tumbling machine in a series of cumulative cycles for a total of 4 h: cycle 1 (0.5′), cycle 2 (0.5′), cycle 3 (1 h) and cycle 4 (2 h). Cut mark cross-section profiles were then obtained using 3D digital microscopy, and a 2D 7-landmark designed for cut mark analyses were employed. Our results show a clear statistical differentiation among non-altered cut marks belonging to Cycle 0, and altered cut marks. The depth and opening angle of cut marks appear to be the features with a higher weight in the differentiation between altered and non-altered cut marks. On the other hand, there are no distinctions when differentiating between the diverse altered cycles. These results, like those previously obtained in an experimental approach using a binocular microscope, point to the limitation of the identification of cut marks and other taphonomic signals in archaeological contexts with altered bone surfaces, caused by the fast alteration and the loss of diagnostic criteria of the bone surface modifications.

Reassessing the role of carnivores in the formation of FLK North 3 (Olduvai Gorge, Tanzania): A pilot taphonomic analysis using Artificial Intelligence tools

FLK North (FLK N) (Bed I, Olduvai Gorge, Tanzania) is one of the best examples of a palimpsest where felids, hyenids and hominins made use of the same space without or with minimal interaction between hominins and the other two carnivores. Felids have been interpreted as the main accumulators and carcass consumers followed by frequent hyenid intervention. The presence of hominins at this site has been documented through the discovery of stone tools. Here, we test previous taphonomic interpretations of this site through the application of Artificial Intelligence (AI) tools (computer vision applied to bidimensional images of tooth pits) to taxonomically discriminate carnivore-made tooth marks. The bones we analyzed constitute a small sample, being a preliminary study of bone surface modifications (BSM) through the application of AI to a sample of the FLK N archaeofaunal assemblage (mostly to Level 3 fossils), pending access to the larger excavated collections. The results obtained in the present study show that the marks analyzed have been generated both by hyenids and felids. The slight predominance of hyena tooth marks is expected, since the bone sample used is dominated by long limb bones, and hyenas are the most likely agent causing long bone breakage, although felids also break bones of carcasses smaller than 150 kg as documented in the site. Felid impact, in at least three cases, is documented with tooth marks imprinted by felids and hyenas occurring on the same specimens. Felid-hyenid interaction is, thus, documented though the deep learning methods applied. The limited number of specimens where both agents are documented suggest that both hyenids and felids were independently breaking a substantial part of the bones at FLK N. This preliminarily modifies previous interpretations that attributed most long bone fragmentation exclusively to hyenas.

Human occupational intensity and palaeoecology at Klasies River from MIS 5–3: Preliminary taphonomic analyses of faunal remains from the Deacon and Wurz excavations

Klasies River is a key site in understanding the behaviour of coastal foragers in the southern Cape of South Africa. Faunal remains from Klasies River, recovered from an almost 70,000 year sequence over the Late Pleistocene, are a valuable dataset in investigating subsistence and occupational dynamics, yet few taphonomic studies have been conducted on the fauna from Klasies River. Here, the first taphonomic analyses of faunal remains from the Deacon and Wurz excavation is presented. The aim of this paper is to link occupational patterns during the Late Pleistocene at Klasies River with coastal ecology using taxonomic data from previous studies, and recently analysed taphonomic data. Taphonomic analyses of samples from the MSA II Lower, the Howiesons Poort and the MSA III periods were used to examine occupational intensity based on proportions of anthropogenic and zoogenic bone surface modification, trampling marks, transverse fractures and faunal density. Declining foraging efficiency or subsistence intensification was also investigated using indicators for expanded diet breadth (e.g., small fauna exploitation and increases in juveniles), more intensive marrow extraction and evidence for greater foraging distances. Herbivore functional types and dietary preferences were examined using taxonomic data and these show three main ecological phases at Klasies River: the MSA I; the upper and lower MSA II; and the HP and MSA III phases. Taphonomic data show increased occupational intensity in the MSA II Lower, with subsistence intensification also more evident here than other layers. Trampling data suggests that occupational intensity was greater in the earlier Howiesons Poort than later phase with little evidence of declining foraging efficiency then. The data indicate that, while humans contributed significantly to the MSA III assemblage, carnivores were the main accumulators here suggesting lower occupational intensity in this phase. The results of this study shows a possible link between increased occupational intensity and declining foraging efficiency at Klasies River but larger taphonomic samples are needed to explore this relationship further.