Citation
Human adaptability in the time of changing climate : a zooarchaeological analysis at Pinnacle Point Site 5-6, Mossel Bay, South Africa

Material Information

Title:
Human adaptability in the time of changing climate : a zooarchaeological analysis at Pinnacle Point Site 5-6, Mossel Bay, South Africa
Creator:
Simeonoff, Sarah Maureen
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Master's ( Master of arts)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Anthropology, CU Denver
Degree Disciplines:
Anthropology
Committee Chair:
Hodgkinds, Jamie M.
Committee Members:
Beekman, Chris
Stone, Tammy

Notes

Abstract:
The adaptability of modern humans is often cited as key to our successful global survival and proliferation; however, little work has been done to quantify this trait through time. This project examines the Marine Isotope Stage (MIS) 5-4 transition occurring ~81 – 71 thousand years ago (ka) on the southern coast of South Africa at an Upper/Middle Paleolithic site on the southern coast of South Africa. During this time, the environment underwent challenges related to transgressing and regressing coastlines, changing faunal availability, and rapid changes in mean temperature. A zooarchaeological analysis was undertaken to determine if and to what extent the modern humans occupying the cave modified their subsistence strategies during this climactic shift. The results of this project indicate a temporary change in subsistence strategies occurred at the onset of glacial MIS 4, followed by a return to the behaviors of MIS 5 as the climate fully transitioned. These findings suggest the population occupying PP5-6 was sufficiently behaviorally elastic as to withstand environmental changes related to resource availability. This study provides insights into the adaptability of modern humans, and further provides a record against which the adaptability of other hominid species can be tested.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
Copyright [name of copyright holder or Creator or Publisher as appropriate]. Permission granted to University of Colorado Denver to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.

Downloads

This item has the following downloads:


Full Text
HUMAN ADAPTABILITY IN THE TIME OF CHANGING CLIMATE:
A ZOOARCHAEOLOGICAL ANALYSIS AT PINNACLE POINT SITE 5-6, MOSSEL BAY, SOUTH AFRICA
by
SARAH MAUREEN SIMEONOFF B.A., Arizona State University, 2013
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Arts Anthropology Department
2019


©2019
SARAH MAUREEN SIMEONOFF ALL RIGHTS RESERVED
11


This thesis for the Masters of Arts degree by Sarah Maureen Simeonoff has been approved for the Anthropology Program
by
Jamie M. Hodgkins, Chair Chris Beekman Tammy Stone
Date: July 27,2019


Simeonoff, Sarah Maureen (M.A., Anthropology)
Human Adaptability in the time of Changing Climate: A Zooarchaeological Analysis at Pinnacle Point site 5-6, Mossel Bay, South Africa
Thesis directed by Associate Professor Jamie M. Hodgkins
ABSTRACT
The adaptability of modern humans is often cited as key to our successful global survival and proliferation; however, little work has been done to quantify this trait through time. This project examines the Marine Isotope Stage (MIS) 5-4 transition occurring ~81 -71 thousand years ago (ka) on the southern coast of South Africa at an Upper/Middle Paleolithic site on the southern coast of South Africa. During this time, the environment underwent challenges related to transgressing and regressing coastlines, changing faunal availability, and rapid changes in mean temperature. A zooarchaeological analysis was undertaken to determine if and to what extent the modern humans occupying the cave modified their subsistence strategies during this climactic shift. The results of this project indicate a temporary change in subsistence strategies occurred at the onset of glacial MIS 4, followed by a return to the behaviors of MIS 5 as the climate fully transitioned. These findings suggest the population occupying PP5-6 was sufficiently behaviorally elastic as to withstand environmental changes related to resource availability. This study provides insights into the adaptability of modern humans, and further provides a record against which the adaptability of other hominid species can be tested.
This form and content of this abstract are approved. I recommend its publication.
Approved: Jamie M. Hodgkins
IV


ACKNOWLEDGEMENTS
I am grateful to my committee, Jamie Hodgkins, Chris Beekman, and Tammy Stone, for their helpful comments and support. In particular, I would like to thank my advisor, Jamie Hodgkins, for her mentorship and guidance. My sincerest gratitude to Curtis Marean for setting the tarsal-identification bar so high that I will be aspiring to it for the rest of my career. The opportunity to work at Pinnacle Point (both in the field and the lab] was a gift.
Without the ceaseless support and encouragement of my parents, Mary and Richard Simeonoff, I wouldn’t have had the confidence to apply for and finish graduate school. I appreciate the constant reminders to work hard and to trust my abilities; making you proud has always motivated me to do better. To the best brothers to ever live, Richard and John, please eat my dust. Thanks for always putting this degree into perspective. I only did it to make you guys look bad.
My sincerest gratitude to HK Gatti for pulling no punches in her approach to encouragement, and for her insistence that I "just do it already.” It was a critical push near the end, but the gentler cheering was also appreciated. In a similar vein, I want to thank Darsita Ryan for always being willing to tell me to get my act together and write, and for offering a sympathetic ear and honest advice.
R. Jackson Bewley, thank you for showing up at the exact right moment. Your support was crucial in keeping me calm and focused. Thank you for always reminding me about the big picture in little ways. The prospect of using this degree to win arguments gave me a strong incentive to finish.
Last and never least, I could not have finished this degree without the theory-deciphering commiseration, comic relief, and true friendship of my partner in crime and archaeology, Breeanna Chantel Charolla. Without you as a teammate has never been an option - thank you for everything. We may as well be co-authors on this bad boy for all of the emotional support, encouragement, editing, and will to live you offered during its labor. There's no one else I would have preferred to suffer alongside.


TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION..............................................................1
Theoretical Basis..........................................................4
Pinnacle Point Complex.....................................................7
Environmental Setting......................................................7
Pinnacle Point Archaeological Context......................................9
Marine Isotope Stage 5-4 Transition at PP5-6..............................13
Light Brown Sand and Roofspall (LBSR).....................................15
Ashy Light Brown Sand (ALBS)..............................................17
Shelly Ashy Dark Brown Sand (SADBS).......................................19
Toba Eruption and Glacial Winter..........................................22
II. METHODS..................................................................25
Excavation Techniques at Pinnacle Point...................................25
Sampling MIS 5-4 Transitional Fauna.......................................27
Zooarchaeological Analysis................................................28
Taxonomic Identification..................................................31
Body Size.................................................................31
Fragment Size.............................................................28
Weathering................................................................29
Surface Visibility........................................................30
Burning...................................................................32
Transportation Strategies.................................................32
Breakage Analysis.........................................................35
Surface Modifications.....................................................37
3D GIS Visualization......................................................42
III. RESULTS..................................................................44
Fragment Size.............................................................44
Weathering................................................................45
vi


Surface Visibility.........................................................46
Taxonomic Identification...................................................47
Body Size..................................................................49
Burning....................................................................52
Nutritive and Non-nutritive Breakage.......................................53
Food Transport.............................................................57
Surface Modification.......................................................60
3DGIS......................................................................62
IV. DISCUSSION AND CONCLUSION...................................................69
LBSR Discussion............................................................69
ALBS Discussion............................................................70
SADBS Discussion...........................................................72
Conclusions................................................................73
V. REFERENCES..................................................................73
vii


LIST OF TABLES
TABLE
Table 1. Coded specimens per StratAgg..............................................21
Table 2. Body size categories (Brain 1981) and example South African fauna (Thompson
2010)....................................................................32
Table 3. Observed weather stages by StratAgg......................................46
Table 4. Comparison between levels of the NISP coded for each Family, Order, Class by
Body Size. Each specimen was coded to the smallest unit of taxonomic unit.51
Table 5. Minimum number of elements, MAU, SFUI and normed MAU values for each
StratAgg, used to create Figure 2. Created using SFUI, as calculated by Metcalfe
and Jones (1981).........................................................58
Table 6. Summary table of the key differences between StratAggs during the MIS 5-4
transition...............................................................75
viii


LIST OF FIGURES
FIGURE
Figure 1. Location of Pinnacle Point relative to the coast. Sites PP13B and PP5-6 are also
shown in relation to each other. Photo published in Karkanas et al. 2015.....13
Figure 2. Lateral view of site, facing east of StratAggs from PP5-6, with the relevant sections highlighted. Figure created by Erich Fisher (University of Colorado
Denver).......................................................................14
Figure 3. Average, minimum, and maximum coastal distances for each StratAgg, as
modeled by Fisher et al. 2010.................................................22
Figure 4. Excavation underway at PP5-6. SubAggs are identified on the wall using tags,
with StratAggs marked with larger tags. Photo facing east....................27
Figure 5. Examples of cut marks from PP5-6, showing the relatively straight morphology
and clustering of marks.......................................................39
Figure 6. Sample of percussion modification marks. Clockwise from top left:
microstriations, conchoidal flake scar, flake with two medullary percussion notches, percussion notch on cortical surface. Examples from unpublished
experimental data by author...................................................41
Figure 7. Average length and width for each StratAgg (millimeters).....................45
Figure 8. Surface visibility by StratAgg...............................................47
Figure 9. Proportion of identifiable taxa for each StratAgg. Taxa representing less than one
percent of the assemblage are not reported in this figure....................49
Figure 10. Relative frequencies of faunal body sizes through the transition...........50
IX


Figure 11. Percentage burned material per StratAgg, including percent fully calcined,
partially calcined, fully carbonized, and partially carbonized.........53
Figure 12. Non-nutritive breakage for each StratAgg, compared to experimental data from
Marean et al. 2000.....................................................55
Figure 13. Nutritive breakage for each StratAgg, compared to experimental data from
Marean et al. 2000.....................................................56
Figure 14. Distance-weighted mean squares curve fitted to each StratAgg and plotted
against Binford’s Food Transport Utility curves (Binford 1978, Faith and Gordon
2007).........................................................................59
Figure 15. The results of the ANCOVA with Poisson distribution to determine the least-squares mean number of cut marks per fragment, controlled for surface area (calculated using the geometric mean). The error bars show the 95% confidence
intervals for each StratAgg...................................................61
Figure 16. The results of the ANCOVA with Poisson distribution to determine the least-squares mean number of percussion marks per fragment, controlled for surface area (calculated using the geometric mean). The error bars show the 95%
confidence intervals for each StratAgg.........................................62
Figure 17. Analyzed fauna from the transition, identified by taxon.......................65
Figure 18. Visualization of faunal remains located within combustion features versus those
located in other contexts......................................................66
Figure 19. Burned bones, displayed by level of burning across the transition.............67
x


Figure 20. Anthropogenically modified bones shown in association with bones recovered
from hearth contexts (in black)............................................68
Figure 21. Summary figure of key differences between StratAggs. Dashed lines indicate
usage of the secondary axis................................................76
xi


CHAPTER I
INTRODUCTION
Anatomically modern humans (AMH) have existed alone for 39,000 years (Higman 2014) and successfully inhabited nearly every environmental niche on the planet. In response to varied environmental dines and fluctuations of resource availability within these dines Homo sapiens have continually adapted. The planet has always undergone fluctuations in mean temperature which have modified environments across the globe (Martinson et al. 1987), necessitating behavioral modifications by humans to ensure their continued survival. Humans, with the help of modern technology, now number in the billions - by far the most successful hominid species ever to proliferate - and inhabit all continents. Enumerating and quantifying the ways in which AMH have modified their behaviors through time to adapt and expand is crucial to human origins research. To that end, this study examines a regularly-occupied Middle/Upper Paleolithic site and focuses on human subsistence strategies across a global climatic transition from an interglacial to a glacial climate.
The extinction of numerous hominid species has been attributed to an inability to successfully adapt to changing environments. No other hominid better exemplifies the debate for human superior adaptability than the demise of Homo neanderthalensis. H. neanderthalensis disappeared from the archaeological record approximately 39 kya, coinciding roughly with the appearance of anatomically modern humans (AMH) in Europe (Higman 2014). In attempting to understand the H. neanderthalensis extinction, researchers have focused on the asserted superior adaptability of AMH (Villa and Roebucks
1


2014 provide a thorough discussion of the argument). Although this comparison is not paramount to this research, it best illustrates the oft-touted adaptability of AMH. With regard to subsistence strategies, several authors have argued both for and against the idea that humans are more adaptable to varied and changing climates than Neanderthals (Mellers 1996, McBrearty and Brooks 2000, Marean 2005). Although no comparison between Neanderthal intelligence, symbology, or strategies are undertaken during this research, it is the adaptability of AMH that frames my research at Pinnacle Point. Did the AMH occupying the site during a glacial-interglacial shift successfully adapt their strategies in the face of environmental change? If so, an example of early human adaptability will be illuminated. Although this study alone cannot adequately solve the debate on human adaptability, it can offer a record against which future studies (from different geographic, temporal, and occupant sites) can be compared.
Specifically, this project examines the Marine Isotope Stage (MIS) 5-4 transition occurring ~81 - 71 thousand years ago (ka) on the southern coast of South Africa. During this time, the environment underwent challenges related to transgressing and regressing coastlines, changing faunal availability, and rapid changes in mean temperature (Karkanas et al. 2015, Lisiecki and Raymo 2005, Martinson et al. 1987, Wilkins et al. 2017). Should the humans at Pinnacle Point exhibit an exceptional ability to mold their behavior in the face of complex climatic change, this may represent a key to understanding modern human’s global success.
Through zooarchaeological analysis, this research seeks to determine if and to what extent environmental changes on the southern coast of South Africa influenced changes in
2


the subsistence strategies of the anatomically modern humans occupying Pinnacle Point Site 5-6 (PP5-6). The Pinnacle Point Site Complex (Pinnacle Point) is a cluster of caves, rockshelters, and open-air sites very near the modern-day South African coast (<30 meters) (Kaplan 1997, Marean et al. 2007, Fisher et al. 2010, Jacobs 2010). The archaeological sequence at PP5-6 begins at approximately 90 ka and continues through 55 ka, encompassing three Marine Isotope Stages (MIS) and two transitions (from 5-4-3, Interglacial-Glacial-Interglacial) (Karkanas etal. 2015, Wilkins etal. 2017), and its location near both terrestrial and coastal resources makes it an ideal site to study human adaptation to different biomes (coastal and plains).
Understanding past subsistence, particularly at a site complex which has been regularly occupied, is crucial to decoding the evolution of modern humans relative to other hominid groups. This study begins by discussing the theoretical basis used to develop the question, then describes the zooarchaeological analysis undertaken, and finally provides the results and implications of this research.
3


CHAPTER II
Background
Theoretical Basis
Pinnacle Point 5-6 (PP5-6), the site at the center of this research, was routinely occupied between 90 - 55 ka (Karkanas et al. 2015, Wilkins et al. 2017). The limitations of studying a site and time period so deep in the past requires that the theoretical approach appropriately match the dearth of culturally-modified material (e.g. ochre, incised shells, beads, etc.). Archaeological assemblages dated to the Middle/Upper Paleolithic are characterized by utilitarian remains; that is, they generally lack culture-signifying artifacts and instead typically contain lithic tools and debris, faunal remains, combustion features, and little else. Because there are no clear social strata, very few instances of cultural material, and a lack of comparable archaeological data, both the symbolic and political economy theoretical schools will be rejected in favor of a processual approach.
In the early 1960’s, Lewis Binford proposed a change to the prevailing theoretical basis of archaeological inquiry (Binford 1962,1964), arguing that "traditional archaeology” needed a systematic framework better suited to the theoretical questions being posed. Traditional archaeology, Binford suggested, suffered from a paradox in which cultural ideology is identified and explained by the material culture alone (Binford 1989). Processual archaeology steers clear of addressing "the ancient mind" in favor of using physical artifacts to inform past socio-economic and cultural conditions (Renfrew 1994). In the processual view, the primary motivation of cultural shifts is environmental change (Binford 1962,1965,1989). Culture, then, is a means of adapting to changing environments
4


- "an extrasomatic adaptive system that is employed in the integration of society with its environment” (Binford 1965:203).
Processualists are positivist in their view of archaeology and believe concrete answers to archaeological inquiry can be found through a study of "total cultural systems” (Binford 1962:224). Because processualism views culture as a response to the environment, the entire system is the preferred unit of analysis, as it constitutes a homeostatic system that responds to environmental change through adaptation (Binford 1964,1989, Renfrew 1994). Binford, in particular, emphasized the need to look at all changes within a system. Because the available data does not allow research to focus on the isolation of one change, the only way to study the full scope of a society is to investigate the sum of its parts (Binford 1965). Objective analysis of archaeological remains as they relate to the past is the driving force behind processual methodology. There is a strong preference for a deductive middle-range approach where "claims that fall between observational descriptions of what the archaeologists find in the present...and the description reconstructions of the past” are routinely employed (Kosso 1991: 622).
The question posed by this research is predicated on the processual notion that humans adapt to changes in their environment; this particular portion of the PP5-6 sequence was selected because of the known climatic changes occurring globally (e.g. interglacial to glacial). Further, the limiting factors of faunal analysis require the scope of the question be focused on subsistence changes between levels. Because little additional data are available on the social strata or culturally-modified artifacts, this analysis will
5


focus on changes in food procurement and processing because it is a sub-system that can be isolated from the broader system and analyzed separately.
The processual approach affects not only the development of the question but also the understanding of the results. Should the analysis reveal humans modified their subsistence strategies in response to the cooler temperatures and changing faunal availability of MIS 4, processual archaeology suggests the change indicates a well-developed ability to respond to the environment. Alternatively, should the occupants show no sign of changing strategies, processual archaeology suggests the environmental changes occurring at Pinnacle Point during the transition were not substantial enough to require behavior modifications. The processual approach is best-suited to this undertaking, given the paucity of cultural evidence and focus on humans’ direct interaction with the environment.
6


Pinnacle Point Complex
Environmental Setting
The Pinnacle Point cave complex (PP) is located approximately 10 km west of the town of Mossel Bay, on the Western Cape of South Africa. The southern coast of South Africa exists in a unique climate, at the "intersection of subtropical and temperate climate regions, as well as the Indian, Atlantic and Southern oceans” (Esteban et al. 2019: 3). The interaction of these various systems means the area is subjected to both winter rains from the west and bimodal rains (a combination of summer and year-round rains) from the center (Copeland et al. 2016), resulting from the warm waters of the Indian Ocean (Engelbracht et al. 2015). Strontium isotope analysis suggests the same rainfall regime would have persisted during the Middle/Upper Paleolithic (Sealy 1996, Bar-Matthews et al. 2010, Copeland et al. 2016). Mossel Bay is situated in the year-round rainfall region, though the boundary for winter rainfall and summer rainfall regimes fall close by (Braun et al. 2018). The climate is subsequently moderate, averaging 18° C (64.4° F), with seldom frost (Esteban et al. 2019).
Mossel Bay, and Pinnacle Point, is situated within the Greater Cape Floristic Region (GCFR), known for its high diversity of plant species (Cowling and Lombard 2002, Esteban et al. 2019, Goldblatt 1997,). The GCFR consists of five distinct biomes: fynbos, renosterveld, succulent karoo, thicket, and forest (Bergh et al. 2014). Of these, fynbos, renosterveld, and thicket are situated near Pinnacle Point, with fynbos comprising the largest proportion of the extant environment. These three biomes represent distinct plant species and environments. Fynbos contains the greatest diversity of plant species and is
7


characterized by "fine-leaved ericoid shrubs and shrublets” with few trees (Bergh et al. 2014:11). It is considered one of the richest floral regions in the world, with a high number of endemic species (Bergh et al. 2014). Renosterveld is also characterized by small-leaved shrubs; however, this biome is dominated by more C3 grasses and features more nutrient-rich soils than fynbos (Bergh et al. 2014: 15). Finally, the thicket biome "is defined by the possession of a large diversity of growth forms, creating a dense canopy of largely evergreen, broadleaved, sclerophyllous, spiny and/or succulent shrubs and low trees” (Bergh etal. 2014: 8).
Of the three biomes, fynbos is the dominate environment surrounding Pinnacle Point and the extant fauna living in these environments are distinct from the rest of Africa (Rector and Reed 2010). Mammalian diversity is relatively low in fynbos (Klein 1983, Marean 2010) and was likely similar through time (Rector and Reed 2010). Large-bodied mammals are uncommon - due to the nutrient-poor soils found in the fynbos biome - thus suggesting the environment was likely never conducive to large mammal hunting (Klein 1983, Marean 2010). Extant small-bodied species include "antelope, hyrax, dune mole rats, and tortoises” (Marean 2010: 431). Klein (1983) called the mammalian fauna in the Pinnacle Point region "notably impoverished” relative to other fynbos regions in South Africa, additionally citing grey duiker, steenbok, and grysbok as common species. In the past, however, large-bodied herbivores (such as Wildebeest, giant zebra, and hartebeest) would have existed on the landscape (Rector and Reed 2010, Copeland et al. 2016). Published reports on the faunal assemblages from Pinnacle Point show that the majority of faunal remains recovered are not identifiable to taxon (Thompson 2010, Rector and Reed
8


2010, Marean 2010); however, the common species and diversity in fynbos is important to understanding the archaeological assemblage at Pinnacle Point.
Pinnacle Point Archaeological Context
Included in the Pinnacle Point Complex are 28 sites (21 of which are dated to the Middle Stone Age ~20 ka - 200 ka) that were identified and recorded during a cultural resource management survey of the area prior to the development of the Pinnacle Point Estate (Jacobs 2010, Kaplan 1997, Marean etal. 2007). In particular, fifteen cave and rockshelter sites are clustered near the coast, averaging 50 meters in elevation and situated within 30 meters from the modern coast below (Fisher et. al 2010). This places them within both coastal and plains biomes, with access to nutrients from both terrestrial and intertidal resources.
This portion of the South African coast is situated against the Agulhas Plain, a "gently sloping and broad continental shelf” (Fisher et al. 2010: 1384). This bank is relatively flat-lying and directly abuts the modern coast. During the change from interglacial to glacial climate, oceans contract to ice cores and cause sea levels to drop (Shackelton and Opdyke 1973). For the coast near PP, this causes a drastic change in environment (Fisher et al. 2010, Copeland et al. 2016) by exposing a broad floodplain (Cawthra et al. 2015), the ecology of which is still unknown (Copeland et al. 2016). A model of the Paleo-Agulhas Plain (PAP) across glacial cycles was undertaken to better understand the impact of changing climate on the exposed landmass of the PAP (Fisher et al. 2010).
This model shows that at peak regression of the ocean approximately 80,000 km2 of grassland would be exposed between PP and the coast (Fisher et al. 2010, Copeland et al.
9


2016). Although the model only accounts for coastal distance in broad ranges (i.e. minimum, maximum, average distances) and cannot account for the exact coastal difference at precise dates, it helpfully provides comparative shoreline distances. This is, while the model does not wholly account for the distance of the coast at any given time, it provides useful comparative distances across large time spans, such as the transition from interglacial to glacial climate. Even subtle shifts (i.e. 5 or 10 km) in the ocean levels would have altered the coastal environment surrounding PP (Marean et al. 2014). The exact nature of the exposed PAP is not well-understood; however, some researchers suggest the broad open grassland would have provided ample resources for large ungulates (Cawthra et al. 2015, Copeland et al. 2016). Thus, the movement of the coast and exposure of the PAP during the MIS 5 to 4 transition at PP is a key factor in understanding the resource procurement of PP5-6 inhabitants.
Excavation of five cave sites from the complex was undertaken by Curtis Marean and Peter Nielssen in 1999. Since the initial excavations, control of the work has shifted to the South African Coast Paleoclimate Paleoenvironment Paleoecology Paleoanthropology Project (SACP4). SACP4’s research focus is early human adaptability at Pinnacle Point through analysis of the use of symbology, lithic technology, and subsistence strategies across various changes in climate and environment (Marean 2010a). To date, several sites from Pinnacle Point have been tested or partially excavated, including PP5-6, the focus of this thesis (see Marean et al. 2004, Marean 2010b, 2010c). Another cave site was previously excavated and analyzed at Pinnacle Point: Pinnacle Point Cave 13B (PP13B), located less than one kilometer from PP5-6. A brief discussion about the information
10


derived from the record at PP13B provides useful insight into the archaeological history of Pinnacle Point and offers a backdrop for the analysis of PP5-6.
Excavations of PP13B revealed the earliest known evidence of intertidal marine shellfish exploitation by humans at approximately 164 ka (Marean et al. 2007), pushing the date of this behavior back in time approximately 40 ka (Erlandson 2001, Walter et al.
2000). Intertidal shellfish exploitation by humans requires an understanding of the lunar cycle, and, Marean (2011:434) suggests:
"a complex cognition characterized by fully modern working memory and executive functions [which] allowed them to link lunar phases to tidal rhythms and thus, develop an effective way to schedule visits to the coast in a manner that maximized returns from the coastal resources.”
The adoption of intertidal foraging shows a change in cognition and behavior from a reliance on terrestrial resources to an expansion towards marine sources - a change which likely required an advanced cognition (cognitive adaption) to understand the tidal schedule (Marean 2014, Marean 2015, Marean 2016, Loftus et al. 2019). PP13B also revealed early evidence of pigment usage (Watts 2010) and bladelet production (Jacobs 2010), both beginning near 162 ka (Marean et al. 2007, Marean 2010c, Thompson et al. 2010). Marean and colleagues have posited that these changes in behavior were related to the changing glacial environment of MIS 6 (~196- 123 ka) (Marean 2010c). This leap in technological and cognitive ability suggests that the occupants of PP13B were adapting to the movement of the Paleolithic coastline and resource availability, laying the groundwork for further investigation into the MIS transition occurring at PP5-6 (Marean 2010c, Marean 2014).
PP13B provides a limited but fascinating record of human occupation in the region (or "snapshots through time” [Marean 2010c:440J) between approximately 164 - 90 ka at
11


which time a sand dune sealed the cave, preventing further occupation until 40 ka (Karkanas and Goldberg 2010, Marean et al. 2007, Marean 2010b). The record provided by PP13B provides interesting insights into early human activity along the coast of South Africa, however, the gap in the record (between 90 - 40 ka) leaves many questions unanswered about the behavior of anatomically modern humans through time (Marean 2010a).
The sequence at PP5-6, in contrast, begins at the time the sand dune sealed PP13B (~90 ka) and continues through 55 ka with little to no occupational interruption (Marean 2010a, Karkanas et al. 2015, Wilkins et al. 2017). The record preserved at PP5-6 presents an opportunity to study changes in subsistence strategies, technology manufacture, and site structure through time with the consistency of a single location (i.e. different cave occupations along the same coastal plain within PP). Of greatest interest to SACP4 is the rarely-sequenced transitional phases present at PP5-6; the site spans MIS stages 5-4-3, representing interglacial-glacial-interglacial transitions (Karkanas et al. 2015, Wilkins et al. 2017). This enables research to be conducted through major climatic shifts on a plethora of different questions by allowing for the isolation of climatic factors on behavioral changes (e.g. Karkanas et al. 2017, Wilkinson et al. 2017).
12


Figure 1. Location of Pinnacle Point relative to the coast. Sites PP13B and PP5-6 are also shown in relation to each other. Photo published in Karkanas et al. 2015.
Excavations at PP5-6 were closed in February 2017, ending the data collection phase at the site (Wilkins et al. 2017). Throughout the work undertaken by SACP4, excavation techniques were devised to maintain consistent control of data "through total station plotting, field stratigraphy, 3D GIS of plotted finds and lenses, and micromorphology with high resolution OSL [optically stimulated luminescence] dating” (Brown et al. 2012:8). The system of excavation has remained consistent, enabling continuity across field seasons, excavators, and within the site.
Marine Isotope Stage 5-4 Transition at PP5-6
Marine Isotope States were identified and described by Martinson et al. (1987), based on deep-sea sediment core analysis. MIS 5 is subdivided into five subcategories (a -
13


e) and spans 130 - 80 ka; however, only MIS 5a is relevant to this research and dates to 82 ka (Lisiecki and Raymo 2005). In contrast, MIS 4 consists of just one stage, dating to 71 ka (Lisiecki and Raymo 2005). Three StratAggs account for the anthropogenically-modified layers dated to the MIS 5-4 transitional period: the LBSR, the ALBS, and the SADBS (See Figure 2). The faunal remains from these layers account for approximately 21,000 individual specimens (33. 8% of the total faunal assemblage at PP5-6). The MIS 5-4 transition at PP5-6 has been characterized in four published reports: speleothem analysis (Bar-Matthews et al. 2010, Braun et al. 2018), lithic technology and raw material sourcing (Wilkins et al. 2017), sediment microfacies (Karkanas etal. 2015), and phytolith analysis (Esteban 2019). Each StratAgg will be characterized in the following sections.
PP5-6lKey Stratigraphic Aggregates
Figure 2. Lateral view of site, facing east of StratAggs from PP5-6, with the relevant sections highlighted. Figure created by Erich Fisher (Arizona State University).
14


Light Brown Sand and Roofspall (LBSR)
The LBSR represents little sedimentary or occupational change throughout the 4.5-meter deposit (Karkanas et al. 2015). The layer is situated at the lower end of the site and was one of the first StratAggs to be identified at PP5-6. The weighted-mean OSL date for this layer is 81 ± 4, placing it in firmly within interglacial MIS 5a. The LBSR is characterized by relatively thin lenses of combustion features, moderate mollusk remains, and frequent roofspall inclusions.
The sediment in the LBSR is fine-grained, with minimal trampling evident from the intact condition of the hearths (Karkanas et al. 2015:13). Study of the sediment shows water movement along the midline of the cave likely disturbed the LBSR sediment, resulting in modest material loss (Karkanas et al. 2015:14). This sediment loss and subsequent compaction near the center of the cave resulted in a shallow depression throughout the LBSR, though this is likely unrelated to human occupation (Karkanas et al. 2015:14). Based on micromorphological study of the layer, occupation was characterized by short, small group occupations evidenced by single-incident hearths (Karkanas et al. 2015:14). This might indicate that the modern humans roaming the landscape were either relatively few or were moving in small bands.
Mollusk shells scattered intermittently throughout the deposit indicate exploitation of the coast was occurring with some regularity but may not have represented a reliable source of nutrients. During this time, the ocean was close to the cave, not much more distant than it is today. The shore averaged 1.1 kilometer, settling at a maximum distance of 1.6 km and a minimum distance of 0.8 km (km) (Fisher et al. 2010, Karkanas et al. 2015).
15


This distance places the coast well within the 6-10 km daily foraging radius for hunter-gatherers (Kelly 1995). Thus, the coast was easily exploitable during this time, though the lack of dense shell middens points to a lower reliance on shellfish than terrestrial fauna during this time (Marean 2010c, Thompson 2010).
During the LBSR, the environment was likely moderate, with little dramatic climatic change through seasons. Analysis of phytoliths from the LBSR indicate the environment was a fynbos-renosterveld mosaic consisting primarily of C3 plants with C4 plants present in lesser proportions (Esteban et al. 2019). This is further substantiated by the speleothem record, which indicated MIS 5 was characterized by C3 vegetation and frequent, year-round rains (Braun et al. 2018.) These environmental indicators suggest fauna would be similar to those species inhabiting the area today.
Focus on the LBSR lithic assemblage has not been robust to date (Wilkins et al. 2017). Blades, flakes, shatter, retouched flakes, cores, and hammers were all noted from the assemblage, though broad trends have not yet been discussed. Quartzite collected from the nearby coast is the most common raw material used during the LBSR (Wilkins et al. 2017). During the LBSR, when the coast was located nearby, quartzite cobbles would have been routinely refreshed by ocean currents, thus providing a consistent source of good quality knapping materials within 10 km of the site (Wilkins et al. 2017). Brown et al. (2012) suggested raw material frequency would correlate to coastal distance; a hypothesis that has proved correct during the LBSR (Wilkins et al. 2017). Relative to the MIS 4, lithic artifacts are larger in size, suggesting less intense processing or termination at earlier reduction stages (Wilkins et al. 2017).
16


Ashy Light Brown Sand (ALBS)
The ALBS immediately overlies the LBSR and consists of 0.8 meter of fine-grained beige sand that slopes slightly towards the inside of the cave. This layer begins after a significant roof fall event that caps the LBSR and represents a major boundary between roofspall-rich layers and aeolian deposits at PP5-6 (Wilkins et al. 2017). The ALBS consists of heavy lenses of human occupation in an ashy sand matrix with frequent shell middens (defined as sediment supported by shell fragments) (Karkanas et al. 2015, Wilkins et al. 2017). The mean-weigh ted age of this layer is 72 ± 3 ka, placing it on the boundary of the transition from MIS 5 to MIS 4. Following previous analyses from PP5-6, though the mean-weighted age of this level is within the transitional period it is grouped in MIS 5 (per Karkanas et al. 2015 and Wilkins etal. 2017).
During the ALBS, the ocean receded to an average distance of 10.6 kilometers (km) from the cave and ranged from a minimum distance of 1.7 km to a maximum of 22.0 km (Fisher et al. 2010, Karkanas et al. 2017). The average distance is feasibly within the daily foraging radius, though very near its limits, while the maximum is well outside. This fact is interesting, giving the greater proportions of mollusk shells present at the site, which appear to signal a higher reliance on coastal resources than in the LBSR. This may indicate a preference for the reliable exploitation of ocean resources over terrestrial, a theory posited by Marean (2016).
Speleothem analysis of nearby caves indicates the environment would have remained mosaic, with a slight increase in summer rains (Braun et al. 2019). This study also indicated an increase in C4 grasses (Braun et al. 2019). Phytolith analyses concur with
17


these results and further indicate the increase in C4 grasses was modest (Esteban et al. 2019). An abundance of C4 grasses should have provided an attractive ecosystem for "gregarious, open-habitat ungulates” (Bar-Matthews etal. 2010: 2143), though the small increase in C4 plants illustrated by these recent studies (Braun et al. 2018, Esteban et al. 2019) may not present a great enough environmental change to support larger fauna. The retreat of the coast and concomitant presence of more abundant grassland on the exposed PAP during the ALBS suggests the environment would have been characterized by dense terrestrial fauna for exploitation (Klein 1983, Copeland etal. 2016, Hodgkins etal. in press). Recent analysis of strontium isotopes finds that the ungulates occupying the grassy PAP remained on the plain (Copeland et al. 2016) and thus would have served as a reliable source of prey for AMH during the ALBS.
Large hearths created in short intervals within the ALBS indicate that occupation of the cave became denser and more frequent than in the LBSR (Karkanas et al. 2015). This change is "roughly concordant with the transition from warmer conditions of MIS 5 to cooler conditions in MIS 4” and may indicate increased sedentism or larger group sizes during the glacial cycle (Karkanas et al. 2015: 19). As yet, the lithic sample from the ALBS is too small to provide insights into the lithic techno-typologies, though a few preliminary analyses are available. Wilkins et al. (2017) noted an absence of backed microliths and bifaces, and very few blades which are typical of sites dated to this time period. This study found the majority of lithic artifacts from the ALBS are comprised of quartzite - the nearest source of which is the coast. Although the coast was more distant than in the LBSR, the inhabitants of PP5-6 still relied on quartzite as their main source of raw material (Wilkins
18


et al. 2017). This suggests that the exposed PAP either did not expose other raw materials or continued to generate quality quartzite within exploitation distance from the site (Wilkins et al. 2017). Lithics from this level are smaller than in levels occurring in MIS 5, perhaps suggesting later reduction stages and/or more intense reduction during MIS 4 (Wilkins et al. 2017). These lithic findings, though preliminary, suggest the occupants at PP5-6 were attempting to maximize the raw materials sourced from the distant coast.
Shelly Ashy Dark Brown Sand (SADBS)
The SADBS is composed of 0.7 m of aeolian sand and consists of dense "trampled combustion microfacies” (Karkanas etal. 2015: 14). These hearths are not individually discernable, suggesting they were being made and replaced quickly by larger group occupations or more intensive occupations. Mollusk remains are frequent, however, they do not form a shell-supported midden as in the ALBS (Karkanas et al. 2015, Wilkins et al. 2017). The sediment of the SABDS is flat-lying, likely due to either intentional flattening by the occupants or caused by continued usage of the site (Karkanas etal. 2015).
During the SADBS, the coast averaged 15.1 km from the cave, ranging between a minimum distance of 2.5 km and a maximum of 23.8 km. Thus, the coast was (at times) outside the average foraging radius range for hunter-gatherers (i.e. 6-10 km, per Kelly 1996), which might have necessitated a lower reliance on coastal resources (Bar-Matthews et al. 2010, Fisher et al. 2010, Wilkins et al. 2017). This modeled coastal distance is interesting given the relatively high proportions of shell present in the SADBS. This may indicate the increased reliance in intertidal resources visible in the shell middens of the ALBS persisted in the SADBS, despite a greater distance to the coast.
19


Phytolith analysis of the SADBS indicated a decrease in C3 grasses and an increase in C4 plants, likely used to stoke the dense hearths occurring in this layer (Esteban et al. 2019). This finding suggests there was an increase in summer rains which indicate the environment was shrubby with low trees (Esteban et al. 2019). Esteban et al. (2019) additionally noted a spike in irregular and unidentifiable phytoliths, which may be explained by a change to more intensive dry wood collection undertaken to maintain the fires at PP5-6. The speleothem study findings were the same, indicating more C4 grasses and an increase in summer rains (Braun et al. 2018). This suggests the environment would have supported larger-bodied ungulates than the earlier levels, though the change to C4 grasses remains small.
Lithics analyzed from this level notably consist of backed microlithics and narrow blades (Wilkins et al. 2017). The SADBS is the first layer at PP5-6 to feature backed microlithic tools, and once they appear they persist for the remainder of the sequence (Wilkins et al. 2017). This indicates a technological shift occurred during the SADBS that was adopted and maintained during subsequent occupations. Silcrete exploitation peaks in the SADBS and is contemporaneous with increases in C4 grasses (Wilkins et al. 2015). One possible explanation for this relationship is that C4 grasses were a preferred combustion agent used to heat-treat lithic materials during the SADBS, thereby making silcrete a preferred material during this time (Wilkins et al. 2017). In addition, the increased coastal distance may indicate that new quartzite cobbles were not being replenished, or silcrete sources were exposed on the open areas of the PAP.
20


Table 1 provides summary data from transitional StratAggs during this undertaking. Broadly, the LBSR represents a sparse short-term occupational level, characterized by exploitation of both coastal and terrestrial resources. The environment was largely mosaic, with little to no input of C4 grasses. As the sequence continues to the glacial ALBS, occupation of the cave increases as does the apparent exploitation of coastal shell. Environmentally, there is a modest increase in C4 grasses, though the mosaic biome is largely intact. Finally, the SADBS represents dense, long-term occupation of the cave where hearths are so tightly packed as to be indistinguishable. During the SADBS, the presence of C4 grasses continues to rise. These grasses and dry wood were likely heavily exploited for fire and hearth fuel throughout the sequence. As the area descended into a glacial, the distance between PP5-6 and the coast changed during each level; these changes are illustrated in Figure 3.
Table 1. Coded specimens per StratAgg.
StratAgg Sediment Description Weighted mean OSL age Faunal NISP Percent Coded
LBSR (MIS 5 -Glacial) Single occupation hearths, shell-rich layers interspersed with roofspall-dense layers with little human input 81 ±4 8, 028 11.0
ALBS (MIS 4 -Interglacial) Sandy sediment with frequent human occupation levels consisting of ashy shell middens 72 ±3 4,314 37.0
SADBS (MIS 4 -Interglacial) Dense, indistinguishable hearths with abundant shell, though not enough to form middens 71 ±3 8,669 24.3
Total 13,860 17.4
21


co
CQ
Q
<
on
on
33
<
Coastal Distance [km] by StratAgg
71 ±3
72 ± 3
on
03
81 ±4
0
10
15
20
25
â–  Max Coastal Distance l Average Coastal Distance â–  Min Coastal Distance
Figure 3. Average, minimum, and maximum coastal distances for each StratAgg, as modeled by Fisher et al. 2010.
Toba Eruption and Glacial Winter
A recent study offers additional insight into the environmental circumstances occurring during the transition from MIS 5-4. Smith et al. (2018) discovered glass shards resulting from the Toba caldera eruption in the ALBS. The volcano, (located approximately 8,937 km away in Sumatra, Indonesia) erupted approximately 74 ka (Chesner et al. 1991, Buhring and Sarnthein 2000) - just before the onset of glacial MIS 4 (Rampino and Self 1992, Ambrose 1998, Smith et al. 2018). Some research suggests the eruption was both caused by the global transition to a glacial climate and helped push the global climate into MIS 4 by causing a volcanic winter (Rampino and Self 1992, Ambrose 1998, Rampino and Ambrose 2000, Ambrose 2003). Ice core evidence shows that the weather dropped 3 - 5° C
22


for several centuries following the eruption (Zielinski etal. 1996, Zielinski 2000). Given this theory, areas impacted by the eruption likely would have seen a contraction of populations to refuge areas (Ambrose 1998, Smith et al. 2018). The presence of the shards within the ALBS and immediately underlying the SADBS suggest the effects of the eruption may have caused human populations around Pinnacle Point to seek the refuge of PP5-6 (Smith et al. 2018). The proximity of the site to coastal and terrestrial resources, as well as the constant shelter would have provided an ideal location for affected populations to group together for survival during the onset of glacial MIS 4.
Anatomically modern humans evolved in Africa approximately 300 kya (Skoglund et al. 2017, Callaway 2017), then expanded to the Levant approximately 100 kya (Lewontin 1972, Relethford 1995, Ambrose 1998). Genetic studies have shown that prior to this major dispersal, one or more population bottlenecks occurred followed by pulses of population growth (Ambrose 1998, Fagundes et al. 2007, Behar et al. 2008, Hammer et al. 1998, Harpending et al. 1998, Hawkes et al. 2000). The approximate date of this bottlenecking is debated, with no concordance between researchers. Ambrose (2005) proposed that one such bottleneck was caused by a glacial winter triggered by the eruption of the Toba volcano in Indonesia. Due to the lack of agreement on the date of these bottlenecks or the demonstrable effects on African populations following the Toba eruption, it is difficult to determine whether Ambrose’s theory is correct (a complete summary is available in Gathorne-Hardy and Horcourt-Smith 2003).
Comparing fauna between MIS stages has not yet been undertaken at Pinnacle Point. This project offers the opportunity to test the reliance on and intensity of terrestrial fauna
23


exploitation during the transitional phase. This study lays the ground work for connections
between human behavior and environment at PP5-6 and further informs our understanding of early humans on the coast of South Africa.
24


CHAPTER III
METHODS
The data available from PP5-6 is tightly controlled with carefully planned excavation techniques, outlined below. Additionally, the StratAggs have been tightly dated and described in three publications from SACP4 (Karkanas et al. 2015, Wilkins et al. 2017, Esteban et al. 2018). This ensures temporal concordance of MIS dates at PP5-6 while also providing comparative metrics for future studies. PP5-6 is an ideal site at which to understand the human response to changing environments because the geological sequence is continuous throughout the MIS 5-4 transition, suggesting the faunal remains will adequately represent the faunal exploitation strategies of the occupants in response to a changing environment. Understanding how humans subsisted across the MIS 5-4 transition at PP5-6 is undertaken by sampling the fauna from the period in question, subjecting it to a thorough zooarchaeological analysis, and visualizing the data using 3D GIS.
Excavation Techniques at Pinnacle Point
Excavation occurs by defining stratigraphic units (StratUnits) that are delineated by natural sediment changes (Brown et al. 2012, Oestmo and Marean 2014). Unique lot numbers are assigned to localize a StratUnit within a 50 cm x 50 cm quadrant excavated to the natural depth of the StratUnit or to a depth of 5 cm if the StratUnit is thick and vertically continuous (Figure 2). The StratUnits are then grouped to a sub-aggregate (SubAgg) of nearby stratigraphic units with similar sedimentation, inclusions, and anthropogenic input (Brown etal. 2012, Oestmo and Marean date). StratAggs are
25


commonly known as "layers” or "levels” at other archaeological sites, however, StratAgg is preferred when discussing PP5-6 to retain consistency across publications. These StratAggs are used to discuss the depositional changes and time periods represented at the site and are named based on sediment characteristics (e.g. Light Brown Sandy Roofspall [LBSR], Ashy Light Brown Sand [ALBS], Shelly Ashy Dark Brown Sand [SADBS]). Upon closure of the site, 11 StratAggs were identified and named (further information is available in Karkanas et al. 2015 and Wilkins et al. 2017). All StratAggs are dated using OSL comprised of 65 single-grain samples throughout the sequence (Karkanas et al. 2015, Wilkins etal. 2017). These dates were found to match blind-tested Uranium-Thorium dates of the caves, confirming the accuracy of the dates (Jacobs 2010, Karkanas et al. 2015, Wilkins etal. 2017).
All artifacts (regardless of size or type) are point-located using a Total Station, assigned catalogue numbers during excavation, and input directly into a database using a barcode scanner (Marean 2015, personal communication). Each excavated artifact is plotted in 3D space, assigned a unique identification number, and associated with its stratigraphic level, SubAgg, and StratAgg at the site (Brown et al. 2012, Oestmo and Marean 2014). This allows each artifact to be associated not only with the level and sediment from which it was removed, but also allows the relocation of artifacts in space relative to the site and each other after excavation. This method is designed to mitigate human errors that regularly occur in archaeology such as measurement inaccuracies and interpretive differences among excavators (Bernatchez and Marean 2011, Dibble et al. 2007, Oestmo and Marean 2014). By using 3D GIS to locate artifacts in space, artifacts are easily
26


associated to dated sections (Brown et al. 2012, Karkanas et al. 2015, Wilkins et al. 2017). In addition, logging 3D coordinates for artifacts preserves the relational data between the artifacts (i.e. the spatial relationship between cutmarked bones and stone tools).
Figure 4. Excavation underway at PP5-6. SubAggs are identified on the wall using tags, with StratAggs marked with larger tags. Photo facing east.
Sampling MIS 5-4 Transitional Fauna
As discussed, the PP5-6 excavation has been closed following thorough excavation. The excavation of 14 meters of sediment (Karkanas et al. 2015) yielded approximately 62,000 faunal remains (SACP4 2017). For this study, fauna will only be analyzed from excavation units that fall within the transition period (~81-71 ka). In fitting with the scope of this project, the upper portions of the LBSR nearest the MIS 5-4 transitional period will
27


be sampled. Likewise, the entirety of the ALBS, and portions of the SADBS which fall closest to the transition will be sampled.
The artifacts from PP5-6 are stored and organized according to excavation units within each SubAgg. Samples of faunal material from each of the StratAggs were selected at random. This ensured no bias was introduced through the selection of more readily-identifiable elements or specimens with relatively higher surface visibility.
Zooarchaeological Analysis
Fauna was analyzed by a suite of attributes, outlined below. Each of these attributes was then quantified and compared across the three StratAggs to determine broad differences in faunal processing at PP5-6 during the MIS 5-4 transition. All analysis was completed using a standardized Access database, and identifications were verified by expert zooarchaeologists prior to inclusion in the dataset. The analysis took place over several seasons and by multiple trained analysts, under the supervision of Drs. Jamie Hodgkins (University of Colorado Denver) and Curtis Marean (Arizona State University).
Fragment Size
Fragment size was recorded for each specimen. The maximum length and width of each fragment was measured using digital calipers and recorded in millimeters. The minimum and maximum, upper and lower quartiles, and means for both the length and width were calculated and plotted to compare the range of fragment sizes between levels. These data provide information on the overall destruction of the bone, which may impact the appearance of identifiable attributes (e.g. anthropogenic or carnivore marks), and/or
28


skeletal element (i.e. very small fragments may possess no features identifiable to element). This metric is also used to compare the preservation between levels; smaller fragments indicate greater levels of post-depositional damage to the bones. Should fragments differ greatly in size between levels, this could skew the results of other analytical procedures such as surface modifications.
Weathering
Weathering stage was coded following Behrensmeyer’s system (1978), to determine how long the bones were exposed on the surface prior to burial (1978:161). Following an extensive study of taphonomic processes (Behrensmeyer et al. 1980), Behrensmeyer standardized the classification of weathering stages for bones to create a comparative system against which all zooarchaeological assemblages could be compared (Behrensmeyer 1978). The system is based on the highest level of weathering visible on the flat portions of bones (1978:152). The weathering stages exist on a continuum and are summarized below, as outlined by Behrensmeyer (1978:151):
Stage 0: No cracking or flaking on bone surface. Bone is still greasy, with skin and muscle ligaments attached.
Stage 1: Parallel cracking of surface begins. Skin and ligaments may or may not still be attached.
Stage 2: Cracking and flaking of surface begins, with flakes attached to shaft on one end. Only remnant skin and ligaments are present, if any.
Stage 3: Surface has a fibrous texture with the outer cortical layer worn away in
patches. Tissues are rarely present. Weathering does not penetrate beyond 1 -1.5 mm.
Stage 4: Surface is coarsely fibrous and the bone splinters when moved. Weathering extends to inner cavities.
Stage 5: Bone is extremely fragile and may be splintered in situ. Spongy inner bone is exposed. Original shape of the bone is difficult or impossible to determine.
29


This weathering rating system is widely used in the zooarchaeological literature and provides a standardized way of discussing the condition of archaeological assemblages. For this analysis, the proportions of each weathering type will be compared across levels to determine whether differences in preservation exist between levels. This is an important piece of data to record because it may explain differences between levels which are not due to human activity, but rather differential preservation between levels (i.e. higher levels of heavy weathering indicates poor preservation of bones, while lower weathering indicates better preservation).
Surface Visibility
Coding for surface visibility was undertaken and assigned a value from 0 - 100% in 10% increments. Surface visibility records the degree to which the cortical surface is observable. Bone surface can be affected by a multitude of factors, the most common of which at Pinnacle Point are calcium carbonate concretions and post-depositional damage that results in the removal of cortical surface. Calcium carbonate concretions at PP5-6 are likely due to the heavy inclusion of mollusk shells, which are composed primarily of calcium carbonate (Andrus 2011, Claassen 1998). Overtime, these shells can breakdown and leach into the sediments, where they can then build up on faunal remains. The utility of coding surface visibility is twofold: to account for differences in the appearance of surface modification for bones with less than 20% surface visibility, and to provide a metric for determining the surface condition of bones between StratAggs. Recording these metrics may provide explanations for differences between levels not due to anthropogenic or weathering input.
30


Taxonomic Identification
Though taxonomic differences may signal changes in available fauna, it is not possible to demonstrate these changes through the analysis of a single site; instead taxonomic representation and the presence of different portions of the animal body provides information about what the occupants of the cave were choosing to transport and/or process at the cave. Taxonomic identification was undertaken at the smallest taxonomic unit possible (e.g., Genus, Family, Order, etc.).
Body Size
Body size is determined by the thickness of cortical bone, the curvature of the bone, and the overall size of each element. Though determining the taxonomic identification of animals is impossible based on body size alone, it may offer insight into the size of fauna chosen for transport to the site by the occupants. It may also provide some insight into the type of fauna transported back to the site by the occupants. Because many bones in archaeological assemblages are not identifiable to species (Klein and Cruz-Uribe 1984), body size can give a rough approximation of the types of fauna being exploited. Body size classes were first developed by Brain (1981), as a means of quantifying the unidentifiable remains to a broad category of species. Body sizes 1-6, representing a range of species, were used to provide data on the unidentifiable fauna remains from PP5-6. Following the analysis undertaken by Thompson (2010) on the fauna at PP13B, examples of South African species (and corresponding size) that fit each size category are presented in Table 2. Every analyzed fragment was assigned a body size class, even if a determination of taxonomy was also made.
31


Table 2. Body si African fauna (T ze categories (Brain 1981) and example South lompson 2010).
Body Size Weight Range (lb) Example Fauna
Body Size 1 5-50 Grysbok/Steenbok
Body Size 2 50-250 Springbok
Body Size 3 250- 750 Wildebeest
Body Size 4 750-2000 Eland
Body Size 5 >2000 Hippopotamus
Body Size 6 >6000 Elephant
Burning
Maximum burning stage was recorded for each fragment as unburned (0%), partially burned (<100%), fully burned (100%), partially calcined (<100%), or fully calcined (100%). Burning stage was determined by the color of the bone (Cain 2005, Stiner et al. 1995). Degree of burning can indicate behavior. Stiner etal. (1995) determined that bones only become calcined if they contact the fire directly; this suggests partially- or fully-calcined faunal remains were used as fire fuel sometime after or as part of the discard process. Additionally, bones can become fully carbonized when buried below the hearth, suggesting these bones are not necessarily attributable to cooking or refuse activities (Stiner et al. 1995). The frequencies of burned to unburned bone were compared and analyzed for differences between layers to determine whether a change in cooking or refuse strategies occurred during the transition.
Transportation Strategies
Based on his work with the Nunamiut, Binford (1978) put forward a theory that the frequency of skeletal elements in an archaeological assemblage can inform archaeologists about the transport strategy employed by past populations. These transport strategies were modeled on the "scatterplot comparisons of the abundance of skeletal elements and
32


the measured utility of the skeletal elements” (Marean and Frey 1997). To quantify these strategies, Binford (1978) calculated the general utility index (GUI) for each skeletal element by summing the nutrient availability of meat, marrow, and grease. This analysis determined that bones fell into two broad categories: high and low utility elements. Low utility bones are defined as bones that have little or no available meat, marrow, or fat (Binford 1978, Metcalf and Jones 1988). The cranium, vertebral column, and distal limbs (e.g. metapodials, tarsals, carpals) account for the lowest utility bones on an animal skeleton. In contrast, high utility bones are those which offer substantial available nutrition and are composed of the scapula, ribs, upper limbs (e.g. humerus, radius, ulna, femur, tibia, fibula), and pelvis of the animal (Binford 1978, Metcalfe and Jones 1988).
In addition to the MUI, minimum number of elements (MNE) is calculated for each skeletal element. The MNE is then divided by the number of times that element appears in a skeleton; this results in the %MAU (normed minimal animal unit) (Binford 1978,1984).
The %MAU calculation is used to determine the actual frequency of each element transported back to the site. Binford then plotted these two calculations against each other (GUI, %MAU) and fitted best curves to the data (1978,1984). As this methodology progressed, alterations to these calculations were made to mitigate common biases in archaeological assemblages. Metcalfe and Jones (1988) suggested a more readily-calculated measure of bone utility, where the dry weight of the element is subtracted from the gross weight of the element. This calculation is then normed and results in the SFUI (standardized food utility index), a ranking of utility for each skeletal element (Metcalfe and Jones 1988). These calculations have displaced Binford’s original methods (1978) and
33


are now the standard indices used to plot against the %MAU. Studies on the differential survivability of bones based on density (Lam etal. 1998, Lyman 1985) and carnivore selection (Marean and Frey 1997, Marean and Assefa 1999, Faith et al. 2007) have since proliferated in zooarchaeology. These studies indicated that not all bones should be included in the determination of transport strategy; instead, only bones with high survivability and high nutrient yield should be considered (Faith and Gordon 2007). Standard methodology is now to assess the frequencies of the femur, tibia, metatarsal, humerus, radius, mandible, skull, and metacarpals (Faith and Gordon 2007). For this research, both the SFUI and high survival skeletal elements were used.
Five transportation strategies result from this research: gourmet, bulk, unbiased, unconstrained, and reverse (Binford 1978, Faith and Gordon 2007). The gourmet strategy consists of only those skeletal elements that have the highest utility and very few or no low utility elements (Binford 1978, Faith and Gordon 2007). Sites which exhibit a bulk transport strategy consist of all bones, except those with the lowest utility (Binford 1978, Faith and Gordon 2007). Unbiased sites are composed of bones which have been transported back to the site in direct correlation to their utility (i.e. the highest utility bones represent the highest proportion, while the lowest utility have the lowest proportion) (Binford 1978, Faith and Gordon 2007). In the unconstrained strategy, the entire animal is transported back to the site for consumption (i.e. no part of the animal is brought back in preference over another) (Faith and Gordon 2007). This suggests that the fauna was either small enough to be transported back in its entirety or the kill site was near enough to the cave that the journey back to PP5-6 would not have been a waste of energy resources
34


(Schoville and Otorola-Castillo 2014). In addition to these transport strategies, a reverse utility curve is noted (Binford 1976,1984, Metcalf and Jones 1988). In the earlier days of food utility curves, the reverse utility curve (where only the lowest utility bones are transported) was thought to indicate a scavenging resource procurement strategy (Binford 1985,1991; Blumenschine 1988, Mellars 1996). It was later established that these reverse curves are instead attributable to an analytical error in which long bone shaft fragments were excluded from MNE counts (Marean and Frey 1997). This skews the sample towards dense carpals, tarsals, and cranial fragments with low utility but high survivability, because identifiable epiphyseal ends of long bones are more susceptible to carnivore scavenging (Marean and Frey Frey 1997, Marean and Frey 1997) and post-depositional deterioration (Lam et al. 1998). Therefore, reverse utility curves are indicative of analytical or sampling biases, not transportation strategies. Establishing which strategy the occupants of PP5-6 employed during each stage of the transition will inform our understanding of changing resource procurement strategies.
Breakage Analysis
Breakage types of long-bone shaft fragments indicate whether the bone was broken while fresh (green) or dry. Epiphyseal ends of bones are excluded from this analysis because they are known to fragment differently than shafts (Villa and Mahieu 1991:34). Bones fragmented while still green indicate they were broken in order to access the still-available nutrients, while bones broken when dry indicate post-depositional damage unrelated to nutrient acquisition (Villa and Mahieu 1991). Villa and Mahieu (1991) were the first researchers to describe and identify the morphological differences between
35


nutritive and non-nutritive breakage by analyzing three known-accumulation assemblages of human remains. The results of this study concluded that the breakage created by nutrient acquisition versus post-depositional damage could be distinguished through analysis of the fracture angle and fracture outline of bones (Villa and Mahieu 1991). Fracture angle was previously known by researchers to display morphological differences between fresh and dry bones (Johnson 1983, Morlan 1984); however, the combination of angle and outline was unique to Villa and Mahieu (1991). Fracture angle is defined as "the angle formed by the fracture surface and the bone cortical surface” (Villa and Mahieu 1991:34). Angles are categorized as either right, oblique (defined as acute or obtuse), or right/oblique (e.g. irregular breakage). Fracture outline refers to the shape of the bone ends and are categorized as either transverse (straight) or v-shaped/curved, and intermediate (i.e. irregular breaks with two or more outline types). Villa and Mahieu (1991) further grouped the fracture angle and outline together to define nutritive versus non-nutritive breakage types.
Curved/v-shaped in outline and obliquely-angled breaks were determined to occur when a bone is broken in order to access meat, marrow and other nutrients (Villa and Mahieu 1991). In contrast, non-nutritive breaks (transverse in outline, right-angled breaks) occur when bones are dry and no longer offer nutrients, suggesting they were broken after their original deposition by factors like cave roof collapse, or sediment compaction by human/animal trampling (Villa and Mahieu 1991). Angles and outlines belonging to intermediate or indeterminate categories were excluded from this analysis because they do not definitively fit into either category and may be the result of multiple processes. For
36


each long bone shaft fragment recorded at PP5-6, the fracture angle and outline were recorded, then each end was identified as either nutritive or non-nutritive breakage.
These frequencies were then compared to data published by Marean et al. (2000), who generated experimental observations of human and carnivore breakage patterns to create expected frequencies of nutritive and non-nutritive breaks to which archaeological assemblages can be compared. By comparing the transitional levels to data generated experimentally, deviations from the norm can be investigated. A change in the amount of nutritive breakage across levels may indicate increasing or decreasing nutritional stress, while the difference from the expected non-nutritive breakage may indicate the processing intensity of the site broadly (as compared to extant populations). Non-nutritive breakage differences between levels will indicate whether more or less post-depositional damage was occurring, which may indicate changing site occupation strategies or preservation biases between levels.
Surface Modifications
Of primary interest to this research are the modifications made by humans at the site during the transition period. In order to understand the differences between layers, several anthropogenic attributes were coded and compared. Surface modifications (i.e. cut and percussion marks) were identified and photographed using incident lighting and a DinoLite digital microscope with up to 40x magnification (though typically a magnification of 20x or less was employed). Each mark was photographed, logged, and confirmed by zooarchaeologists with advanced experience to validate the marks (per Blumenschine et al. 1996).
37


Cutmarks
The identification of cutmarks created by humans and hominids has been crucial to zooarchaeology. Since 1981, descriptions of cutmark morphology have proliferated (Bunn 1981, Potts and Shipman 1981), due largely to the direct indication of anthropogenic input on bones. While early zooarchaeologists recognized the morphology of cutmarks, there was much debate on whether these marks could be reliably distinguished from sediment compaction (Behrensmeyer etal. 1989, Fisher 1995) or carnivore tooth marks (Blumenschine and Selvaggio 1988,1991, Blumenschine etal. 1996, Potts 1988, Potts and Shipman 1988). Even the equipment used to identify cutmarks was debated. Some researchers (Shipman 1981, Shipman and Rose 1983) suggested only scanning electron microscopes (SEM) could provide powerful enough magnification to identify cutmarks, while others (Blumenschine 1995, Blumenschine and Marean 1993, Blumenschine and Selvaggio 1988,1991) suggested a hand lens and incident lighting were sufficient. Further debate regarding the conspicuousness of the marks also surrounded cutmark identification. Some argued that only obvious marks (those identifiable with the naked eye) should be considered in the total count of marks recorded on bone fragments, while others advocated that all marks (regardless of clarity or location) should be considered. A study conducted by Blumenschine et al. (1996) tested these long-argued points by blind-testing known accumulator marks using low magnification and incident lighting. This experiment definitively showed that by following specific protocol for each mark type, surface modifications could be accurately identified with 95% confidence. Additionally, this study advocated for the inclusion of all marks, regardless of prominence on the bone surface.
38


Blumenschine et al. (1996) proved the accuracy of this method for identifying surface modifications and this is now the widely used technique in zooarchaeology.
Cutmarks are described as "deep [and] v-shaped [in] cross-section” linear marks occurring on the cortical surface of bones (Figure 5) (Blumenschine etal. 1996:496). Marks created by different types of stone tools have slightly different morphologies. Bifaces produce close parallel cuts that are relatively curved and overlapping, while lithic flakes produce long, straight lines. Scrapers (often unifacially flaked) produce marks that fit both of these descriptions, depending on which side is facing the cortical surface (Hodgkins and Riel-Salvatore 2008). For the purposes of this study, differentiation between tool type for each cutmark is not reported.
Figure 5. Examples of cut marks from PP5-6, showing the relatively straight morphology and clustering of marks.
39


Percussion Marks
Hominids have broken open long bones to access the bone marrow inside since they first began hunting (Outram 2001, Thompson et al. 2019). The fat-rich nutrients from marrow reside in the medullary cavity of long bones and are accessed by percussing the bones open with a cobble in order to process the marrow (Binford 1978, Blumenschine and Selvaggio 1988, Outram 2001). Percussing bones to access nutrients has been observed across the globe, from the Arctic to the desert of Africa and every place between (Binford 1978, O’Connell et al. 1988, O’Connell and Marshall 1989, Yellen 1991). Not only is bone marrow an excellent source of calorie-rich nutrition for hunter-gatherers, it serves a number of other purposes such as waterproofing (Binford 1978) and tanning skins (Levin and Potapov 1964). Percussion damage has been of interest to zooarchaeologists since the late 1980’s when they were first quantified by Johnson (1985) and Blumenschine and Selvaggio (1988) using actualistic research. Two types of percussion damage relevant to this research have been identified through these actualistic studies: microstriations and percussion notches (Blumenschine 1995, Blumenschine 1995). Percussion microstriations are described as "pits, grooves, or isolated patches of microstriations” (Blumenschine 1995: 29)], usually identified on the cortical surface opposite the striking surface. Percussion notches (per Capaldo and Blumenschine 1994) are u-shaped damage marks occurring on bone surfaces and are the result of hammerstone or anvil blows. Notches generally occur on the edges of bones, where the bone has split from the blow. Percussion marks have been reliably distinguished from other surface modifications (i.e. carnivore
40


r
Figure 6. Sample of percussion modification marks. Clockwise from top left: microstriations, conchoidal flake scar, flake with two medullary percussion notches, percussion notch on cortical surface. Examples from unpublished experimental data by author.
tooth marks, through experimental studies (Capaldo and Blumenschine 1994, Pickering 2002, Galan et al. 2009).
Analysis of Surface Modifications
Comparisons on the frequency of both cutmarks and percussion marks were then made between levels. In order to cull the sample to only those bones with adequate surface visibility, analyzed specimens with less than 20% observable surface were removed from the statistical analysis. Controlling for surface visibility ensured evenness of the sample
41


across all three StratAggs. A chi-square analysis on the percentage of bones with one or more cutmark or percussion mark was undertaken. Further, an ANCOVA test with a Poisson distribution was conducted to test whether a statistically significant difference exists between levels, controlling for fragment size (the observable surface area of bone analyzed) between levels. Calculation of the surface area for each fragment was undertaken by calculation of the geometric mean (V(l*w)) (per Hodgkins etal. 2016:6). The geometric mean serves as a proxy for surface area of flat objects. To test whether this method can be used on bone fragments from PP5-6, Bunn’s (1983) circumference system was employed during analysis. Shaft fragment circumference is coded as (1) <50% total circumference (flat), (2) >50% total circumference (somewhat curved), and (3) 100% circumference (curved). The majority (83.5% n=5,119) of all bones coded from PP5-6 are fragmentary and coded as (1) <50% of the total circumference. This indicates that fragments have lost most of their curvature and are primarily flat. Thus, calculation of the geometric mean operates as a proxy for surface area (Hodgkins et al. 2016). If a significant difference is observed by the ANCOVA analysis, a Tukee’s post-hoc test is employed to determine where the difference exists between levels within the transition.
3D GIS Visualization
Once analysis of the fauna at PP5-6 was completed, coded bones were visualized in relation to each other using ArcGIS. This allowed faunal remains to be visually grouped by coded characteristics. First, all coded bones across the StratAggs were mapped in relation to each other to visualize the sample of bones through the sequence. This was undertaken in order to illustrate the physical relationship of sampled remains and to identify sections
42


of the MIS 5-4 transition sequence that should be further explored in the future. Based on the results of the zooarchaeological analysis, four visual representations were made to examine the relationship of combustion features, degree of burning, taxon, and anthropogenically-modified remains. These visualizations were undertaken to better understand the layout of the rockshelter and to potentially identify areas for future investigation.
43


CHAPTER 4
RESULTS
Fragment Size
The maximum length and width of each fragment was measured for each StratAgg. Each level averages to the same approximate fragment size, in both length and width (Figure 7). The average length of fragments is approximately 20 mm and the average width is approximately 10 mm, suggesting remains across the transition are highly fragmented, thus potentially obscuring surface modification data and taxonomic identification. This suggests there was little change in the amount of post-depositional damage occurring between levels; rock fall events, human and animal trampling, increased population size, or increased sedentism did not impact the faunal remains disproportionately. The range for length and widths of fragments from each level are also reported. The SADBS shows the largest range in fragment length, relative to the other levels. This indicates there are a few relatively longer fragments in the SADBS, suggesting greater variation in preservation during this time though it does not result in changes in the mean. This range, combined with a similar average, suggests the SADBS may be affected by outliers.
44


SADBS Length
SADBS Width i—i i 1
ALBS Length

ALBS Width

LBSR Length

LBSR Width HI 1
0 25 50 75 100 125 150 175 200 225
Figure 7. Average length and width for each StratAgg (millimeters).
Weathering
Weathering for each fragment was recorded using Behrensmeyer’s criteria (Behrensmeyer 1978). Stage 0 accounts for the majority of all weathering at PP5-6 (Table 3). There is a small increase in the percentage of fauna exhibiting Stage 1 weathering in the ALBS, and this increase continues in the SADBS. The difference between these two stages is unlikely to obscure the bone surface or ability of analysts to diagnose surface modifications or identify remains. The remaining weathering stages account for less than five percent of the total assemblage, suggesting faunal remains between StratAggs were not disproportionately affected by weathering. These data indicate bones likely spent little time exposed on the site surface before deposition (Behrensmeyer 1978), probably due to the protection provided by the rockshelter roof of PP5-6.
45


Table 3. Observed weather stages by StratAgg.
Weathering Stage LBSR (n) ALBS (n) SADBS (n)
Stage 0 97.4% (759) 84.6% (1,351) 87.1% (1,833)
Stage 1 0.4% (3) 14.0% (223) 11.9% (251)
Stage 2 1.5% (12) 0.8% (13) 0.9% (18)
Stage 3 0.1% (1) 0.4% (7) 0.1% (18)
Stage 4 0.5% (4) 0.3% (4) 0.0% (0)
Surface Visibility
Figure 8 shows the proportion of surface visibility for each level. For reporting, the surface visibility was grouped into four categories (i.e. less than 20 percent, 20 - 50 percent, 50 - 90 percent, and 100 percent visible). Approximately 32.5% of LBSR remains have full (100 percent) visibility. Fauna from the ALBS has the most obscured surfaces, with only 25% of remains exhibiting 100 percent surface visibility. The SADBS accounts for the least surficially obscured remains: 50% displayed complete visibility. Surfaces were obscured primarily by calcium carbonate build up which creates a hard crust on the cortical surface of bones. The high rates of partially- or fully-obscured bones in the ALBS indicates a post-depositional process occurring in this level is affecting the surface visibility recorded during this study. This may be explained by the shell middens that comprise much of the ALBS; calcium carbonate is the primary ingredient in shells and when ground (as through trampling or natural processes) or burned (through cooking or heating activities), the calcium carbonate may be released into the surrounding soil and can subsequently be deposited onto faunal remains during post-deposition (Claassen 1998). Moderate surface visibility (e.g. visibility ranging between 20 - 90%) remained constant
46


through the transition and accounts for a combined total of ~25% of the remains from each level.
Surface Visibility
SADBS
ALBS
LBSR
n=531
n=193
n=1059 n=865
n=81
n=392
n=327
n=66
n=261
0.00% 10.00% 20.00% 30.00% 40.00% 50.00%
â–  >20% â–  20%50% 50%>x<80% 100%
Figure 8. Surface visibility by StratAgg.
Taxonomic Identification
Specimens were coded to the smallest taxonomic category possible for each fragment. In many cases no determination could be made, due to the lack of diagnostic characteristics and fragmentary nature of the remains. These bones were assigned to the broad category "terrestrial mammals.” In each StratAgg, these account for the largest proportion of coded remains (LBSR - 74.3%, ALBS - 75.2%, SADBS - 80.2%, Table 4). This is not unusual for South African faunal assemblages of this age (Thompson 2010) and is likely attributable to poor preservation. Remains belonging to Family Bovidae (e.g.
47


wildebeest, gazelles, impala) and Family Testudinidae (i.e. tortoises) account for the highest proportion (53.6% of the total assemblage; n=l,057) of remains identifiable beyond class (Figure 9). An exhaustive list of identified fauna is available in (Table 4), however, because they do not represent large proportions of the sample, they are not discussed further.
Despite the changing climate and environment, the difference in identifiable taxa is negligible. While the proportions of Family Bovidae change between levels, this is likely attributable to identifiability of the remains. The proportion of tortoises (Family Testudinidae) decreases in the ALBS (15.2%) but remains similar between the LBSR (23.6%) and the SADBS (21.9%). This change might be related to either a change in resource procurement strategies or the availability of tortoises at the initial onset of glacial MIS 4.
48


SADBS
n=196
n=80
0 20 40 60 80 100
â–  Family Bovidae Family T estudinidae
Figure 9. Proportion of identifiable taxa for each StratAgg. Taxa representing less than one percent of the assemblage are not reported in this figure.
Body Size
Figure 10 summarizes the body size class of the remains recorded for each level. The LBSR is dominated by size 2 fauna (38.4%), while size 1 (16.7%) and size 3 (16.6%) comprise the remainder of the determined fauna. Approximately 27.5% of the LSBR assemblage is comprised of indeterminate size class fauna. The ALBS is similarly dominated by size 2 fauna (40.6%), with size 1 (17.8%) and 3 (4.9%) also represented. Indeterminate size fragments account for 35.9% of the ALBS assemblage. The SADBS exhibits a slightly different patterning: fauna of indeterminate size composes the highest proportion of bones in this level (44.6%), with size 2 representing 39.2%, size 1 accounting for 11.4%, and size class 3 at 3.5%. Little difference is observed between levels for size classes 1 and 2,
49


suggesting the availability of these small mammals was not greatly disrupted by either the
coastal retreat or the increasingly mosaic environment. Body size 3, however, shows an unexpected and marked decrease in proportion as the climate shifts into a moderate glacial with an increase in C4 grasses.
BodySize
â–  BodyBize Indeterminate
BodyBizeEL
(51003b)
â–  BodySizeS (50 US 503b]
BodySizeS
(25O@05Olb)
0% 10% 20% 30% 40% 50%
Figure 10. Relative frequencies of faunal body sizes through the transition.
50


Table 4. Comparison between levels of the NISP coded for each Family, Order, Class by Body Size. Each specimen was coded to the smallest unit of taxonomic unit.
Body Size LBSR ALBS SADBS MIS 5 MIS 4 MIS 4
Indet. 0 Terrestrial Mammal ^ 2 3 116 236 633 0 6 47 54 135 118 224 428 415 92 81 45
2 Marine Mammal 0 0 2 10 1
Indet. 1 Family Bovidae 3 4 14 6 8 6 31 18 50 152 276 30 17 17 0 0 1
Indet. Family Testudinidae o (Tortoises) i 2 8 14 26 0 2 4 52 52 50 110
Order Carnivora Indet. 1 0 10 13 2
Family Procaviidae (Hyrax) 2 0 3 1
Family Spheniscidae (Penguin) 2 0 10
Family Equidae (Horse/Donkey/Zebra) 4 10 0
Family Rhinocerotidae (Rhino) 5 0 10
Family Otariidae/Phocidae (Seal) 5 0 10
Order Primates (Non-human Primates) 1 0 12
1 Class Aves (Birds) 0 3 0 0 0 2
NISP Total — 654 1177 1567
51


Burning
Burned bones were analyzed according to the maximum burning stage exhibited on the bone (i.e. >50% calcined is coded as calcined). The level of burning and proportion of burned to unburned bones are reported in Figure 11. There is a statistically significant difference in frequency of burned bones across all three StratAggs at the site (x2=64.97, df =2, p<0.001, v=0.145 Figure 11). The LBSR contained 56.1% burned bone, the ALBS preserved 30.3%burned bone, and the SADBS exhibited 56.3%. Although the strength of the difference between levels is a modest, given the age of the deposits the association is important. Because the significant change between StratAggs is related to the proportion of burned to unburned bones, not to the proportions of each level of burning, an additional examination on the proportions of each level of burning was also undertaken. This showed that the proportions of each level of burning (i.e. calcined, carbonized, etc.) remain unchanged between StratAggs, which suggests that the manner in which bones are being cooked or discarded was the same or similar for each level. Instead, only the amount of total burning is reduced during the ALBS.
52


Maximum Burning Stage
SADBS
ALBS
LBSR
â–¡ Fully Calcined
HU Partially Calcined (<100%)
â–  Fully Carbonized
â–  Partially Carbonized (<100%)
â–  Total Burned
0.0% 20.0% 40.0% 60.0% 80.0%
Figure 11. Percentage burned material per StratAgg, including percent fully calcined, partially calcined, fully carbonized, and partially carbonized.
Nutritive and Non-nutritive Breakage
Percentages of nutritive and non-nutritive breakage patterns were plotted for each StratAgg and compared to experimental data from Marean et al. 2000 ( Figure 12-Figure 13).
Non-nutritive breakage for each level is reported in Figure 12. All three levels exhibit higher percentages of non-nutritive breakage than the experimental data sets, suggesting the bones underwent a high level of taphonomic damage. Approximately 20% of long-bone shaft breaks are non-nutritive in the LBSR. The ALBS features approximately 23% non-nutritive breakage. The SADBS has the highest proportion of dry bone breaks: approximately 33%.
53


Nutritive breakage for each layer is reported in Figure 13. All three StratAggs display fewer nutritive breaks than the experimental data. The LBSR has the highest proportion of nutritive breaks (approximately 79%). The ALBS has nearly the same: approximately 72% of breaks were nutritive. The SADBS has the lowest percentage of nutritive breakage at roughly 55%. Numbers lower than the expected (Marean et al. 2010) could indicate differing processing techniques or obscuring of nutritive breakage by the high levels of non-nutritive breakage.
54


Non-Nutritive Breaks
40%
35%
30%
to
8 25%
CQ
§ 20%
£
I 15%
â–² SADBS
ALBS
LBSR
sp
0s*-
10% Hominid to Carnivore
5%
0%
| - Hominid Only
l
I1 Carnivore Only
0%
10% 20% 30%
% Right Angle Breaks
40%
Figure 12. Non-nutritive breakage for each StratAgg, compared to experimental data from Marean et al. 2000.
55


Nutritive Breaks
1/1
nj
cu
i-
CO
-a
cu
Q.
re
~o
0)
>
s_
3
u
sP
oS
100%
90%
80%
70%
60%
50%
40%
30%
30%
Carnivore O
Hominid On *T
♦ ♦ ALBS LBSR

SADBS


Hominid to Carnivore
40% 50% 60% 70% 80% 90% 100%
% Oblique Angle Breaks
Figure 13. Nutritive breakage for each StratAgg, compared to experimental data from Marean etal. 2000.


Food Transport
Food Transport Utility curves were calculated for each StratAgg based on the %MAU and SFUI (Faith and Gordon 2007), these figures are reported in Table 5. Distance-weighted mean squares curves were fitted to each StratAgg and plotted against Binford’s Food Transport Utility curves in Figure 14 (Binford 1978, Faith and Gordon 2007). The results show the LBSR followed a bulk strategy, where all but the lowest utility bones are transported back to the site (Binford 1978, Faith and Gordon 2007). The ALBS tracks most closely to the unbiased (following it nearly exactly), suggesting occupants of PP5-6 were transporting skeletal remains in direct proportion to their utility. The SADBS closely aligns with the bulk strategy. The food utility lines for both the LBSR and the SADBS are brought down by lower than expected proportions of femora. This may be due to difficulties in assigning skeletal element because of the fragmentation of bones and extensive post-depositional damage, though it may also be related to butchering strategy. The low levels of femora present at PP5-6 may be explained by ethnographic study conducted by Binford (1978:54) who noted that Nunamiut hunters’ broke femora to access nutrients during the butchering process and thus never transported these bones back to site. A similar patterning of missing femora has been noted in other archaeological assemblages, suggesting this maybe be a consistent explanation applicable to numerous contexts (Costamagno et al. 2006, Hodgkins 2012).
57


Table 5. Minimum number of elements, MAU, SFUI and normed MAU values for each StratAgg, used to create Figure 2. Created using SFUI, as calculated by Metcalfe and Jones f!981J.__________________________________________________________________________________
Element MNEL MNER Side N/A MAU SFUI %MAU
Cranial 2 1 9.1 27.9
Mandible 2 1 11.5 27.9
C/5 CO Q Humerus 1 1 1 36.8 27.9
Radius 1 5 3 25.8 85.7
C/5 Metacarpal 2 2 2 5.2 57.1
Femur 1 0.5 100 14.3
Tibia 5 2 3.5 62.8 100
Cranial 3 1.5 9.1 11.1
Mandible 3 1.5 11.5 11.1
Humerus 1 8 4.5 36.8 100
C/5 CO Radius 4 3 3.5 25.8 77.8
nO < Metacarpal 4 2 5.2 44.4
Femur 2 1 100 22.2
Tibia 3 5 4 62.8 88.9
Metatarsal 1 3 2 37 44.4
Cranial N/A N/A 9 4.5 9.1 100
Mandible 1 0.5 11.5 11.1
QZ Humerus 2 1 36.8 22.2
pd hJ Femur 1 0.5 100 11.1
Tibia 2 1 1.5 62.8 33.3
Metatarsal 1 0.5 37 11.1
58


♦ LfiSR All - ALBS ALL
• SADBS ALL
----LBSR
----ALBS
----SADBS
7igure 14. Distance-weighted mean squares curve fitted to each StratAgg and plotted against Binford’s Food Transport Utility curves (Binford 1978, Faith and Gordon 2007).
59


Surface Modification
Cut and percussion mark frequency was recorded for all levels. Only faunal remains exhibiting surface visibility greater than 20% were included in this analysis. The frequency of cutmarked bones through the transition decreases significantly but modestly (x2=21.51, df=2, p<0.0001, v=0.106) from the LBSR (12.1%) to the ALBS (8.7%) to the SADBS (7.3%). Percussion marks do not exhibit a significant difference between levels. Approximately 5.1% of the LBSR assemblage exhibits percussion damage. Roughly 3.8% of the ALBS has percussion damage, and the SADBS shows approximately 3.2%. An analysis of the number of cutmarks (Figure 15) and percussion marks (Figure 16) per fragment (controlling for surface area, using the geometric mean) was also undertaken to determine whether bones were being more intensely processed across StratAggs.
The analysis of cutmarks per fragment shows a statistically significant difference between levels (Wald x2=6.020, p=0.049, Figure 15). The p-value for this test (p=0.049) is close to the boundary of significance; however, this may indicate a larger pattern of behavioral change. A Tukey’s post-hoc test was performed to determine where the significant difference is located in the transition. This showed the change in cutmarks per fragment is significant (p=0.016, df=l) between the ALBS and the SADBS, with cutmarks per fragment increasing 0.06 marks per fragment. The analysis on the number of percussion marks per fragment was not significant across levels (Wald x2=2.115, p=0.347, Figure 16).
60


Wald x2 = ©.020,^=0.049
^ g
s=
o
s
CJD
re
â– i-t
a
o
CJD
re
o
S- fe
a e
t: "a re «
[§ o
-4-» 5-<
5 § u
LBSR ALBS SADBS
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
Figure 15. The results of the ANC0VA with Poisson distribution to determine the least-squares mean number of cut marks per fragment, controlled for surface area (calculated using the geometric mean). The error bars show the 95% confidence intervals for each StratAgg.
61


Wald x2 =l2.115,lj)=0.347
0.30 0.25 0.20 0.15 0.10 0.05 0.00
LBSR ALBS SADBS
Figure 16. The results of the ANCOVA with Poisson distribution to determine the least-squares mean number of percussion marks per fragment, controlled for surface area (calculated using the geometric mean). The error bars show the 95% confidence intervals for each StratAgg.
3DGIS
The spatial relationship between artifacts was mapped using ArcScene 10.3.1 software. First, a 3D visualization of the three different StratAggs was undertaken, and points that were wrongly input into the Total Station were removed. These were identified by their location outside excavated areas and verified against the zooarchaeological data. This visualization showed that while the faunal remains analyzed account for a large proportion of the continuous transition, some gaps remain in the transitional data. While a complete analysis of the PP5-6 fauna is outside the scope of this project, these data suggest
E3
O)
O)
s .a
i®
2 c M £ £
i bx> S< cb
& ^ u.
CB tS

â– a c
o s
(/5
(/5
3
u
s-
o
CLn
o
!-
a
o
u
62


these findings may be strengthened with further analysis of missing portions of the sequence, as illustrated by the GIS data.
The spatial relationship between the largest proportions of taxonomic identifications was undertaken (Figure 17). For ease of visualization, only the two highest proportion of taxon were mapped: Family Bovidae and Family Testitudidae. This showed a normal distribution of Family Bovidae and Family Testitudidae, with identifiable remains clustering closely together. The LBSR shows a concentration of tortoise remains in the lower portion of the StratAgg, suggesting a relatively higher reliance on tortoise during MIS 5, which gives way to a more varied sample of both tortoises and Bovidae as the LBSR gives way to MIS 4. The GIS data also show that the tortoise remains from the ALBS are, in general, situated more closely to the LBSR while the later portion of the ALBS is dominated by Bovidae. This suggests that the lower proportion of tortoise remains in the ALBS is a function of the changing environment, which is later mitigated as MIS 4 wears on in the SADBS, where tortoise and Bovidae remains are equally distributed.
To examine the relationship of faunal remains to hearth features, the assemblage was then mapped by bones recorded from a combustion feature versus those coded from all other contexts (Figure 18). This shows the concentration of hearth features (and bones either discarded in the fire or used as fuel) for each level. To see if the degree of burning would concord with the combustion feature data, the levels were mapped according to the highest observed level of burning (Figure 20). Despite the location of hearths within the transition, burned bones are dispersed throughout the sequence. There is some aggregation of burned bone at the locations of the hearths; however, the location of burned
63


bones is not confined to them. This suggests the bones were being discarded throughout the cave after consumption and do not appear to be either systematically used for fuel or discarded in known refuse piles. With regard to the lower proportion of burned versus unburned bone in the ALBS, the GIS data does not offer valuable explication. There is a concentration of unburned bone in the ALBS that appears to account of a large number of the coded specimens. This cluster is located outside of known hearth features and very near the onset of MIS 4, suggesting the lack of preparation with fire might be the result of climactic change. The exact date of this grouping of unburned bones is not known, but it may represent an effect of the Toba Eruption (e.g. animals reacting to effects of the eruption seek shelter in the rockshelter). Another possible explanation is that the bones were coded as unburned due to a coding error; whole lots were often undertaken by the same analyst therefore if an error was not corrected quickly it may have not been noted. This analytical error is unlikely as all coding was confirmed and overseen by a lead zooarchaeologist.
Finally, the location of bone marked by cut or percussion marks were visualized across the transition. This shows an even distribution across all three StratAggs; the anthropogenically modified bones do not appear especially aggregated between the levels. Within each level, however, the bones are situated near or within the combustion features illustrated in Figure 18. This supports the use of the cave by humans as a food processing and preparation site. This spatial relationship is an obvious one; bones percussed for marrow or cut to access meat were subsequently heated for consumption.
64


Figure 17. Analyzed fauna from the transition, identified by taxon.


Scale: 4.78 m
Figure 18. Visualization of faunal remains located within combustion features versus those located in other contexts.
66


Scale: 4.78 m
SADBS (MIS 5)
ALBS (MIS 5)
LBSR (MIS 4)
Maximum Burning Stage
Fully Calcined Fully Carbonized Partially Calcined Partially Carbonized Unbu med
Figure 19. Burned bones, displayed by level of burning across the transition.


Scale: 4.78 m
Figure 20. Anthropogenically modified bones shown in association with bones recovered from hearth contexts (in black).


CHAPTER V
DISCUSSION AND CONCLUSION
The results from the zooarchaeological analysis are discussed within each StratAgg, then broad trends or changes are identified and discussed.
LBSR Discussion
The LBSR, a relatively static layer of moderate anthropogenic input dating to MIS 5, serves as the layer against which the MIS 4 levels can be compared to determine changes in behavior related to climate. During this level, the coast was very near the shore (Fisher et al. 2010) with the environment typified by a fynbos-renosterveld mosaic (Braun et al.
2018, Esteban et al. 2019). Though the intertidal zone was close and easily accessible, the LBSR sediments display only moderate exploitation of coastal resources such as mollusks (Karkanas et al. 2015, Wilkins et al. 2017). The faunal analysis suggests that during the LBSR, occupants were exploiting small-bodied mammals, primarily under 250 kg. According to the fauna identifiable to taxon and body size, these animals are comprised primarily of Bovidae (grysbok/steenbok/springbok) and Testudinidae (tortoise). A majority of fragments from this level were not assignable to taxon (66.8% of the total LBSR assemblage) or body size classes (27.5% indeterminate), due to fragmentation and lack of identifiable features. Fragment size and non-nutritive breakage rates suggest preservation in the LBSR is poor, though the weathering rates show bones likely were not exposed on the surface long before burial. Burning data indicates approximately 56.1% of the bones from the LBSR were heated; however, only 9.3% of the bones are calcined to some degree. These calcined remains were in direct contact with the hearth (Stiner et al. 1995), possibly
69


indicating their use as fuel. The remaining heated fragments (46.8%) are likely the result of cooking or processing of nutrients. Unburned bones do not necessarily indicate a lack of processing, as burning signatures on bone can be obscured by tissue. During the LBSR, humans at PP5-6 were employing a bulk transport strategy, suggesting PP5-6 occupants during this time were choosing to transport whole or nearly whole carcasses back to the site. Schoville and Otorola-Castillo (2014) have demonstrated that transport strategies calculated with assemblages comprised primarily of body sizes 1 and 2 are not reliable indicators of distance to kill sites, as proposed and used by Binford. Instead, this transport strategy is a better indication of the decision by PP5-6 occupants to transport whole carcasses back to the site for processing. The number of nutritional breaks and percentage of cutmarked bones (12.1%) further indicate fauna were being processed more heavily than in later levels.
ALBS Discussion
The ALBS is dated to the onset of MIS 4 and contains relatively frequent hearths interspersed with dense shell middens (Karkanas et al. 2017). During the ALBS, the coast was located a moderate distance away (Fisher et al. 2015), and the environment shows an increase in C4 grasses though the impact of this increase has not been well demonstrated (Braun et al. 2018, Esteban et al. 2019). It is also during the ALBS that the Toba eruption occurs in Sumatra (Chesner et al. 1991, Buhring and Sarnthein 2000), potentially plunging the area into more extreme glacial conditions (Ambrose 1998, Ambrose 2000, Ambrose 2000, Rampino and Self 1992).
70


Though fragment size remains constant in the ALBS (from the LBSR), non-nutritive breakage, weathering, and surface obstruction increase, possibly signaling poorer preservation than in the LBSR. Surface visibility may be affected by leeching of calcium carbonate from shells into the sediment (Claassen 1990). Faunal analysis shows nearly the same amount of remains identifiable to taxon as in the LBSR; however, proportions of Testudinidae decrease to 15.2%. This might indicate either an availability disruption or procurement difficulty during the ALBS. None of the Testudinidae specimens were identifiable to species; however, faunal analysis conducted at nearby sites suggests the sample would include the angulate tortoise (Chersina angulata), and the pancake tortoise [Homopus arelatus) (Henshilwood et al. 2001, Halkett et al. 2003, Klein and Cruz-Uribe 2000, Klein et al. 2004, Thompson 2010). These species of tortoise are common to modern fynbos and thrive in temperate, coastal, semi-arid environments (Branch 1984, Thompson 2010). This suggests the change in Testudinidae during the ALBS is likely attributable to a decrease in tortoise availability during the initial cooling of MIS 4. Additionally, the ALBS shows an increase in cutmarks per fragment (0.31 per fragment) and nutritive breakages from the LBSR, both of which indicate increasing processing intensity (Hodgkins et al. 2016). This seems to track well with the change to an unbiased food transport strategy, where occupants transport food back in direct relation to its nutritive value (Binford 1978, Faith and Gordon 2007). This level has the lowest proportion of total burned bone across the transition (30.3%), which may be partially explained by the decreased surface visibility. However, the rate of decrease in burned bone cannot be explained entirely by obscured surfaces but may instead indicate nutritional stress. The lower proportion of burned bone
71


could illustrate less time spent cooking or processing the meat before consumption; decreasing the cooking time of meat to the shortest time is a possible tactic for minimizing the time between hunting and nutrient acquisition. Another explanation may be related to the relationship between AMH and fire- European assemblages show that evidence of fire during glacial periods is low (Sandgathe et al. 2011a, Sandgathe et al. 2011b). Sandgathe et al. (2011a) propose that this is due to an opportunistic (rather than controlled) use of fire. The lower proportion of burned bone seems to roughly accord with these data, possibly suggesting a more limited availability of natural fire during the initial onset of MIS 4. In general, the ALBS might be marked by increasing nutritional stress coupled with decreased preservation of faunal remains.
SADBS Discussion
The SADBS is squarely situated in MIS 4, when the coast has retreated to an average distance of 15.1 km (Fisher et al. 2010). The occupation of the cave is denser, evidence by thick hearths and increased sediment trampling (Karkanas et al. 2017). As with the ALBS, the environment shows a continuing increase in C4 plants and shrubby vegetation (Braun et al. 2018, Esteban et al. 2019). Though fragment size, weathering, and surface visibility in the SADBS closely tracks with that of the LBSR, non-nutritive breakage is at the highest level throughout the transition. This is likely due to the increased trampling of sediments caused by longer occupation times or larger group sizes (Karkanas etal. 2017). In many ways, the SADBS is very similar to the LBSR; taxa, body size, cut marks per fragment, proportion of burned bones, and transport strategy are much the same. However, the proportion of nutritive breakage in the SADBS is the highest across the transition. A
72


possible explanation for this increase could be the proposed larger population sizes of the SADBS; more occupants would require all possible nutrients to be extracted from fauna.
The decrease in cut marks per fragment from the ALBS to the SADBS, however, seems to indicate an increase in nutrient acquisition only as it relates to shaft breakage. Three explanations may account for the decrease in cutmarks per fragment. The advent of microlithic technology occurring the SADBS (Wilkins et al. 2017) may have affected the way meat was removed from bones such that cut marks on bone are not as frequently left behind as in earlier levels. Alternatively, a socially- or culturally-determined change in butchering technique could account for the decrease in marks. Finally, there may have been an increased reliance on marrow extraction but not on flesh; nutritive breakage may have increased to access the fatty, nutrient-dense marrow while no change in defleshing occurred. It is difficult to determine which of these explanations best fits the scenario at PP5-6, though it is clear a change in behavior did occur during the SADBS.
Conclusions
The results of the zooarchaeological analysis show that there was likely some behavioral mitigation occurring during the transition from interglacial to glacial climate at PP5-6. The data show several trends relating to subsistence strategies during the change from interglacial to glacial climate - the key differences are summarized in Table 6 and visualized in Figure 21. There is an overall trending decrease throughout the transition to a lower proportion of body size 3 fauna, suggesting a change in reliance on larger fauna by the occupants. This is counter to the changing environment that would have produced more C4 grasses; a change which, coupled with a more exposed Agulhas plain, should have
73


supported larger-bodied mammals. Though teeth were not included in this analysis, they have been recovered and identified as large-bodied ungulates through the transition, suggesting the occupants of PP5-6 were exploiting body sizes 3-6. This is likely explained, in part, by the high level of fragmentation caused by increased trampling and post-depositional damage of faunal remains through the transition, which obscure the identification of larger body sizes. Additionally, there is a steady decrease in both nutritive breakage and percentage of cut marked bones, suggesting less nutritional stress as the transition carried on. Backed microlith tools, with a superior ability to hunt large game may explain the decrease in cut marked bone, either through an increased success rate of hunters or by changing the way butchering is undertaken (i.e. new strategies could result in fewer cuts on bone surfaces). Though the overall proportion of cut marked bones decreases steadily through time, the ALBS shows a significant increase in number of cutmarks per fragment (controlled for fragment size), relative to both the LBSR and SADBS. This indicates fragments were being more intensively processed during the ALBS, relative to both the LBSR and SADBS.
As Pinnacle Point became more mosaic, the PP5-6 occupants may have experienced an initial increase in nutritional stress that was mitigated during the ALBS. The occupants appear to have returned to the same strategies used during MIS 5 during the SADBS, either by changing subsistence strategies or a return to steadier climatic conditions. Increased nutritional stress is also visible in the total percentage of burned bone, which experiences a statistically significant decrease in the ALBS. Cooking time may have been abbreviated
74


during times of increased nutritional stress, or AMH’s access to fire during the initial onset of the glacial period may have altered the patterning of burning during the ALBS.
Table 6. Summary table of the key differences between StratAggs during the MIS 5-4 transition.
LBSR ALBS SADBS
MIS 5 (81 ± 4) MIS 5 (71 ± 3) MIS 5 (72 ± 3)
Environment Primarily C3 vegetation with bimodal rain Increase in C4 vegetation and summer rains Highest proportion of C4 grasses and summer rains
Shell exploitation Thin lenses of shell exploitation Dense, shell-supported matrices Abundant shell but less than in ALBS
Average Distance to Coast (Fisher et al. 2010] 1.1km 10.5 km 15.1 km
Body Size 3 16% 5% 3%
100% Surface Visibility 35% 25% 50%
Non-nutritive Breakage 22% 24% 32%
Nutritive Breakage 78% 74% 55%
Cutmarked bone 12.10% 0.1 7.30%
Cutmarks per fragment 0.26 0.32 (Statistically significant difference] 0.27
Heated Bone 56.1% 30.0% (Statistically significant drop in overall proportion of burned bones] 56.3%
Transport Strategy Bulk Unbiased Bulk
This analysis suggests humans were transporting food back to PP5-6 differently across the transition. Similar transport strategies were being undertaken in both the LBSR and the SABDS, while during the ALBS occupants are exhibiting preferential transport of higher utility elements. This may suggest an energy-saving measure intended to reduce
75


expended energy on the lowest utility elements, possibly to mitigate changing fauna availability or increased nutritional stress (Schoville and Otorola-Castillo 2014). Based on the Food Utility Curves, the data show there may have been at least some change in the ability of the occupants to access nutrients, causing modified behavior at the onset of MIS 4.
MIS 5-4 Transition Summary
100
90
0
LBSR
18
16
ALBS
0
SADBS
-•-100% Surface Visibility -•-Non-nutritive Breakage
-•-Nutritive Breakage —•-Burning
-•-Percentage of Body Size 3 -•-Cutmarked bone
Figure 21. Summary figure of key differences between StratAggs. Dashed lines indicate usage of the secondary axis.
The most tumultuous time of change during the transition occurs in the ALBS, where the environment was likely the most unpredictable. This accords well with the proposed glacial winter spurned by the Toba Caldera eruption (Ambrose 1998, 2000). The
subsistence behaviors and environmental indicators revealed during the ALBS suggest the glacial winter hypothesis should be further investigated. Taken in sum, the results of this study indicate there was a temporary change in subsistence strategies occurring at the
76


onset of glacial MIS 4, followed by a return to the behaviors of MIS 5 as the climate settled into the glacial period. The implications of these findings are, as yet, preliminary but suggest the population continuously occupying PP5-6 was sufficiently behaviorally elastic as to withstand environmental changes related to resource availability. There is no evidence for gaps in the occupation at PP5-6 (Karkanas et al. 2015, Wilkins et al. 2017), suggesting one group or related groups were occupying the cave, rather than representing unlinked bands of AMH. Further work is required at this site before meaningful conclusions can be drawn on the multitude of ways that humans at PP5-6 modified their behavior. Despite distant coastlines, changing flora, and climate, the population at this site was not diminished.
These results indicate humans on the southern coast of South Africa had the cognitive ability to adapt their resource procurement (e.g. transport strategies), cooking (e.g. burning), and nutrient acquisition (e.g. increasing nutritive breakage and cut marks) to an environment of shifting resource availability. This suggests this group of early anatomically modern humans possess all of the necessary cognition and behavioral elasticity to survive varied climates and niches. While this study alone cannot prove AMH possessed superior adaptability, it can serve as a useful case study of human resilience in the face of rapidly changing climate. The adaptability of AMH, as demonstrated by this thesis, can be compared to other sites, time periods, and hominid occupations to further illuminate our understanding of modern human’s expansion across the globe.
77


REFERENCES
Ambrose, S.H., 1998. Late Pleistocene human population bottlenecks, volcanic winter, and differentiation of modern humans. Journal of human evolution, 34(6), pp.623-651.
Ambrose, S.H., 2000. Volcanic winter in the Garden of Eden: the Toba supereruption and the late Pleistocene human population crash. Volcanic hazards and disasters in human antiquity, 345, p.71.
Ambrose, S.H., 2003. Did the super-eruption of Toba cause a human population bottleneck? Reply to Gathorne-Hardy and Har court-Smith. Journal of Human Evolution, 45(3), pp.231-237.
Andrus, C.F.T., 2011. Shell midden sclerochronology. Quaternary Science Reviews, 30(21-22), pp.2892-2905.
Bar-Matthews, M., Marean, C.W., Jacobs, Z., Karkanas, P., Fisher, E.C., Herries, A.I., Brown, K., Williams, H.M., Bernatchez, J., Ayalon, A. and Nilssen, P.J., 2010. A high resolution and continuous isotopic speleothem record of paleoclimate and paleoenvironment from 90 to 53 ka from Pinnacle Point on the south coast of South Africa. Quaternary Science Reviews, 29(17), pp.2131-2145.
Bernatchez, J. and Marean, C.W., 2011. Total station archaeology and the use of digital photography. SAA Archaeological Record, 11(3), pp.16-21.
Behar, D.M., Villems, R., Soodyall, H., Blue-Smith, J., Pereira, L., Metspalu, E., Scozzari, R., Makkan, H., Tzur, S., Comas, D. and Bertranpetit, J., 2008. The dawn of human matrilineal diversity. The American Journal of Human Genetics, 32(5), pp.1130-1140.
Behrensmeyer, A.K., 1978. Taphonomic and ecologic information from bone weathering. Paleobiology, 4(2), pp.150-162.
Behrensmeyer, A.K., Dechant Boaz, D.E. and Hill, A.P., 1980. The recent bones of Amboseli Park, Kenya, in relation to East African paleoecology. Fossils in the making: Vertebrate taphonomy and paleoecology, pp.72-92.
Bergh, N.G., Verboom, G.A., Rouget, M., Cowling, R.M., 2014. Vegetation types of the Greater Cape Floristic Region. In: Allsop, N., Colville, J.F., Verboom, G.A. (Eds.), Fynbos: Ecology, Evolution, and Conservation of a Megadiverse Region. Oxford University Press, Oxford, pp. 1-25.
78


Binford, L.R., 1978. Nunamiut: Ethnoarchaeology. New York: Academic Press.
Binford, L.R., 1962. Archaeology as anthropology. American antiquity, 28(2], pp.217-225.
Binford, L.R., 1964. A consideration of archaeological research design. American antiquity, 29(4), pp.425-441.
Binford, L.R., 1965. Archaeological Systematics and the Study of Culture Process 1. American antiquity, 31 (2Partl), pp.203-210.
Binford, L.R., 1989. Isolating the transition to cultural adaptations: an organizational approach. The emergence of modern humans: biocultural adaptations in the later Pleistocene, pp.18-41.
Binford, L.R., 1991. When the going gets tough, the tough get going: Nunamiut local groups, camping patterns and economic organization. Ethnoarchaeological Approaches to Mobile Campsites. International Monographs in Prehistory, Ann Arbor, pp.25-137.
Blumenschine, R.J., 1988. An experimental model of the timing of hominid and carnivore influence on archaeological bone assemblages. Journal of Archaeological Science, 15(5), pp.483-502.
Blumenschine, R.J., 1995. Percussion marks, tooth marks, and experimental determinations of the timing of hominid and carnivore access to long bones at FLK Zinjanthropus, Olduvai Gorge, Tanzania. Journal of Human Evolution, 29(1), pp.21-51.
Blumenschine and Marean 1993
Blumenschine, R.J. and Selvaggio, M.M., 1988. Percussion marks on bone surfaces as a new diagnostic of hominid behaviour. Nature, 333(6175), pp.763-765.
Blumenschine, R.J. and Selvaggio, M.M., 1991. On the marks of marrow bone processing by hammerstones and hyenas: their anatomical patterning and archaeological implications. Cultural Beginnings: Approaches to understanding early hominid life-ways in the African savanna. Bonn, pp.17-32.
Blumenschine, R.J., Marean, C.W. and Capaldo, S.D., 1996. Blind tests of inter-analyst correspondence and accuracy in the identification of cut marks, percussion marks, and carnivore tooth marks on bone surfaces. Journal of Archaeological Science, 23(4), pp.493-507.
79


Braun, K., Bar-Matthews, M., Matthews, A., Ayalon, A., Cowling, R.M., Karkanas, P., Fisher, E.C., Dyez, K., Zilberman, T. and Marean, C.W., 2019. Late Pleistocene records of speleothem stable isotopic compositions from Pinnacle Point on the South African south coast. Quaternary Research, 91(1), pp.265-288.
Brain, C.K., 1981. The hunters or the hunted?. University of Chicago Press.
Brown, K.S., Marean, C.W., Jacobs, Z., Schoville, B.J., Oestmo, S., Fisher, E.C., Bernatchez, ]., Karkanas, P. and Matthews, T., 2012. An early and enduring advanced technology originating 71,000 years ago in South Africa. Nature, 491(7425), pp.590-593.
Buhring and Sarnthein 2000
Bunn, H.T. 1981. Archaeological evidence for meat-eating by Plio-Pleistocene hominids from Koobi Fora and Olduvai Gorge. Nature 291: 574-577.
Bunn, H.T., 1983. Comparative analysis of modern bone assemblages from a San hunter-gatherer camp in the Kalahari Desert, Botsuwana, and from a spotted hyena den near Nairobi, Kenya. Animal and Archaeology, 1. Hunters and Their Prey; BAR International Series, 163, pp.141-148.
Cain, C.R., 2005. Using burned animal bone to look at Middle Stone Age occupation and behavior. Journal of Archaeological Science, 32(6], pp.873-884.
Capaldo, S.D. and Blumenschine, R.J., 1994. A quantitative diagnosis of notches made by hammerstone percussion and carnivore gnawing on bovid long bones. American Antiquity, 59(4), pp.724-748.
Chesner, C., Rose, W.I., Deino, A.L., Drake, R. and Westgate, J.A., 1991. Eruptive history of Earth's largest Quaternary caldera (Toba, Indonesia) clarified. Geology, 19(3), pp.200-203.
Claassen, C., 1998. Shells. Cambridge University Press.
Cowling, R.M. and Lombard, A.T., 2002. Heterogeneity, speciation/extinction history and climate: explaining regional plant diversity patterns in the Cape Floristic Region. Diversity and Distributions, 5(3), pp.163-179.
Dibble, Harold L., Curtis W. Marean, and Shannon P. McPherron. "The use of barcodes in excavation projects: examples from Mossel Bay (South Africa) and Roc de Marsal (France)." The SAA Archaeological Record 7.1 (2007): 33-38.
Erlandson, J.M., 2001. The archaeology of aquatic adaptations: paradigms for a new millennium. Journal of Archaeological Research, 9(4), pp.287-350.
80


Esteban, I., Marean, C.W., Cowling, R.M., Fisher, E.C., Cabanes, D. and Albert, R.M., 2019. Palaeoenvironments and plant availability during MIS 6 to MIS 3 on the edge of the Palaeo-Agulhas Plain (south coast, South Africa) as indicated by phytolith analysis at Pinnacle Point. Quaternary Science Reviews.
Fagundes, N.J., Ray, N., Beaumont, M., Neuenschwander, S., Salzano, F.M., Bonatto, S.L. and
Excoffier, L., 2007. Statistical evaluation of alternative models of human
evolution. Proceedings of the National Academy of Sciences, 104(45), pp.17614-17619.
Faith, J.T. and Gordon, A.D., 2007. Skeletal element abundances in archaeofaunal assemblages: economic utility, sample size, and assessment of carcass transport strategies. Journal of Archaeological Science, 34(6), pp.872-882.
Faith, J.T., Marean, C.W. and Behrensmeyer, A.K., 2007. Carnivore competition, bone destruction, and bone density. Journal of Archaeological Science, 34(12), pp.2025-2034.
Fisher, J.W., 1995. Bone surface modifications in zooarchaeology. Journal of Archaeological method and theory, 2(1), pp.7-68.
Fisher, E.C., Bar-Matthews, M., Jerardino, A. and Marean, C.W., 2010. Middle and Late Pleistocene paleoscape modeling along the southern coast of South Africa. Quaternary Science Reviews, 29(11), pp.1382-1398.
Galan, A.B., Rodriguez, M., De Juana, S. and Dominguez-Rodrigo, M., 2009. A new experimental study on percussion marks and notches and their bearing on the interpretation of hammerstone-broken faunal assemblages. Journal of Archaeological Science, 36(3), pp.776-784.
Gathorne-Hardy, F.J. and Harcourt-Smith, W.E.H., 2003. The super-eruption of Toba, did it cause a human bottleneck?. Journal of Human Evolution, 45(3), pp.227-230.
Goldblatt, P., 1997. Floristic diversity in the Cape flora of South Africa. Biodiversity & Conservation, 6(3), pp.359-377.
Hammer, M.F., Karafet, T., Rasanayagam, A., Wood, E.T., Altheide, T.K., Jenkins, T., Griffiths, R.C., Templeton, A.R. and Zegura, S.L., 1998. Out of Africa and back again: nested cladistic analysis of human Y chromosome variation. Molecular biology and evolution, 15(4), pp.427-441.
Harpending, H.C., Batzer, M.A., Gurven, M., Jorde, L.B., Rogers, A.R. and Sherry, S.T., 1998. Genetic traces of ancient demography. Proceedings of the National Academy of Sciences, 95(4), pp.1961-1967.
81


Hawks, J., Hunley, K., Lee, S.H. and Wolpoff, M., 2000. Population bottlenecks and Pleistocene human evolution. Molecular biology and evolution, 17(1), pp.2-22.
Higham, T., Douka, K., Wood, R., Ramsey, C.B., Brock, F., Basell, L., Camps, M., Arrizabalaga, A., Baena, J., Barroso-Rulz, C. and Bergman, C., 2014. The timing and spatiotemporal patterning of Neanderthal disappearance. Nature, 512(7514), p.306.
Hodgkins J, Marean CW, Turq A, Sandgathe D, McPherron SP, Dibble HL, 2016. Climate-mediated shifts in Neandertal subsistence behaviors at Pech de l’Aze IV and Roc de Marsal (Dordogne Valley, France) Journal of Human Evolution. 96:1-18.
Jacobs, Z., 2010. An OSL chronology for the sedimentary deposits from Pinnacle Point Cave 13B—a punctuated presence. Journal of Human Evolution, 59(3), pp.289-305.
Johnson, E., 1985. Current developments in bone technology. In Advances in archaeological method and theory (pp. 157-235). Academic Press.
Kaplan, J., 1997. Archaeological study: proposed Pinnacle Point development. Riebeek West, South Africa: Agency for Cultural Resource Management.
Karkanas, P. and Goldberg, P., 2010. Site formation processes at Pinnacle point Cave 13B (Mossel Bay, Western Cape Province, South Africa): resolving stratigraphic and depositional complexities with micromorphology. Journal of Human Evolution, 59(3), pp.256-273.
Karkanas, P., Brown, K.S., Fisher, E.C., Jacobs, Z. and Marean, C.W., 2015. Interpreting human behavior from depositional rates and combustion features through the study of sedimentary microfacies at site Pinnacle Point 5-6, South Africa .Journal of human evolution, 85, pp.1-21.
Kelly, R.L., 1995. The foraging spectrum.
Klein, R.G., 1983. Palaeoenvironmental implications of Quaternary large mammals in the fynbos region. Fynbos palaeoecology: a preliminary synthesis, pp.116-138.
Klein, R.G. and Cruz-Uribe, K., 1984. The analysis of animal bones from archeological sites. University of Chicago press.
Kosso, P., 1991. Method in archaeology: Middle-range theory as hermeneutics. American Antiquity, 56(4), pp.621-627.
82


Lam, Y.M., Chen, X., Marean, C.W. and Frey, C1998. Bone density and long bone representation in archaeological faunas: comparing results from CT and photon densitometry. Journal of Archaeological Science, 25(6), pp.559-570.
Levin, M.G., Potapov, L.P. and Dunn, S.P., 1964. The peoples of Siberia. University of Chicago Press.
Lewontin, R.C., 1972. The apportionment of human diversity. In Evolutionary biology (pp. 381-398). Springer, New York, NY.
Lisiecki, L.E. and Raymo, M.E., 2005. A Pliocene-Pleistocene stack of 57 globally distributed benthic 5180 records. Paleoceanography, 20(1).
Loftus, E., Lee-Thorp, ]., Leng, M., Marean, C. and Sealy, ]., 2019. Seasonal scheduling of shellfish collection in the Middle and Later Stone Ages of southern Africa. Journal of human evolution, 128, pp.1-16.
Lyman, R.L., 1985. Bone frequencies: differential transport, in situ destruction, and the MGUI. Journal of Archaeological Science, 12(3), pp.221-236.
Marean, CW., (Ed.). 2010a. The Middle Stone Age at Pinnacle Point Site 13B, a Coastal Cave near Mossel Bay (Western Cape Province, South Africa) [Special Issue]. Journal of Human Evolution, 59, (3-4).
Marean, C.W., 2010b. Introduction to the Special Issue—The Middle Stone Age at Pinnacle Point Site 13B, a Coastal Cave near Mossel Bay (Western Cape Province, South Africa).
Marean, C.W., 2010c. Pinnacle Point Cave 13B (Western Cape Province, South Africa) in context: the Cape floral kingdom, shellfish, and modern human origins. Journal of Human Evolution, 59(3), pp.425-443.
Marean, C.W., 2011. Coastal South Africa and the coevolution of the modern human lineage and the coastal adaptation. In Trekkingthe shore (pp. 421-440). Springer, New York, NY.
Marean, C.W., 2014. The origins and significance of coastal resource use in Africa and Western Eurasia. Journal of Human Evolution, 77, pp.17-40.
Marean, C.W., 2015. The most invasive species of all. Scientific American, 313(2), pp.32-39.
Marean, C.W. and Assefa, Z., 1999. Zooarcheological evidence for the faunal exploitation behavior of Neandertals and early modern humans. Evolutionary Anthropology: Issues, News, and Reviews: Issues, News, and Reviews, 5(1), pp.22-37.
83


Marean, C.W. and Frey, C1997. Animal bones from caves to cities: reverse utility curves as methodological artifacts. American Antiquity, 62(4), pp.698-711.
Marean, C.W. and Kim, S.Y., 1998. Mousterian large-mammal remains from Kobeh Cave behavioral implications for Neanderthals and early modern humans. Current anthropology, 29(S1), pp.S79-S114.
Marean, C.W., Abe, Y., Frey, C.J. and Randall, R.C., 2000. Zooarchaeological and taphonomic analysis of the Die Kelders Cave 1 Layers 10 and 11 Middle Stone Age larger mammal fauna .Journal of Human Evolution, 38[ 1), pp.197-233.
Marean, C.W., Bar-Matthews, M., Bernatchez, ]., Fisher, E., Goldberg, P., Herries, A.I., Jacobs, Z., Jerardino, A., Karkanas, P., Minichillo, T. and Nilssen, P.J., 2007. Early human use of marine resources and pigment in South Africa during the Middle Pleistocene. Nature, 449(7164), pp.905-908.
Marean, Curtis W., Nilssen, P.J., and Kyle Brown, Antonieta Jerardino, and Deano Stynder. 2004. Paleoanthropological investigations of Middle Stone Age sites at Pinnacle Point, Mossel Bay (South Africa): archaeology and hominid remains from the 2000 field season. Paleoanthropology 2, no. 1.
Martinson, D.G., Pisias, N.G., Hays, J.D., Imbrie,}., Moore, T.C. and Shackleton, N.J., 1987. Age dating and the orbital theory of the ice ages: development of a high-resolution 0 to 300,000-year chronostratigraphy. Quaternary research, 27(1), pp.1-29.
Mellars, P., 1996. Symbolism, language, and the Neanderthal mind. Modelling the early human mind, pp.15-32.
Metcalfe, D. and Jones, K.T., 1988. A reconsideration of animal body-part utility indices. American Antiquity, 52(3), pp.486-504.
Morlan, R.E., 1984. Toward the Definition of Criteria for the Recognition of Artificial Bone Alterations 1. Quaternary Research, 22(2), pp.160-171.
O'Connell, J.F., Hawkes, K. and Jones, N.B., 1988. Hadza hunting, butchering, and bone transport and their archaeological implications. Journal of Anthropological research, 44(2), pp.113-161.
O'Connell, J.F. and Marshall, B., 1989. Analysis of kangaroo body part transport among the Alyawara of Central Australia. Journal of Archaeological Science, 16(4), pp.393-405.’
84


Oestmo, S. and Marean, Curtis W., 2014. Pinnacle Point: Excavation and Survey Methods. In Smith C. (ed.) Encyclopedia of Global Archaeology, New York: Springer.
Outram, A.K., 2001. A new approach to identifying bone marrow and grease exploitation: why the "indeterminate” fragments should not be ignored. Journal of archaeological science, 28(4), pp.401-410.
Pickering, T.R., 2002. Reconsideration of criteria for differentiating faunal assemblages accumulated by hyenas and hominids. International Journal of Osteoarchaeology, 12(2), pp.127-141.
Potts, R., 2017. Early Hominid Activities at Olduvai: Foundations of Human Behaviour. Routledge.
Potts, R. and Shipman, P., 1981. Cutmarks made by stone tools on bones from Olduvai Gorge, Tanzania. Nature, 291(5816), p.577.
Potts, R., Shipman, P. and Ingall, E., 1988. Taphonomy, paleoecology, and hominids of Lainyamok, Kenya .Journal of Human Evolution, 17(6), pp.597-614.
Rampino, M.R., Ambrose, S.H., McCoy, F. and Heiken, G., 2000. Volcanic hazards and disasters in human antiquity. Geological Society of America Special Paper, 345, pp.71-82.
Rampino, M.R. and Self, S., 1992. Volcanic winter and accelerated glaciation following the Toba super-eruption. Nature, 359(6390), p.50.
Rector, A.L. and Reed, K.E., 2010. Middle and late Pleistocene faunas of Pinnacle Point and their paleoecological implications. Journal of Human Evolution, 59(3), pp.340-357.
Relethford, J.H., 1995. Genetics and modern human origins. Evolutionary Anthropology: Issues, News, and Reviews, 4(2), pp.53-63.
Renfrew, C. and Zubrow, E.B.W., 1994. Towards a cognitive archaeology.
South African Coast Paleoclimate Paleoenvironment Paleoecology Paleoanthropology Project, 2017. [Pinnacle Point Excavation Database]. Unpublished Raw Data.
Schoville, B.J. and Otarola-Castillo, E., 2014. A model of hunter-gatherer skeletal element transport: The effect of prey body size, carriers, and distance. Journal of human evolution, 73, pp.1-14.
85


Selvaggio, M.M., 1998. Evidence for a three-stage sequence of hominid and carnivore involvement with long bones at FLKZinjanthropus, Olduvai Gorge, Tanzania. Journal of Archaeological Science, 25(3), pp.191-202.
Shackleton, N.J. and Opdyke, N.D., 1973. Oxygen isotope and palaeomagnetic stratigraphy of Equatorial Pacific core V28-238: Oxygen isotope temperatures and ice volumes on a 105 year and 106 year scale. Quaternary research, 3(1), pp.39-55.
Shipman, P., 1981. Applications of scanning electron microscopy to taphonomic problems. Annals of the New York Academy of Sciences, 276(1), pp.357-385.
Shipman, P. and Rose, ]., 1983. Early hominid hunting, butchering, and carcass-processing behaviors: approaches to the fossil record. Journal of anthropological Archaeology, 2(1), pp.57-98.
Smith, E.I., Jacobs, Z., Johnsen, R., Ren, M., Fisher, E.C., Oestmo, S., Wilkins, J., Harris, J.A., Karkanas, P., Fitch, S. and Ciravolo, A., 2018. Humans thrived in South Africa through the Toba eruption about 74,000 years ago. Nature, 555(7697), p.511.
Stiner, M.C., Kuhn, S.L., Weiner, S. and Bar-Yosef, O., 1995. Differential burning, recrystallization, and fragmentation of archaeological bon e. Journal of Archaeological Science, 22(2), pp.223-237.
Thompson, J.C., 2010. Taphonomic analysis of the Middle Stone Age faunal assemblage from Pinnacle Point Cave 13B, Western Cape, South Africa .Journal of Human Evolution, 59(3), pp.321-339.
Thompson, E., Williams, H.M. and Minichillo, T., 2010. Middle and late Pleistocene middle stone age lithic technology from Pinnacle point 13B (Mossel Bay, western Cape province, South Africa f Journal of Human Evolution, 59(3), pp.358-377.
Villa, P. and Mahieu, E., 1991. Breakage patterns of human longb ones. Journal of human evolution, 21(1), pp.27-48.
Walter, R.C., Buffler, R.T., Bruggemann, J.H., Guillaume, M.M., Berhe, S.M., Negassi, B., Libsekal, Y., Cheng, H., Edwards, R.L., von Cosel, R. and Neraudeau, D., 2000. Early human occupation of the Red Sea coast of Eritrea during the last interglacial. Nature, 405(6782), p.65.
Watts, I., 2010. The pigments from Pinnacle Point Cave 13B, Western Cape, South Africa. Journal of Human Evolution, 59(3), pp.392-411.
86


Wilkins, J., Brown, K.S., Oestmo, S., Pereira, T., Ranhorn, K.L., Schoville, B.J. and Marean, C.W., 2017. Lithic technological responses to Late Pleistocene glacial cycling at Pinnacle Point Site 5-6, South Africa. PloS one, 12(3), p.e0174051.
Yellen, J.E., 1991. Small mammals: !Kung San utilization and the production of faunal assemblages. Journal of Anthropological Archaeology, 10(1), pp.1-26.
Zielinski, G.A., 2000. Use of paleo-records in determining variability within the volcanism-climate system. Quaternary Science Reviews, 19(1-5), pp.417-438.
Zielinski, G.A., Mayewski, P.A., Meeker, L.D., Whitlow, S., Twickler, M.S. and Taylor, K., 1996. Potential atmospheric impact of the Toba Mega-Eruption~ 71,000 years ago. Geophysical Research Letters, 23(8), pp.837-840.
87


Full Text

PAGE 1

HUMAN ADAPTABILITY IN THE TIME OF CHANGING CLIMATE: A ZOOARCHAEOLOGICAL ANALYSIS AT PINNACLE POINT SITE 5 6, MOSSEL BAY, SOUTH AFRICA by SARAH MAUREEN SIMEONOFF B.A. , Arizona State University, 2013 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Arts Anthropology Department 201 9

PAGE 2

ii 201 9 SARAH MAUREEN SIMEONOFF ALL RIGHTS RESERVED

PAGE 3

iii This thesis for the Masters of Arts degree by Sarah Maureen Simeonoff has been approved for the Anthropology Program by Jamie M. Hodgkins, Chair Chris Beekman Tammy Stone Date: July 27 , 201 9

PAGE 4

iv Simeonoff, Sarah Maureen (M.A., Anthropology) Human A daptability in the time of C hanging C limate: A Z ooarchaeological A nalysis at P innacle P oint site 5 6, M ossel B ay, S outh A frica Thesis directed by Associate Professor Jamie M. Hodgkins ABSTRACT The adaptability of modern humans is often cited as key to our successful global survival and proliferation ; however, little work has been done to quantify this trait through time. This project exa mines the Marine Isotope Stage (MIS) 5 4 transition occurring ~81 71 thousand years ago (ka) on the southern coast of South Africa at a n Upper/Middle Paleolithic site on the southern coast of South Africa. During this time, the environment underwent chal lenges related to transgressing and regressing coastlines, changing faunal availability, and rapid changes in mean temperature. A zooarchaeological analysis was undertaken to determine if and to what extent the modern humans occupying the cave modified the ir subsistence strategies during this climactic shift. The results of this project indicate a temporary change in subsistence strategies occurred at the onset of glacial MIS 4, followed by a return to the behaviors of MIS 5 as the climate fully transitione d. These findings suggest the population occupying PP5 6 was sufficiently behaviorally elastic as to withstand environmental changes related to resource availability. This study provides insights into the adaptability of modern humans, and further provides a record against which the adaptability of other hominid species can be tested. This form and content of this abstract are approved. I recommend its publication. Approved: Jamie M. Hodgkins

PAGE 5

ACKNOWLEDGEMENTS I am grateful to my committee, J amie Hodgkins, Chris Beekman, and Tammy Stone , for their helpful comments and support. I n particular, I would like to thank my advisor, Jamie Hodgkins, for her mentorship and guidance. My sincerest gratitude to Curtis Marean for setting the tarsal identifi cation bar so high that I will be aspiring to it for the rest of my career. T he opportunity to work at Pinnacle Point ( both in the field and the lab ) was a gift. W ithout the ceaseless support and encouragement of my parents, Mary and Richard Simeonoff, I he confidence to apply for and finish graduate school . I appreciate the constant reminder s to work hard and to trust my abilities ; making you proud has always motivated me to do better. To the best brothers to ever live, Richard and Joh n, please eat my dust. T hanks for always putting this degree into perspective. I only did it to make you guys look bad. My sincerest gratitude to HK Gatti for pulling no punches in her approach to encouragement, It was a critical push near the end, but the gentler cheering was also appreciated. In a similar vein, I want to thank Darsita Ryan for always being willing to tell me to get my act together and write , and for offering a sympathetic ear and honest advice . R . J ackson B ewley , thank you for showing up at the exact right moment . Your support was crucial in keeping me calm and focused . Thank you for always reminding me about the big picture in little ways . The prospect of using this degree to win argumen ts gave me a strong incentive to finish. Last and never least, I could not have finished this degree without the theory d eciphering commiserat ion, comic relief, and true friendship of my partner in crime and archaeology , Breeanna Chantel Charolla. Without you as a teammate has never been an option thank you for everything. We may as well be co authors on this bad boy for all of the emotional support, encouragement, editing, and will to live you offered during its labor . There's no one else I would have preferred to suffer alongside .

PAGE 6

vi TABLE OF CONTENTS CHAPTER I. INTRODUCTION ................................ ................................ ................................ ................................ ......... 1 Theoretical Basis ................................ ................................ ................................ ................................ ........ 4 Pinnacle Point Complex ................................ ................................ ................................ .......................... 7 Environmental Setting --------------------------------------------------------------------------7 Pinnacle Point Arc haeological Context -------------------------------------------------------9 Marine Isotope Stage 5 4 Transition at PP5 6 ................................ ................................ ........... 13 Light Brown Sand and Roofspall (LBSR) ---------------------------------------------------15 Ashy Light Brown Sand (ALBS) --------------------------------------------------------------17 Shelly Ash y Dark Brown Sand (SADBS) ----------------------------------------------------19 Toba Eruption and Glacial Winter -----------------------------------------------------------22 II. METHODS ................................ ................................ ................................ ................................ ................... 25 Excavation Techniques at Pinnacle Point ................................ ................................ ..................... 25 Sampling MIS 5 4 Transitional Fauna ................................ ................................ ............................ 27 Zooarchaeological Analysis ................................ ................................ ................................ ................ 28 Taxonomic Identification ---------------------------------------------------------------------31 Body Size -----------------------------------------------------------------------------------------31 Fragment Size -----------------------------------------------------------------------------------28 Weathering --------------------------------------------------------------------------------------29 Surface Visi bility -------------------------------------------------------------------------------30 Burning -------------------------------------------------------------------------------------------32 Transportation Strategies --------------------------------------------------------------------32 Breakage Ana lysis ------------------------------------------------------------------------------35 Surface Modifications -------------------------------------------------------------------------37 3D GIS Visualization ................................ ................................ ................................ .............................. 42 III. RESULTS ................................ ................................ ................................ ................................ ..................... 44 Fragment Size ................................ ................................ ................................ ................................ ........... 44 Weathering ................................ ................................ ................................ ................................ ................ 45

PAGE 7

vii Surface Visibility ................................ ................................ ................................ ................................ ..... 46 Taxonomic Identification ................................ ................................ ................................ .................... 47 Body Size ................................ ................................ ................................ ................................ .................... 49 Burning ................................ ................................ ................................ ................................ ........................ 52 Nutritive and Non nutritive Breakage ................................ ................................ ........................... 53 Food Transport ................................ ................................ ................................ ................................ ........ 57 Surface Modification ................................ ................................ ................................ .............................. 60 3D GIS ................................ ................................ ................................ ................................ ........................... 62 IV. DISCUSSION AND CONCL USION ................................ ................................ ................................ ....... 69 LBSR Discussion ................................ ................................ ................................ ................................ ...... 69 ALBS Discussion ................................ ................................ ................................ ................................ ...... 70 SADBS Discussion ................................ ................................ ................................ ................................ ... 72 Conclusions ................................ ................................ ................................ ................................ ................ 73 V. REFERENCES ................................ ................................ ................................ ................................ ............ 73

PAGE 8

viii LIST OF TABLES TABLE Table 1. Coded specimens per StratAgg. ................................ ................................ ................................ .... 21 Table 2. Body size categories (Brain 1981) and example South African fauna (Thompson 2010). ................................ ................................ ................................ ................................ ...................... 32 Table 3. Observed weather stages by StratAgg. ................................ ................................ ...................... 46 Table 4. Com parison between levels of the NISP coded for each Family, Order, Class by Body Size. Each specimen was coded to the smallest unit of taxonomic unit. ........ 51 Table 5. Minimum number of elements, MAU, SFUI and normed MAU values for each StratAgg, used to create Figure 2. Created using SFUI, as calculated by Metcalfe and Jones (1981). ................................ ................................ ................................ ............................... 58 Table 6. Summary table of the key differences between StratAggs during the MIS 5 4 transition. ................................ ................................ ................................ ................................ .............. 75

PAGE 9

ix LIST OF FIGURES FIGURE Figure 1. Location of Pinnacle Point relative to the coast. Sites PP13B and PP5 6 are also shown in relation to each other. Photo published in Karkanas et al. 2015. ............. 13 Figure 2. Lateral view of site, facing east of StratAggs from PP5 6, with the relevant sections highlighted. Figure created by Erich Fisher (University of Colorado Denver). ................................ ................................ ................................ ................................ .................. 14 Figure 3. Average, minimum, and maximum coastal distances for each StratAgg, as modeled by Fisher et al. 2010. ................................ ................................ ................................ ...... 22 Figure 4. Excavation underway at PP5 6. SubAggs are identified on the wall using tags, with StratAggs marked with larger tags. Photo facing east. ................................ ............ 27 Figure 5. Examples of cut marks from PP5 6, showing the relatively straight morphology and clustering of marks. ................................ ................................ ................................ .................. 39 Figure 6. Sample of percussion modification marks. Clockwise from top left: microstriations, conchoidal flake scar, flake with two medullary percussion notche s, percussion notch on cortical surface. Examples from unpublished experimental data by author. ................................ ................................ ................................ ........ 41 Figure 7. Average length and width for each StratAgg (millimeters). ................................ .......... 45 Figure 8 . Surface visibility by StratAgg. ................................ ................................ ................................ ...... 47 Figure 9. Propor tion of identifiable taxa for each StratAgg. Taxa representing less than one percent of the assemblage are not reported in this figure. ................................ .............. 49 Figure 10. Relative frequencies of faunal body sizes through the transition. ........................... 50

PAGE 10

x Figure 11. Percentage burned material per Str atAgg, including percent fully calcined, partially calcined, fully carbonized, and partially carbonized. ................................ ....... 53 Figure 12. Non nutritive breakage for each StratAgg, compared to experimental data from Marean et al. 2000. ................................ ................................ ................................ ............................ 55 Figure 13 . Nutritive breakage fo r each StratAgg, compared to experimental data from Marean et al. 2000. ................................ ................................ ................................ ............................ 56 Figure 14. Distance weighted mean squares curve fitted to each StratAgg and plotted 2007). ................................ ................................ ................................ ................................ ...................... 59 Figure 15. The r esults of the ANCOVA with Poisson distribution to determine the least squares mean number of cut marks per fragment, controlled for surface area (calculated using the geometric mean). The error bars show the 95% confidence intervals for each StratAgg. ................................ ................................ ................................ ............ 61 Figure 16. The results of the ANCOVA with Poisson distribution to determine the least squares mean number of percussion marks per fragment, controlled for surface area (calculated using the geometric mean). The error bars show the 95% confidence intervals for each StratAgg. ................................ ................................ .................... 62 Figure 17. Analyzed fauna from the transition, identified by taxon. ................................ ............. 65 Figure 18. Visualization of faunal remains located within combustion features versus those located in other contexts. ................................ ................................ ................................ ................ 66 Figure 19. Burned bones, displayed by level of burning across the transition. ........................ 67

PAGE 11

xi Figure 20. Anthrop ogenically modified bones shown in association with bones recovered from hearth contexts (in black). ................................ ................................ ................................ .. 68 Figure 21. Summary figure of key differences between StratAggs. Dashed lines indicate usage of the secondary axis. ................................ ................................ ................................ .......... 76

PAGE 12

1 CHAPTER I INTRODUCTION Anatomically modern humans (AMH) have existed alone for 39,000 years (Higman 201 4 ) and successfully inhabited nearly every environmental niche on the planet. In response to varied environmental clines and fluctuations of resource availability within these clines Homo sapiens have continually adapted. The planet has always undergone flu ctuations in mean temperature which have modified environments across the globe (Martinson et al. 1987) , necessitating behavioral modifications by humans to ensure their continued survival. Humans, with the help of modern technology, now number in the bill ions by far the most successful hominid species ever to proliferate and inhabit all continents. Enumerating and quantifying t he ways in which AMH have modified their behaviors through time to adapt and expand is crucial to human origins research. To th at end , this study examines a regularly occupied Middle / Upper Paleolithic site and focuses on human subsistence strategies across a global climatic transition from an interglacial to a glacial climate. The extinction of numerous hominid species has been a ttributed to an inability to successfully adapt to changing environments. No other hominid better exemplifies the debate for human superior adaptability than the demise of Homo neanderthalensis. H. neanderthalensis disappeared from the archaeological recor d approximately 39 kya, coinciding roughly with the appearance of anatomically modern humans (AMH) in Europe (Higman 201 4 ). In attempting to understand the H. neanderthalensis extinction, researchers have focused on the asserted superior adaptability of AM H (Villa and Roebucks

PAGE 13

2 2014 provide a thorough discussion of the argument). Although this comparison is not paramount to this research, it best illustrates the oft touted adaptability of AMH. With regard to subsistence strategies, several authors have argue d both for an d against the idea that humans are more adaptable to varied and changing climates than Neanderthals (Mellers 1996, McBrearty and Brooks 2000, Marean 2005). Although no comparison between Neanderthal intelligence, symbology, or strategies are u ndertaken during this research, it is the adaptability of AMH that frames my research at Pinnacle Point. Did the AMH occupying the site during a glacial interglacial shift successfully adapt their strategies in the face of environmental change? If so, an e xample of early human adaptability will be illuminated. Although this study alone cannot adequately solve the debate on human adaptability , it can offer a record against which future studies (from different geographic, temporal, and occupant sites) can be compared. Specifically, this project examines the Marine Isotope Stage (MIS) 5 4 transition occurring ~81 71 thousand years ago (ka) on the southern coast of South Africa. During this time, the environment underwent challenges related to transgressing a nd regressing coastlines, changing faunal availability, and rapid changes in mean temperature (Karkanas et al . 2015, Lisiecki and Raymo 2005, Martinson et al. 1987, Wilkins et al. 2017). Should the humans at Pinnacle Point exhibit an exceptional ability to mold their behavior in the face of global success. Through zooarchaeological analysis, this research seeks to determine if and to what ext ent environmental changes on the so uthern coast of South Africa influenced changes in

PAGE 14

3 the subsistence strategies of the anatomically modern humans occupying Pinnacle Point Site 5 6 (PP5 6). The Pinnacle Point Site Complex (Pinnacle Point) is a cluster of caves, rockshelters, and open air si tes very near the modern day South African coast (<30 meters) (Kaplan 1997, Marean et al. 2007, Fisher et al. 2010, Jacobs 2010). The archaeological sequence at PP5 6 begins at approximately 90 ka and continues through 55 ka, encompassing three Marine Isot ope Stage s (MIS) and two transitions ( from 5 4 3 , Interglacial Glacial Interglacial) (Karkanas et al. 2015, Wilkins et al. 2017), and its location near both terrestrial and coastal resources makes it an ideal site to study human adaptation to different bio mes (coastal and plains). Understanding past subsistence, particularly at a site complex which has been regularly occupied, is crucial to decoding the evolution of modern humans relative to other hominid groups. This study begin s by discussing the theoret ical basis used to develop the question, then describe s the zooarchaeological analysis undertaken, and finally provide s the results and implications of this research.

PAGE 15

4 CHAPTER II B ackground Theoretical Basis Pinnacle Point 5 6 ( PP5 6 ) , the site at the center of this research, was routinely occupied between 90 55 ka (Karkanas et al. 2015, Wilkins et al. 2017). The limitations of studying a site and time period so deep in the past requires that the theoretical approach appropriately ma tch the dearth of culturally modified material (e.g. ochre, incised shells, beads, etc.). Archaeological assemblages dated to the Middle/Upper Paleolithic are characterized by utilitarian remains ; t hat is, they generally lack culture signifying artifacts a nd instead typically contain lithic tools and debris, faunal remains, combustion features, and little else . Because there are no clear social strata, very few instances of cultural material, and a lack of comparable archaeological data, both the symbolic a nd political economy theoretical school s will be rejected in favor of a processual approach. I Lewis Binford proposed a change to the prevailing theoretical basis of archaeolog ical inquiry (Binford 1962, 1964), arguing that "traditional need ed a systematic framework better suited to the theoretical questions being posed. Traditional archaeology, Binford suggested, suffered from a paradox in which cultural ideology is identified and explained by the material culture alone (Binford 1989) . Processual archaeology steers clear of addressing "the ancient mind" in favor of using physical artifacts to inform past socio economic and cultural conditions (Renfrew 1994). In the processual view, t he primary motivation of cultural shift s is environmental change (Binford 1962, 1965, 1989). Culture, then, is a means of adapting to changing environments

PAGE 16

5 ( Binford 1965:203). Processualists are positivist in their view of archaeology and believe concrete (Binford 1962:224). Because processualism views culture as a response to the environment, the entir e system is the preferred unit of analysis, as it constitutes a homeostatic system that responds to environmental change through adaptation (Binford 196 4 , 1989, Renfrew 1994). Binford , in particular , emphasized the need to look at all changes within a syst em . Because the available data does not allow research to focus on the isolation of one chang e, the only way to study the full scope of a society is to investigate the sum of its parts (Binford 1965). Objective analysis of archaeological remains as they re late to the past is the driving force behind processual methodology . There is a strong preference for a deductive middle range approach where descri ption reconstructions are routinely employed (Kosso 1991: 622). The question posed by this research is predicated on the processual notion that humans adapt to changes in their environment; this particular portion of the PP5 6 sequence was se lected because of the known climatic changes occurring globally (e.g. interglacial to glacial). Further, the limiting factors of faunal analysis require the scope of the question be focused on subsistence changes between levels. Because little additional d ata are available on the social strata or culturally modified artifacts , this analysis will

PAGE 17

6 focus on changes in food procurement and processing because it is a sub system that can be isolated from the broader system and analyzed separately . The processual approach affects not only the development of the question but also the understanding of the results. Should the analysis reveal human s modified their subsistence strategies in response to the cooler temperatures and changing faunal availabil ity of MIS 4, processual archaeology suggests the change indicates a well developed ability to respond to the environment. Alternatively, should the occupants show no sign of changing strategies, processual archaeology suggests the environmental changes o ccurring at Pinnacle Point during the transition were not substantial enough to require behavior modifications. The processual approach is best suited to this undertaking, given the environment.

PAGE 18

7 Pinnacle Point Complex Environmental Setting The Pinnacle Point cave complex (PP) is located approximately 10 km west of the town of Mossel Bay, on the W estern Cape of South Africa. The southern coast of South Africa exists in a unique climate , interaction of these various systems means the area is subject ed to both winter rains from the we st and bimodal rains ( a combination of summer and year round rains ) from the center (Copeland et al. 2016) , resulting from the warm waters of the Indian Ocean (Engelbracht e t al. 2015) . Strontium isotope analysis suggests the same rainfall regime would hav e persisted during the Middle/Upper Paleolithic (Sealy 1996, Bar Matthews et al. 2010, Copeland et al. 2016). Mossel Bay is situated in the year round rainfall region, though the boundary for winter rainfall and summer rainfall regimes fall close by (Braun et al . 2018). The climate is su bsequently moderate, averaging 18° C (64.4° F) , with seldom frost (Esteban et al. 2019). Mossel Bay , and Pinnacle Point, is situated within the G r e ater C ape F loristic R egion (GCFR), known for its high diversity of plant species (Cowling and Lombard 2002, Esteban et al . 2019 , Goldblatt 1997,). The GCF R consists of five distinct biomes: f ynbos, r enosterveld, s ucculent k aroo, t hicket, and f orest (Bergh et al. 2014). Of these, f ynbos, r enosterveld, and t hicket are situated ne ar Pinnacle Point , with fynbos comprising the largest proportion of the extant environment . These three biomes represent distinct plant species and environments. Fynbos contains the greatest diversity of plant species and is

PAGE 19

8 2014:11). It is considered one of the richest floral regions in the world, with a high number of endemic species (Bergh et al. 2014). Renosterveld is also characteriz ed by small leaved shrubs; however, this biome is dominate d by more C3 grasses and features more nutrient rich soils than fynbos (Bergh et al. 2014: 15). Finally, the t possess ion of a large diversity of growth forms, creatin g a dense canopy of largely evergreen, broadleaved, sclerophyllous, spiny and/or succulent shrubs and low trees (Bergh et al. 2014: 8). Of the three biomes, f ynbos is the dominate environment surrounding Pinnacle Point and the extant fauna living in thes e environments are distinct from the rest of Africa (Rector and Reed 2010) . Mammalian diversity is relatively low in fynbos (Klein 1983, Marean 2010) and was likely similar through time (Rector and Reed 2010). Large bodied mammals are uncommon due to th e nutrient poor soils found in the fynbos biome thus suggesting the environment was likely never conducive to large mammal hunting (Klein 1983, Marean 2010). Extant small 0: 431). Klein (1983) called the mammalian fauna in the y in South Africa, additionally citing grey duiker, steenbok, and grysbok as common species. In the past, however, large bod ied herbivores (such as Wildebeest, giant zebra, and hartebeest) would have existed on the landscape (Rector and Reed 2010, Copeland et al. 2016). Published reports on the faunal assemblages from Pinnacle Point show that the majority of faunal remains reco vered are not identifiable to taxon (Thompson 2010, Rector and Reed

PAGE 20

9 2010, Marean 2010) ; however, the common species and diversity in fynbos is important to understanding the archaeological assemblage at Pinnacle Point . Pinnacle Point Archaeological Context Included in the Pinnacle Point Complex are 28 sites (21 of which are dated to the Middle Stone Age ~20 ka 200 ka) that were identified and recorded during a cultural resource management survey of the area prior to the development of the Pinnacle Point E state (Jacobs 2010, Kaplan 1997, Marean et al . 2007). In particular, f ifteen cave and rockshelter sites are clustered near the coas t , averaging 50 meters in elevation and situated within 3 0 meters from the modern coast below (Fisher et. al 2010). This places them within both coastal and plains biomes, with access to nutrients from both terrestrial and intertidal resources. This portion of the South African coast is situated against the Agulhas Plain , a ing ( Fisher et al. 2010: 1384) . This bank is relatively flat lying and directly abuts the modern coast. During the change from interglacial to glacial climate, oceans contract to ice cores and cause sea levels to drop (Shackelton and Opdyke 1973). For the coas t near PP, this causes a drastic change in environment (Fisher et al. 2010, Copeland et al. 2016) by exposing a broad floodplain (Cawthra et al. 2015), the ecology of which is still unknown (Copeland et al. 2016). A model of the Paleo Agulhas Plain (PAP) a cross glacial cycles was undertaken to better understand the impact of chang ing climate on the exposed landmass of the PAP (Fisher et al. 201 0 ). This model shows that at peak regression of the ocean approximately 80,000 km 2 of grassland would be exposed be tween PP and the coast (Fisher et al. 201 0 , Copeland et al.

PAGE 21

10 2016). Although the model only accounts for coastal distance in broad ranges (i.e. minimum, maximum, average distances) and cannot account for the exact coastal difference at precise dates , it hel pfully provides comparative shoreline distances. This is, while the model does not wholly account for the distance of the coast at any given time, it provide s useful comparative distances across large time spans, such as the transition from interglacial t o glacial climate. E ven subtle shifts (i.e. 5 or 10 km) in the ocean levels would have altered the coastal environment surrounding PP (Marean et al. 2014). T he exact nature of the exposed PAP is not well understood ; however, some researchers suggest the br oad open grassland would have provided ample resources for large ungulates (Cawthra et al. 2015, Copeland et al. 2016). Thus, the movement of the coast and exposure of the PAP during the MIS 5 to 4 transition at PP is a key factor in understanding the reso urce procurement of PP5 6 inhabitants. Excavatio n of five cave sites from the complex was undertaken by Curtis Marean and Peter Nielssen in 199 9. Since the initial excavations, control of the work has shifted to the South African Coast Paleoclimate Paleo environment Paleoecology Paleoanthropology P is early human adaptability at Pinnacle Point through analysis of the use of symbology, lithic technolog y , and subsistence strategies across various changes in climate and e nvironment (Marean 2010 a ). To date, several site s from Pinnacle Point have been tested or partially excavated, including PP 5 6 , the focus of this thesis ( see Marean et al . 2004 , Marean 2010b, 2010c ). Another cave site was previously excavated and analyzed a t Pinnacle Point: Pinnacle Point Cave 13B (PP13B) , located less than one kilometer from PP5 6. A brief discussion about the information

PAGE 22

11 derived from the record at PP13B provide s useful insight into the archaeological history of Pinnacle Point and offer s a backdrop for the analysis of PP5 6. Excavations of PP13B revealed the earliest known evidence of intertidal marine sh ellfish exploitation by humans at approximately 164 ka (Marean et al. 2007), pushing the date of this behavior back in time approximatel y 40 ka (Erlandson 2001, Walter et al . 2000). Intertidal shellfish exploitation by humans requires an understanding of the lunar cycle, and , Marean (2011:434) suggests : funct ions [which] allowed them to link lunar phases to tidal rhythms and thus, develop an effective way to schedule visits to the coast in a manner that maximized returns from the coastal resources . Th e adoption of intertidal foraging shows a change in cognit ion and behavior from a reliance on terrestrial resources to an expansion towards marine sources a change which likely required an advanced cognition (cognitive adaption) to understand the tidal schedule (Marean 2014, Marean 2015, Marean 2016 , Loftus et al. 2019 ) . PP13B also revealed early evidence of pigment usage (Watts 2010) and bladelet production (Jacobs 2010) , both beginning near 16 2 ka ( Marean et al . 2007, Marean 2010 c, Thompson et al . 2010 ). Marean and colleagues have posited that these changes in behavior were related to the changing glacial environment of MIS 6 (~196 123 ka) (Marean 2010c). This leap in technological and cognitive ability suggests that the occupants of PP13B were adapting to the movement of the Paleolithic coastline and resource availability, laying the groundwork for further investigation into the MIS transition occurring at PP5 6 (Marean 2010c, Marean 2014). PP13B provides a limited but fascinating record of human occupation in the region 010c:440] ) between approximately 164 90 ka at

PAGE 23

12 which time a sand dune sealed the cave, preventing further occupation until 40 ka ( Karkanas and Goldberg 2010, Marean et al . 200 7 , Marean 2010 b ). The record provided by PP13B provides interesting insights in to early human activity along the coast of South Africa, however, the gap in the record (between 90 40 ka) leaves many questions unanswered about the behavior of anatomically modern humans through time (Marean 2010 a ). The sequence at PP5 6, in contrast, begins at the time the sand dune sealed PP13B (~90 ka) and continues through 55 ka with little to no occupational interruption (Marean 2010 a , Karkanas et al . 2015, Wilkins et al . 2017). The record preserved at PP5 6 presents an opportunity to study change s in subsistence strategies, technology manufacture, and site structure through time with the consistency of a single location (i.e. different cave occupations along the same coastal plain within PP ). Of greatest interest to SACP4 is the rarely sequenced t ransitional phases present at PP5 6 ; t he site spans MIS stages 5 4 3, representing interglacial glacial interglacial transition s (Karkanas et al . 2015, Wilkins et al . 2017). This enables research to be conducted through major climatic shifts on a plethora of different questions by allowing for the isolation of climatic factors on behavioral change s ( e.g. Karkanas et al. 2017, Wilkinson et al . 2017) .

PAGE 24

13 Figure 1 . Location of Pinnacle Point relative to the coast. Sites PP13B and PP5 6 are also shown in relation to each other. Photo published in Karkanas et al. 2015. Excavations at PP5 6 were closed in February 2017, ending the data collection phase at the site (Wi lkins et al . 2017). Throughout the work undertaken by SACP4, excavation techniques were devised station plotting, field stratigraphy, 3D GIS of plotted finds and lenses, and micromorphology with high re solution OSL [optically stimulated luminescence] (Brown et al. 2012:8). The system of excavation has remained consistent, enabling continuity across field seasons , excavators, and within the site. Marine Isotope Stage 5 4 Transition at PP5 6 Marin e Isotope States were identified and described by Martinson et al. ( 1987 ) , based on deep sea sediment core analysi s . MIS 5 is subdivided into five subcategories (a

PAGE 25

14 e) and spans 130 80 ka; however, only MIS 5a is relevant to this research and dates to 82 ka (Lisiecki and Raymo 2005). In contrast, MIS 4 consists of just one stage, d at ing to 71 ka ( Lisiecki and Raymo 2005 ). Three StratAggs account for the anthropogenically modified layers dated to the MIS 5 4 transitional period: the LBSR, t he ALBS, and the SADBS (See Figure 2 ). The faunal remains from these layers account for approximately 21,000 individual specimens (33. 8% of the total faunal assemblage at PP5 6 ) . The MIS 5 4 transition at PP5 6 has been characterized in four published reports : speleothem analysis (Bar Matthews et al. 2010 , Braun et al. 201 8 ), l ithic technol ogy a nd raw material sourcing (Wilkins et al. 2017) , sediment microfac ies (Karkanas et al . 2015) , and phytolith analysis (Esteban 201 9 ) . Each StratAgg will be characterized in the following sections. Figure 2 . Lateral view of site, facing east of StratAggs from PP5 6, with the relevant sections highlighted. Figure created by Erich Fisher ( Arizona State University ).

PAGE 26

15 L ight Brown Sand and Roofspall (LBSR ) The LBSR represents little sedimentary or occupational change throughout the 4.5 meter deposit (Karkanas et al. 2015). The layer is situated at t he lower end of the site and was one of the first StratAggs to be identified at PP5 6. The weighted mean OSL date for this layer is 81 ± 4, placing it in firmly within interglacial MIS 5a. The LBSR is characterized by relatively thin lenses of combustion features , moderate mollusk remains, and frequent roofspall inclusions . The sediment in the LBSR is fine grained, with minimal trampling evident from the intact condition of the hearths (Karkanas et al. 2015 :13 ). Study of the se diment shows water movement along the midline of the cave likely disturbed the LBSR sediment, result ing in modest material loss (Karkanas et al. 2015 :14 ) . This sediment loss and subsequent compaction near the center of the cave resulted in a shallow depres sion throughout the LBSR, though this is likely unrelated to human occupation (Karkanas et al. 2015 :14 ). Based on micromorphological study of the layer, occupation was characterized b y short, small group occupations evidenced by single incident hearths (Ka rkanas et al . 2015 :14 ). This might indicate that the modern humans roaming the landscape were either relatively few or were moving in small bands. Mollusk shells scattered intermittently throughout the deposit indicate exploitation of the coast was occurring with some regularity but may not have represented a reliable source of nutrients. During this time, the ocean was close to the cave , not much mo re distant than it is today . The shore averaged 1.1 kilometer , settling at a maximum distance of 1.6 km and a minimum distance of 0.8 km (km ) ( Fisher et al . 2010 , Karkanas et al. 2015 ) .

PAGE 27

16 This distance places the coast well within the 6 10 km daily foragin g radius for hunter gatherers ( Kelly 1995 ). Thus, the coast was easily exploit able during this time, though the lack of dense shell middens points to a lower reliance on shellfish than terrestrial fauna during this time (Marean 2010c, Thompson 2010). Duri ng the LBSR, the environment was likely moderate , with little dramatic climatic change through seasons . Analysis of phytoliths from the LBSR indicate the environment was a fynbos r enosterveld mosaic consisting primarily of C3 plants with C4 plants present in lesser proportions (Esteban et al. 2019). This is further substantiated by the speleothem record, which indicated MIS 5 was characterized by C3 vegetation and frequent, year round rains (Braun et al . 201 8 . ) These environmental indicators suggest fauna would be similar to those species inhabiting the area today . Focus on the LBSR lithic assemblage has not been robust to date (Wilkins et al. 2017). Blades, flakes, shatter, retouched flakes, cores, and hammers were all noted from the assemblage, though br oad trends have not yet been discussed. Quartzite collected from the nearby coast is the most common raw material used during the LBSR (Wilkins et al. 2017). During the LBSR , when the coast was located nearby, quartzite cobbles would have been routinely re freshed by ocean currents, thus providing a consistent source of good quality knapping materials within 10 km of the site (Wilkins et al. 2017). Brown et al . ( 2012) suggested raw material frequency would correlate to coastal distance; a hypothesis that has proved correct during the LBSR (Wilkins et al. 2017). Relative to the MIS 4, lithic artifacts are larger in size, suggesting less intense processing or termin ation at earlier reduction stages (Wilkins et al. 2017).

PAGE 28

17 Ashy Light Brown Sand (ALBS) The ALBS immediately overlies the LBSR and consists of 0.8 meter of fine grained beige sand that slopes slightly towards the inside of the cave . Th is layer be gin s after a significant roof fall event that caps the LBSR and represents a major boundary between roofspall rich layers and aeolian deposits at PP5 6 (Wilkins et al. 2017). The ALBS consists of heavy lenses of human occupation in an ashy sand matrix with frequent s hell middens (defined as sediment supported by shell fragments) (Karkanas et al. 2015, Wilkins et al. 2017). The mean weighted age of this layer is 72 ± 3 ka, placing it on the boundary of the transition from MIS 5 to MIS 4. Following previous analyses fro m PP5 6, though the mean weighted age of this level is within the transitional period it is grouped in MIS 5 (per Karkanas et al. 2015 and Wilkins et al. 2017). During the ALBS , the ocean receded to an average distance of 10 .6 kilo meters (km) from the cav e and ranged from a minimum distance of 1.7 km to a maximum of 22.0 km (Fisher et al. 2010, Karkanas et al. 2017). The average distance is feasibly within the daily foraging radius, though very near its limits, while the maximum is well outside. This fact i s interesting, giving the greater proportions of mollusk shells present at the site, which appear to signal a higher reliance on coastal resources than in the LBSR. This may indicate a preference for the reliable exploitation of ocean resources over terres trial , a theory posited by Marean (2016) . Speleothem analysis of nearby caves indicates the environment would have remained mosaic, with a slight increase in summer rains (Braun et al. 2019). This study also indicated an increase in C4 grasses (Braun et a l. 2019). Phytolith analyses concur with

PAGE 29

18 these results and further indicate the increase in C4 grasses was modest (Esteban et al. 2019). An abundan ce of C4 grasses should have provided an attractive ecosystem for M atthews et al . 2010: 2143) , though th e small increase in C4 plants illustrated by these recent studies (Braun et al. 2018, Esteban et al. 2019) may not present a great enough environmental change to support larger fauna . The retreat of the coast and concom itant presence of more abundant grassland on the exposed PAP during the ALBS suggests the environment would have been characterized by dense terrestrial fauna for exploitation (Klein 1983 , Copeland et al. 2016, Ho dgk ins et al. in press ) . Recent analysis of strontium isotopes finds th at the ungulates occupying the grassy PAP remained on the plain (Copeland et al. 2016) and thus would have served as a reliable source of prey for AMH during the ALBS. L arge hearths created in short intervals within the ALBS in dicate that occupation of the cave bec a me denser and more frequent than in the LBSR (Karkanas et al . 2015). This may indicate increased sed entism or larger group sizes during the glacial cycle (Karkanas et al . 2015: 19). As yet, the lithic sample f ro m the ALBS is too small to provide insights into the lithic techno typologies , though a few preliminary analyses are available. Wilkins et al. (2 017) noted an absence of backed microliths and bifaces, and very few blades which are typical of sites dated to this time period. This study found the majority of lithic artifacts from the ALBS are comprised of quartzite the nearest source of which is th e coast. Although the coast was more distant than in the LBSR, the inhabitants of PP5 6 still relied on quartzite as their main source of raw material (Wilkins

PAGE 30

19 et al. 2017). This suggests that the exposed PAP either did not expose other raw materials or co ntinued to generate quality quartzite within exploitation distance from the site (Wilkins et al. 2017). Lithics from this level are smaller than in levels occurring in MIS 5, perhaps suggesting later reduction stages and/or more intense reduction during MI S 4 (Wilkins et al. 2017). These lithic findings, though preliminary, suggest the occupants at PP5 6 were attempting to maximize the raw materials sourced from the distant coast. Shelly Ashy Dark Brown Sand ( SADBS ) The SADBS is composed of 0.7 m of aeolia n sand and consists of These hearths are not individually discernable, suggesting they were being made and replaced quickly by larger group occupations or more intensive occupations. Mollu sk remains are frequent, however, they do not form a shell supported midden as in the ALBS (Karkanas et al. 2015, Wilkins et al. 2017). The sediment of the SABDS is flat lying, likely due to either intentional flattening by the occupants or caused by conti nued usage of the site (Karkanas et al. 2015). During the SADBS, the coast averaged 15.1 km from the cave, ranging between a minimum distance of 2.5 km and a maximum of 23.8 km. Thu s, the coast was ( at time s) outside the average foraging radius range for hunter gatherers (i.e. 6 10 km, per Kelly 1996) , which might have necessitated a lower reliance on coastal resources (Bar Matthews et al . 2010, Fisher et al . 2010, Wilkins et al . 2017). This modeled coastal distance is interesting given the relative ly high proportions of shell present in the SADBS. This may indicate the increased reliance in intertidal resources visible in the shell middens of the ALBS persisted in the SADBS, despite a greater distance to the coast.

PAGE 31

20 Phytolith analysis of the SADBS i ndicated a decrease in C3 grasses and an increase in C4 plants, likely used to stoke the dense hearths occurring in this layer (Esteban et al. 2019). This finding suggests there was an increase in summer rains which indicate the environment was shrubby wit h low trees (Esteban et al. 2019). Esteban et al. (2019) additionally noted a spike in irregular and unidentifiable phytoliths, which may be explained by a change to more intensive dry wood collection undertaken to maintain the fires at PP5 6. The speleoth em study findings were the same, indicating more C4 grasses and an increase in summer rains (Braun et al. 2018). This suggests the environment would have supported larger bodied ungulates than the earlier levels, though the change to C4 grasses remains sma ll. Lithics analyzed from this level notably consist of backed microlithics and narrow blades (Wilkins et al. 2017). The SADBS is the first layer at PP5 6 to feature backed microlithic tools , and once they appear they persist for the remainder of the sequ ence ( Wilkins et al . 2017). This indicates a technological shift occurred during the SADBS that was adopted and maintained during subsequent occupations. Silcrete exploitation peaks in the SADBS and is contemporaneous with increases in C4 grasses (Wilkins et al. 2015). One possible explanation for this relationship is that C4 grasses were a preferred combustion agent used to heat treat lithic materials during the SADBS, thereby making silcrete a preferred material during this time (Wilkins et al . 2017). In addition, the increase d coastal distance may indicate that new quartzite cobbles were not being replenished, or silcrete sources were exposed on the open areas of the PAP.

PAGE 32

21 Table 1 provides summary data from transitional StratAggs during this undertaking. Broadly, the LBSR represents a sparse short term occupational level, characterized by exploitation of both coastal and terrestrial resources. The environment was largely mosaic, with little to no input of C4 grasses. As the sequence continues to the glaci al ALBS , occupation of the cave increases as does the apparent exploitation of coastal shell. Environmentally, there is a modest increase in C4 grasses, though the mosaic biome is largely intact. Finally, the SADBS represents dense, long term occupation of the cave where hearths are so tightly packed as to be indistinguishable. During the SADBS, the presence of C4 grasses continues to rise. These grasses and dry wood were likely heavily exploited for fire and hearth fuel throughout the sequence. As the area descended into a glacial, the distance between PP5 6 and the coast changed during each level; these changes are illustrated in Figure 3 . Table 1 . Coded specimens per StratAgg. StratAgg Sediment Description Weighted mean OSL age Faunal NISP Percent Coded LBSR (MIS 5 Glacial) Single occupation hearths, shell rich layers interspersed with roofspall dense layers with little human input 81 ± 4 8, 028 11.0 ALBS (MIS 4 Interglacial) Sandy sediment with frequent human occupation levels consisting of ashy shell middens 72 ± 3 4,314 37.0 SADBS (MIS 4 Interglacial) Dense, indistinguishable hearths with abundant shell, though not enough to form middens 71 ± 3 8,669 24.3 Total 13,860 17.4

PAGE 33

22 Figure 3 . Average, minimum, and maximum coastal distances for each StratAgg, as modeled by Fisher et al. 2010. Toba Eruption and Glacial Winter A recent study offers additional insight into the environmental circumstances occurring during the transition from MIS 5 4. Smith et al. (2018) discovered glass shards resulting from the Toba caldera eruption in the ALBS. The volcano, ( located approximatel y 8,937 km away in Sumatra, Indonesia) erupted approximately 74 ka (Chesner et al. 1991, Buhring and Sarnthein 2000) just before the onset of glacial MIS 4 (Rampino and Self 1992, Ambrose 1998, Smith et al. 2018). Some research suggests the eruption was both caused by the global transition to a glacial climate and helped push the global climate into MIS 4 by causing a volcanic winter (Rampino and Self 1992 , Ambrose 1998, Rampino and Ambrose 2000, Ambrose 200 3 ). Ice core evidence shows that the weather dro pped 3 5° C

PAGE 34

23 for several centuries following the eruption (Zielinski et al. 1996, Zielinski 2000) . Given this theory , areas impacted by the eruption likely would have seen a contraction of populations to refuge areas (Ambrose 1998, Smith et al. 2018). The presence of the shards within the ALBS and immediately underlying the SADBS suggest the effects of the eruption may have caused human populations around Pinnacle Point to seek the refuge of PP5 6 (Smith et al. 201 8 ). The proximity of the site to coastal and terrestrial resources, as well as the constant shelter would have provided an ideal location for affected populations to group together for survival during the onset of glacial MIS 4. Anatomically modern humans evolved in Africa approximately 3 00 kya ( Skoglund et al. 2017, Callaway 2017 ), then expanded to the Levant approximately 100 kya (Lewontin 1972, Relethford 1995, Ambrose 1998). Genetic studies have shown that prior to this major dispersal, one or more population bottlenecks occurred followed by pulses of population growth (Ambrose 1998, Fagundes et al. 2007, Behar et al. 2008 , Hammer et al . 1998, Harpending et al . 1998, Hawkes et al. 200 0 ) . The approximate date of this bottlenecking is debated, with no concordance between researchers. Ambrose (2005) proposed that one such bottleneck was caused by a glacial winter triggered by the eruption of the Toba volcano in Indonesia. Due to the lack of agreement on the date of these bottlenecks or the demonstrable effects on Africa n populations following the Toba eruption, it is difficult to Gathorne Hardy and Horcourt Smit h 2003). Comparing fauna between MIS stages has not yet been undertaken at Pinnacle Point. This project offers the opportunit y to test the reliance on and intensity of terrestrial fauna

PAGE 35

24 exploitation during the transitional phase. This study lay s the ground work for connections between human behavior and environment at PP5 6 and further inform s our understanding of early humans on the coast of South Africa.

PAGE 36

25 CHAPTER III METHODS The data available from PP5 6 is tightly controlled with carefully planned excavation techniques , outlined below . Additionally, the StratAggs have been tightly dated and described in three publications from SACP4 (Karkanas et al. 2015, Wilkins et al. 2017, Esteban et al. 2018). This ensure s temporal concordance of MIS dates at PP 5 6 while also providing comparative metrics for future studies. PP5 6 is an ideal site at which to understand the human response to changing environments because the geological sequence is continuous throughout the MIS 5 4 transition, suggesting the faun al remains will adequately represent the faunal exploitation strategies of the occupants in response to a changing environment . Understanding how humans subsisted across the MIS 5 4 transition at PP5 6 is undertaken by sampling the fauna from the period in question, subjecting it to a thorough zooarchaeological analysis, and visualizing the data using 3D GIS. Excavation Techniques at Pinnacle Point Excavation occurs by defining stratigraphic units (StratUnits) that are delineated by natural sediment change s (Brown et al . 2012, Oestmo and Marean 2014). Unique l ot n umbers are assigned to localize a StratUnit within a 50 cm x 50 cm quadrant excavated to the natural depth of the StratUnit or to a depth of 5 cm if the StratUnit is thick and vertically continuous ( Figure 2 ). The StratUnits are then grouped to a sub aggregate (SubAgg) of nearby stratigraphic units with similar sedimentation, inclusions, and anthropogenic input (Brown et al. 2012, Oestmo and Marean date) . StratAggs are

PAGE 37

26 preferred when discussing PP5 6 to retain consistency across publications. These StratAggs are used to discuss the depositional changes and time periods represented at the site and are named base d on sediment characteristics (e.g. Light Brown Sandy Roofspall [LBSR], Ashy Light Brown Sand [ALBS], Shelly Ashy Dark Brown Sand [SADBS] ). Upon closure of the site, 11 StratAggs were identified and named (further information is available in Karkanas et al . 2015 and Wilkins et al. 2017). All StratAggs are dated using OSL comprised of 65 single grain samples throughout the sequence (Karkanas et al . 2015, Wilkins et al . 2017). These dates were found to match blind tested Uranium Thorium dates of the caves, co nfirming the accuracy of the dates (Jacobs 2010, Karkanas et al. 2015, Wilkins et al. 2017) . All artifacts (regardless of size or type) are point located using a Total Station , assigned catalogue numbers during excavation, and input directly into a databas e using a barcode scanner (Marean 2015, personal communication) . E a ch excavated artifact is plotted in 3D space, assigned a unique identification number, and associated with its stratigraphic level , SubAgg, and StratAgg at the site (Brown et al. 2012, Oest mo and Marean 20 14). This allows each artifact to be associated not only with the level and sediment from which it was removed, but also allows the relocation of artifacts in space relative to the site and each other after excavation. This method is design ed to mitigate human errors that regularly occur in archaeology such as measurement inaccuracies and interpretive differences among excavators (Bernatchez and Marean 2011, Dibble et al. 2007, Oestmo and Marean 2014). By using 3D GIS to locate artifacts in space , artifacts are easily

PAGE 38

27 associated to dated sections (Brown et al. 2012 , Karkanas et al. 2015, Wilkins et al. 2017 ). In addition, logging 3D coordinates for artifacts preserves the relational data between the artifacts (i.e. the spatial relationship be tween cutmarked bones and stone tools). Figure 4 . Excavation underway at PP5 6. SubAggs are identified on the wall using tags, with StratAggs marked with larger tags. Photo facing east. Sampling MIS 5 4 Transitional Fauna A s discussed, t he PP5 6 excavation has been closed following thorough excavation . The excavation of 14 meters of sediment (Karkanas et al . 2015) yielded approximately 62,000 faunal remains ( SACP4 2017). For this study , fauna will only be analyzed from excav ation units that fall within the transition period ( ~ 81 71 ka). In fitting with the scope of this project, the upper portions of the LBSR nearest the MIS 5 4 transitional period will

PAGE 39

28 be sampled. Likewise, the entirety of the ALBS, and portions of the SADB S which fall closest to the transition will be sampled. The artifacts from PP5 6 are s tored and organized according to excavation units within each SubAgg. Samples of faunal material from each of the StratAggs were selected at random. This ensured no bias was introduced through the selection of more readily identifiable elements or specimens with relatively higher surface visibility. Zooarchaeological Analysis Fauna was analyzed by a suite of attributes, outlined below . Each of these attributes was then quantified and compared across the three StratAggs to determine broad differences in faunal processing at PP5 6 during the MIS 5 4 transition. All analysis was completed using a standardized Access database, and identifications were verified by expert zooa rchaeologists prior to inclusion in the dataset. The analysis took place over several seasons and by multiple train ed analysts, under the supervision of Drs. Jamie Hodgkins (University of Colorado Denver) and Curtis Marean (Arizona State University). Frag ment Size Fragment size was recorded for each specimen. The maximum length and width of each fragment was measured using digital calipers and recorded in millimeters. The minimum and maximum, upper and lower quartiles, and means for both the length and wid th were calculated and plotted to compare the range of fragment sizes between levels. These data provide information on the overall destruction of the bone, which may impact the appearance of identifiable attributes (e.g. anthropogenic or carnivore marks), and/or

PAGE 40

29 skeletal element (i.e. very small fragments may possess no features identifiable to element). This metric is also used to compare the preservation between levels; smaller fragments indicate greater levels of post depositional damage to the bones. Should fragments differ greatly in size between levels, this could skew the results of other analytical procedures such as surface modifications. Weathering Weathering stage was coded following Behrensmeyer system (1978), to determine how long the bones were exposed on the surface prior to burial (1978:161). Following an extensive study of taphonomic processes ( Behrensmeyer et al. 19 80), Behrensmeyer standardized the classification of weathering stages for bones to create a comparative system against which all zooarchaeological assemblages could be compared (Behrensmeyer 1978). The system is based on the highest level of weathering visible on the flat portions of bones (1978:152). The weathering stages exist on a continuum and are summarized below, as outlined by Behrensmeyer (1978:151): Stage 0: No cracking or flaking on bone surface. Bone is still greasy, with skin and muscle ligaments attached. Stage 1: Parallel cr acking o f surface begins. S kin and ligaments may or may not still be attached. Stage 2: Cracking and flaking of surface begins, with flakes attached to shaft on one end . Only remnant skin and ligaments are present, if any. Stage 3: Surface has a fibrous texture with the outer cortical layer worn away in patches. Tissues are rarely presen t. Weathering does not penetrate beyond 1 1 .5 mm. Stage 4: Surface is coarse ly fibrous and the bone splinters when moved. Weather ing extends to inner cavities. Stage 5: Bone is extremely fragile and may be splintered in situ. Spongy inner bone is expose d. Original shape of the bone is difficult or impossible to determine .

PAGE 41

30 This weathering rating system is widely used in the zooarchaeological literature and provides a standardized way of discussing the condition of archaeological assemblages. For this an alysis, the proportions of each weathering type will be compared across levels to determine whether differences in preservation exist between levels. This is an important piece of data to record because it may explain differences between levels which are n ot due to human activity, but rather differential preservation between levels (i.e. higher levels of heavy weathering indicates poor preservation of bones, while lower weathering indicates better preservation). Surface Visibility Coding for surface visibil ity was undertaken and assigned a value from 0 100% in 10% increments. Surface visibility records the degree to which the cortical surface is observable. Bone surface can be affected by a multitude of factors, the most common of which at Pinnacle Point a re calcium carbonate concretions and post depositional damage that results in the removal of cortical surface. Calcium carbonate concretions at PP5 6 are likely due to the heavy inclusion of mollusk shells, which are composed primarily of calcium carbonate (Andrus 2011, Claassen 1998). Over time, these shells can breakdown and leach into the sediments, where they can then build up on faunal remains. The utility of coding surface visibility is twofold: to account for differences in the appearance of surface modification for bones with less than 20% surface visibility, and to provide a metric for determining the surface condition of bones between StratAggs. Recording these metrics may provide explanations for differences between levels not due to anthropogenic or weathering input.

PAGE 42

31 Taxonomic Identification Though taxonomic differences may signal changes in available fauna, it is not possible to demonstrate these changes through the analysis of a single site; instead taxonomic representation and the presence of different portions of the animal body provide s information about what the occupants of the cave were choosing to transport and/or process at the cave. Taxonomic identification was undertaken at the smallest taxonomic unit possible (e.g., Genus, Family, Ord er, etc.). Body Size Body size is determined by the thickness of cortical bone , the curvature of the bon e, a nd the overall size of each element. Though determining the taxonomic identification of animals is impossible based on body size alone , it may offer insight into the size of fauna chosen for transport to the site by the occupants. It may also provide some insight into the type of fauna transported back to the site by the occupants. Because many bones in archaeological assemblages are not identifiable to species (Klein and Cruz Uribe 1984) , body size can give a rough approximation of the types of fauna being exploited. Body size classes were first developed by Brain (1981), as a means of quantifying the unidentifiable remains to a broad cat egory of species. Body sizes 1 6, representing a range of species, were used to provide data on the unidentifiable fauna remains from PP5 6. Following the analysis undertaken by Thompson (2010) on the fauna at PP13B, examples of South African species (a nd corresponding size) that fit each size category are presented in Table 2 . Every analyzed fragment was assigned a body size class, even if a determination of taxonomy wa s also made.

PAGE 43

32 Table 2 . Body size categories (Brain 1981) and example South African fauna (Thompson 2010). Body Size Weight Range (lb ) Example Fauna Body Size 1 5 50 Grysbok/Steenbok Body Size 2 50 250 Springbok Body Size 3 250 750 Wildebeest Body Size 4 750 2000 Eland Body Size 5 > 2000 Hippopotamus Body Size 6 > 6000 Elephant Burning Maximum burning stage was recorded for each fragment as unburned (0%), partially burned (<100%), fully burned (100%), partially calcined (<100%), or fully calcined (100%). Burning stage was determined by the color of the bone ( Cain 2005, Stiner et al. 1995). Degree of burning can indicate behavior . Stiner et al. (1995) determined that bones only becom e calcined if they contact the fire directly; this suggests partially or fully calcined faunal remains were used as fire fuel sometime after or as part of the discard process. Additionally, bones can become fully carbonized when buried below the hearth, su ggesting these bones are not necessarily attributable to cooking or refuse activities (Stiner et al. 1995). The frequencies of burned to unburned bone were compared and analyzed for differences between layers to determine whether a change in cooking or re fuse strategies occurred during the transition. Transportation Strategies Based on his work with the Nunamiut, Binford (1978) put forward a theory that the frequency of skeletal elements in an archaeological assemblage can inform archaeologists about the transport strategy employed by past populations. These transport strategies

PAGE 44

33 stra tegies, Binford (1978) calculated the general utility index (GUI) for each skeletal element by summing the nutrient availability of meat, marrow, and grease. This analysis determined that bones fell into two broad categories: high and low utility elements. Low utility bones are defined as bones that have little or no available meat, marrow, or fat (Binford 1978, Metcalf and Jones 1988). The cranium, vertebral column, and distal limbs (e.g. metapodials, tarsals, carpals) account for the lowest utility bones on an animal skeleton. In contrast, high utility bones are those which offer substantial available nutrition and are composed of the scapula, ribs, upper limbs (e.g. humerus, radius, ulna, femur, tibia, fibula), and pelvis of the animal (Binford 1978, Metc alfe and Jones 1988). In addition to the MUI, minimum number of elements (MNE) is calculated for each skeletal element. The MNE is then divided by the number of times that element appears in a skeleton; this results in the %MAU (normed minimal animal unit) (Binford 1978, 1984). The %MAU calculation is used to determine the actual frequency of each element transported back to the site. Binford then plotted these two calculations against each other (GUI, %MAU) and fitted best curves to the data (1978, 1984). As this methodology progressed, alterations to these calculations were made to mitigate common biases in archaeological assemblages. Metcalfe and Jones (1988) suggested a more readily calculated measure of bone utility, where the dry weight of the element is subtracted from the gross weight of the element. This calculation is then normed and results in the SFUI (standardized food utility index), a ranking of utility for each skeletal element (Metcalfe and Jones 1988). These calculations have displaced Binfo

PAGE 45

34 are now the standard indices used to plot against the %MAU. Studies on the differential survivability of bones based on density (Lam et al. 199 8 , Lyman 19 8 5) and carnivore selection (Marean and Frey 1997 , Marean and Assefa 1999, Faith et al. 2007) have since proliferated in zooarchaeology. These studies indicated that not all bones should be included in the determination of transport strategy; instead, only bones with high survivability and high nutrient yield should be cons idered (Faith and Gordon 2007). Standard methodology is now to assess the frequencies of the femur, tibia, metatarsal, humerus, radius, mandible, sk u ll, and metacarpals (Faith and Gordon 2007). For this research, both the SFUI and high survival skeletal el ements were used. Five transportation strategies result from this research: gourmet, bulk, unbiased, unconstrained , and reverse (Binford 1978, Faith and Gordon 2007). The gourmet strategy consists of only those skeletal elements that have the highest uti lity and very few or no low utility elements (Binford 1978, Faith and Gordon 2007). Sites which exhibit a bulk transport strategy consist of all bones, except those with the lowest utility (Binford 1978, Faith and Gordon 2007). Unbiased sites are composed of bones which have been transported back to the site in direct correlation to their utility (i.e. the highest utility bones represent the highest proportion, while the lowest utility have the lowest proportion) (Binford 1978, Faith and Gordon 2007). In th e unconstrained strategy, the entire animal is transported back to the site for consumption (i.e. no part of the animal is brought back in preference over another ) (Faith and Gordon 2007). This suggests that the fauna was either small enough to be transpor ted back in its entirety or the kill site was near enough to the cave that the journey back to PP5 6 would not have been a waste of energy resources

PAGE 46

35 ( Schoville and Otorola Castillo 2014) . In addition to these transport strategies, a reverse utility curve i s noted (Binford 1976, 1984, Metcalf and Jones 1988). In the earlier days of food utility curves, the reverse utility curve (where only the lowest utility bones are transported) was thought to indicate a scavenging resource procurement strategy (Binford 19 85, 1991; Blumenschine 198 8 , Mellars 1996). It was later established that these reverse curves are instead attributable to an analytical error in which long bone shaft fragments were excluded from MNE counts (Marean and Frey 1997). This skews the sample to wards dense carpals, tarsals, and cranial fragments with low utility but high survivability , because identifiable epiphyseal ends of long bones are more susceptible to carnivore scavenging (Marean and Frey Frey 1997, Marean and Frey 1997 ) and post depositi onal deterioration (Lam et al. 199 8 ). Therefore, reverse utility curves are indicative of analytical or sampling biases, not transportation strategies. Establishing which strategy the occupants of PP5 6 employed during each stage of the transition will inf orm our understanding of changin g resource procurement strategies. Breakage Analysis Breakage types of long bone shaft fragments indicate whether the bone was broken while fresh (green) or dry . Epiphyseal ends of bones are excluded from this analysis because they are known to fragment differently than shafts (Villa and Mahieu 1991:34). Bones fragmented while still green indicate they were broken in order to access the still available nutrients, while bones broken when dry indicate post depositional da mage un related to nutrient acquisition (Villa and Mahieu 1991). Villa and Mahieu (1991) were the first researchers to describe and identify the morphological differences between

PAGE 47

36 nutritive and non nutritive breakage by analyzing three known accumulation ass emblages of human remains. The results of this study concluded that the breakage created by nutrient acquisition versus post depositional damage could be distinguished through analysis of the fracture angle and fracture outline of bones (Villa and Mahieu 1 99 1 ). Fracture angle was previously known by researchers to display morphological differences between fresh and dry bones (Johnson 1983, Morlan 1984); however, the combination of angle and outline was unique to Villa and Mahieu (1991). Fracture angle is de 1991:34). A ngles are categorized as either right, oblique (defined as acute or obtuse), or right/oblique (e.g. irregular breakage). Fracture outline refers to the shape of the bone ends and are categorized as either transverse (straight) or v shaped/curved, and intermediate (i.e. irregular breaks with two or more outline types ) . Villa and Mahieu (1991) further grouped the fracture angle and outline together t o define nutritive versus non nutritive breakage types. C urved/v shaped in outline and obliquely angled breaks were determined to occur when a bone is broken in order to access meat, marrow and other nutrients (Villa and Mahieu 1991). In contrast, n on nut ritive breaks (transverse in outline, right angled breaks) occur when bones are dry and no longer offer nutrients, suggesting they were broken after their original deposition by factors like cave roof collapse, or sediment compaction by human/animal trampl ing (Villa and Mahieu 1991) . Angles and outlines belonging to intermediate or indeterminate categories were excluded from this analysis because they do not definitively fit into either category and may be the result of multiple processes. For

PAGE 48

37 each long bon e shaft fragment recorded at PP5 6, the fracture angle and outline were recorded, then each end was identified as either nutritive or non nutritive breakage. These frequencies were then compared to data published by Marean et al. (2000), who generated exp erimental observations of human and carnivore breakage patterns to create expected frequencies of nutritive and non nutritive breaks to which archaeological assemblages can be compared. By comparing the transitional levels to data generated experimentally, deviations from the norm can be investigated. A change in the amount of nutritive breakage across levels may indicate increasing or decreasing nutritional stress, while the difference from the expected non nutritive breakage may indicate the processing in tensity of the site broadly (as compared to extant populations). Non nutritive breakage differences between levels will indicate whether more or less post depositional damage was occurring, which may indicate changing site occupation strategies or preserva tion biases between levels. Surface Modifications Of primary interest to this research are the modifications made by humans at the site during the transition period. In order to understand the differences between layers, several anthropogenic attributes were coded and compared. Surface modifications (i.e. cut and percussion marks) were identified and photographed using incident lighting and a DinoLite digital microscope with up to 40x magnification (though typically a magnification of 20x or less was empl oyed). Each mark was photographed, logged, and confirmed by zooarchaeologists with advanced experience to validate the marks (per Blumenschine et al. 1996).

PAGE 49

38 Cutmarks The identification of cut marks created by humans and hominids has been crucial to zooarchaeology. Since 1981, descriptions of cutmark morphology have proliferated (Bunn 1981, Potts and Shipman 1981) , due largely to the direct indication of anthropogenic input on bones. While early zooarchaeologists recognized the morphology of cutmarks, there was much debate on whether these marks could be reliabl y distinguished from sediment compaction (Behrensmeyer et al. 1989, Fisher 199 5 ) or carnivore tooth marks (Blumenschine and Selvaggio 1988, 1991, Blumenschine et al. 1996 , Potts 1988, Potts and Shipman 1988). Even the equipment used to identify cutmarks was debated . S ome researchers (Shipman 1981, Shipman and Rose 1983) suggested only scanning electron microscopes (SEM) could provide powerful enough magnification to identify cutmarks, while other s (Blumenschine 1995, Blumenschine and Marean 1993, Blumenschine and Selvaggio 1988, 1991) suggested a hand lens and incident lighting were sufficient. Further debate regarding the conspicuousness of the marks also surrounded cutmark identification. Some a rgued that only obvious marks (those identifiable with the naked eye) should be considered in the total count of marks recorded on bone fragments, while others advocated that all marks ( regardless of clarity or location ) should be considered. A study condu cted by Blumenschine et al . ( 1996 ) tested these long argued points by blind testing known accumulator marks using low magnification and incident lighting. This experiment definitively showed that by following specific protocol for each mark type, surface m odifications could be accurately identified with 95% confidence. Additionally, this study advocated for the inclusion of all marks, regardless of prominence on the bone surface.

PAGE 50

39 Blumenschine et al. (199 6 ) proved th e accuracy of this method for identifying surface modifications and this is now the widely used technique in zooarchaeology . [and] v shaped [in] cross occurring on the cortical surface of bones ( Figure 5 )(Blumenschine et al . 1996:496). Marks created by different types of stone tools have slightly different morphologies. Bifaces produce close parallel cuts that are relatively curved and overlapp ing, while lithic flakes produce long, straight lines. Scrapers (often unifacially flaked) produce marks that fit both of these descriptions , depending on which side is facing the cortical surface (Hodgkins and Riel Salvatore 2008) . For the purposes of thi s study, differentiation between tool type for each cutmark is not reported. Figure 5 . Examples of cut marks from PP5 6, showing the relatively straight morphology and clustering of marks.

PAGE 51

40 Percussion Marks Hominids have broken open long bones to access the bone marrow inside since they first began hunting (Outram 2001 , Thompson et al. 2019 ). The fat rich nutrients from marrow reside in the medullary cavity of long bones and are accessed by percussing the bones open with a cobble in order to process the marrow (Binford 1978, Blumenschine and Selvaggio 1988, Outram 2001). Percussing bones to access nutrients has been observed across the globe, from the Arctic to the desert of Africa and every place between ( Binford 1978, et al. marrow an excellent source of calorie ric h nutrition for hunter gatherers, it serves a n umber of other purposes such as waterproofing (Binford 1978) and tanning skins (Levin and Potapov 1964). Percussion damage has been of interest to zooarchaeologists since the uantified by Johnson (1985) and Blumenschine and Selvaggio (1988) using actualistic research. Two types of percussion damage relevant to this research have been identified through these actualistic studies: microstriations and percussion notches (Blumensch ine 1995 , Blumenschine 19 9 5). Percussion microstriations are described as 1995: 29)], usually identified on the cortical surface opposite the striking surface. P ercussion notches (per C apaldo and Blumenschine 1994) are u shaped damage marks occurring on bone surfaces and are the result of hammerstone or anvil blows. Notches generally occur on the edges of bones, where the bone has split from the blow. Percussion marks have been reliably distinguished from other surface modifications (i.e. carnivore

PAGE 52

41 tooth marks, through experimental studies ( Capaldo and Blumenschine 1994, Pickering 2002, Gal a n et al . 200 9 ) . Analysis of Surface Modificat ions Comparisons on the frequency of both cutmarks and percussion marks were then made between levels . In order to cull the sample to only those bones with adequate surface visibility, analyzed specimens with less than 20% observable surface were removed f rom the statistical analysis. C ontrolling for surface visibility ensured evenness of the sample Figure 6 . Sample of percussion modification marks. Clockwise from top left: microstriations, conchoidal flake scar, flake with two medullary percussion notches, percussion notch on cortical surface. Examples from unpublished experimental data by author.

PAGE 53

42 across all three StratAggs. A chi square analysis on the percentage of bones with one or more cutmark or percussion mark was undertaken. Further, an ANCOVA tes t with a Poisson distribution was conducted to test whether a statistically significant difference exists between levels, controlling for fragment size ( the observable surface area of bone analyzed) between levels. Calculation of the surface area for each fragment was undertaken (per Hodgkins et al. 2016:6) . T he geometric mean serves as a proxy for surface area of flat objects. To test whether this method can be used on bone fragments from PP5 6 , ) circumference system was employed during analysis. Shaft fragment circumference is coded as (1) <50% total circumference (flat) , (2) >50% total circumference (somewhat curved) , and (3) 100% circumference (curved) . The majority (83.5% n=5,119) of all bone s coded from PP5 6 are fragmentary and coded as (1) <50% of the total circumference . T his indicates that fragments have lost most of their curvature and are primarily flat. Thus, calculation of the geometric mean operates as a proxy for surface area (Hodgk ins et al. 2016). If a significant difference is hoc test is employed to determine where the difference exists between levels within the transition. 3D GIS Visualization Once analysis of the fauna at PP5 6 w as completed, coded bones were visualized in relation to each other using ArcGIS. This allowed faunal remains to be v isually grouped b y coded characteristics . First, all coded bones across the StratAggs were mapped in relation to each other to visualize th e sample of bones through the sequence. This was undertaken in order to illustrate the physical relationship of sampled remains and to identify sections

PAGE 54

43 of the MIS 5 4 transition sequence that should be further explored in the future. Based on the results of the zooarchaeolog ical analysis, four visual representations were made to examine the relationship of combustion features, degree of burning, taxon, and anthropogenically modified remains. These visualizations were undertaken to better understand the la yout of the rockshelter and to potentially identify areas for future investigation.

PAGE 55

44 CHAPTER 4 RESULTS Fragment Size The maximum length and width of each fragment was measured for each StratAgg. Each level averages to the same approximate fragment size, in both length and width ( Figure 7 ). The average length of fragments is approximately 20 mm and the average width is approximately 10 mm, suggesting remains across the transit ion are highly fragmented , thus potentially obscur ing surface modification data and taxonomic identification . This suggests there was little change in the amount of post depositional damage occurring between levels; rock fall events, human and animal tramp ling, increased population size, or increased sedentism did not impact the faunal remains disproportionately. The range for length and widths of fragments from each level are also reported. The SADBS shows the largest range in fragment length, relative to the other levels. This indicates there are a few relatively longer fragments in the SADBS, suggesting greater variation in preservation during this time though it does not result in changes in the mean. This range, combined with a similar average, suggest s the SADBS may be affected by outliers.

PAGE 56

45 Figure 7 . Average length and width for each StratAgg (millimeters). Weathering (Behrensmeyer 1978). Stage 0 accoun ts for the majority of all weathering at PP5 6 ( Table 3 ). There is a small increase in the percentage of fauna exhibiting Stage 1 weathering in the ALBS, and this increase continues in the SADBS. The difference between these two stages is unlikely to obscure the bone surface or ability of analysts to diagnose surface modifications or identify remains. The remaining weathering stages account for less than five percent of th e total assemblage, suggesting faunal remains between StratAggs were not disproportionately affected by weathering. These data indicate bones likely spent little time exposed on the site surface before deposition (Behrensmeyer 1978), probably due to the pr otection provided by the rockshelter roof of PP5 6.

PAGE 57

46 Table 3 . Observed weather stages by StratAgg. Weathering Stage LBSR (n) ALBS (n) SADBS (n) Stage 0 97.4% (759) 84.6% (1,351) 87.1% (1,833) Stage 1 0.4% (3) 14.0% (223) 11.9% (251) Stage 2 1.5% (12) 0.8% (13) 0.9% (18) Stage 3 0.1% (1) 0.4% (7) 0.1% (18) Stage 4 0.5% (4) 0.3% (4) 0.0% (0) Surface Visibility Figure 8 shows the proportion of surface visibility for each level. For reporting, the surface visibility was grouped into four categories (i.e. less than 20 percent, 20 50 percent, 50 90 percent, a nd 100 percent visible). Approximately 32.5% of LBSR remains have full (100 percent) visibility. Fauna from the ALBS has the most obscured surfaces, with only 25% of remains exhibiting 100 percent surface visibility. The SADBS accounts for the least surfi cially obscured remains: 50% displayed complete visibility. Surfaces were obscured primarily by calcium carbonate build up which creates a hard crust on the cortical surface of bones. The high rates of partially or fully obscured bones in the ALBS indica tes a post depositional process occurring in this level is affecting the surface visibility recorded during this study. This may be explained by the shell middens that comprise much of the ALBS; calcium carbonate is the primary ingredient in shells and whe n ground (as through trampling or natural processes) or burned (through cooking or heating activities), the calcium carbonate may be released into the surrounding soil and can subsequently be deposited onto faunal remains during post deposition (Claassen 1 998). Moderate surface visibility (e.g. visibility ranging between 20 90%) remained constant

PAGE 58

47 through the transition and accounts for a combined total of ~25% of the remains from each level. Figure 8 . Surface visibility by StratAgg. Taxonomic Identification Specimens were coded to the smallest taxonomic category possible for each fragment. In many cases no determination could be made , due to the lack of diagnostic characteristics and fragmentary nature of the remains. These bones were assigned to the . I n each StratAgg , these account for the largest proportion of coded remains (LBSR 74.3%, ALBS 75.2%, SADBS 80.2% , Table 4 ) . This is not unusual for South African faunal assemblages of this age (Thompson 2010) and is likely attributable to poor preservation. Remains belonging to Family Bovidae ( e.g.

PAGE 59

48 wildebeest, g azelles, impala) and Family Testudinidae (i.e. tortoises) account for the highest proportion ( 5 3 .6 % of the total assemblage; n=1,057 ) of remains identifiable beyond class ( Figure 9 ) . An exhaustive list of identified fauna is available in ( Table 4 ), however, because they do not represent large proportions of the sample, they are not discussed further. Despite the changing climate and environment, the difference in identifiable taxa is negligible. While the proportions of Family Bovidae change between levels, this is likely attributable to identifiability of the remains. The proportion of tortoises (Family Testudinidae) decreases in the ALBS (15.2%) but remains similar between the LBSR (23.6%) and the SADBS (21.9%). This change might be related to either a change in resource procurement str ategies or the availability of tortoises at the initial onset of glacial MIS 4.

PAGE 60

49 Figure 9 . Proportion of identifiable taxa for each StratAg g . Taxa representing less than one percent of the assemblage are not reported in this figu re. Body Size Figure 10 summarizes the body size class of the remains recorded for each level. The LBSR is dominated by size 2 fauna (38.4%), while size 1 (16.7%) and size 3 (16.6%) comprise the remainder of the deter mined fauna. Approximately 27.5% of the LSBR assemblage is comprised of indeterminate size class fauna. The ALBS is similarly dominated by size 2 fauna (40.6%), with size 1 (17.8%) and 3 (4.9%) also represented. Indeterminate size fragments account for 35. 9% of the ALBS assemblage. The SADBS exhibits a slightly different patterning: fauna of indeterminate size compose s the highest proportion of bones in this level (44.6%), with size 2 representing 39.2%, size 1 accounting for 11.4%, and size class 3 at 3.5% . Little difference is observed between levels for size classes 1 and 2,

PAGE 61

50 suggesting the availability of these small mammals was not greatly disrupted by either the coastal retreat or the increasingly mosaic environment. Body size 3, however, show s a n unex pected and marked decrease in proportion as the climate shifts into a moderate glacial with a n increase in C4 grasses. Figure 10 . Relative frequencies of faunal body sizes through the transition.

PAGE 62

51 Table 4 . Comparison between levels of the NISP coded for each Family, Order, Class by Body Size. Each specimen was coded to the smallest unit of taxonomic unit. Body Size LBSR MIS 5 ALBS MIS 4 SADBS MIS 4 Terrestrial Mammal Indet. 116 236 633 0 0 6 47 1 54 135 118 2 224 428 415 3 92 81 45 Marine Mammal 2 0 0 2 3 1 0 1 Family Bovidae Indet. 14 6 8 1 6 3 1 1 8 2 50 1 52 276 3 30 1 7 1 7 4 0 0 1 Family Testudinidae (Tortoises) Indet. 8 14 26 0 0 2 4 1 52 52 50 2 1 1 0 Order Carnivora Indet. 0 1 0 1 1 3 2 Family Procaviidae (Hyrax) 2 0 3 1 Family Spheniscidae (Penguin) 2 0 1 0 Family Equidae (Horse/Donkey/Zebra) 4 1 0 0 Family Rhinocerotidae (Rhino) 5 0 1 0 Family Otariidae/Phocidae (Seal) 5 0 1 0 Order Primates (Non human Primates) 1 0 1 2 Class Aves (Birds) 1 0 3 0 2 0 0 2 NISP Total 654 1177 1567

PAGE 63

52 Burning Burned bones were analyzed according to the maximum burning stage exhibited on the bone (i.e. >50% calcined is coded as calcined). The level of burning and proportion of burned to unburned bones are reported in Figure 11 . There is a statistically significant difference in frequency of burned bones 64.97 , df =2, p<0 .001 , v= 0.145 Figure 11 ). T he LBSR contained 56.1% burned bone, the ALBS preserved 30.3% burned bon e, and the SAD BS exhibited 56.3 % . Although the strength of the difference between levels is a modest, given the age of the deposits the association is important. Because the significant change between StratAggs is related to the proportion of burned to unburned bones, n ot to the proportions of each level of burning, an additional examination on the proportions of each level of burning was also undertaken. This showed that the proportions of each level of burning (i.e. calcined, carbonized, etc.) remain unchanged between StratAggs , which suggests that the manner in which bones are being cook ed or discard ed was the same or similar for each leve l. Instead, only the amount of total burning is reduced during the ALBS.

PAGE 64

53 Figure 11 . Percentage burned material per StratAgg, including percent fully calcined, partially calcined, fully carbonized, and partially carbonized. Nutritive and Non nutritive Breakage Percentages of nutritive and non nutritive breakage patterns were plotted for each StratAgg and compared to experimental data from Marean et al. 2000 ( Figure 12 Figure 13 ). Non nutritive breakage for each level is reported in Figure 12 . All three levels exhibit higher percentages of non nutritive breakage than the experimental data sets , suggesting the bones underwent a high level of taphonomic damage. Approximately 20% of long bone shaft breaks are non nutritive in the LBSR. The ALBS features approximately 23% non nutritive breakage. The SADBS has the highest proportion of dry bone brea ks: approximately 33%.

PAGE 65

54 Nutritive breakage for each layer is reported in Figure 13 . All three StratAggs display fewer nutritive breaks than the experimental data. The LBSR h as the highest proportion of nutritive breaks (approximately 79%). The ALBS has nearly the same: approximately 72% of breaks were nutritive. The SADBS has the lowest percentage of nutritive breakage at roughly 55%. Numbers lower than the expected (Marean e t al. 2010) could indicate differing processing techniques or obscuring of nutritive breakage by the high levels of non nutritive breakage.

PAGE 66

55 Figure 12 . Non nutritive breakage for each StratAgg, compared to experimental data from Marean et al. 2000.

PAGE 67

56 Figure 13 . Nutritive breakage for each StratAgg, compared to experimental data from Marean et al. 2000.

PAGE 68

57 Food Transpor t Food Transport Utility curves were calculated for each StratAgg based on the %MAU and SFUI (Faith and Gordon 2007), these figures are reported in Table 5 . Distance weighted mean squares curve s were Food Transport Utility curves in Figure 14 (Binford 1978 , Faith and Gordon 2007 ). The results show the LBSR followed a bulk strategy , where all but the lowest utility bones are transported back to the site (Binford 1978, Faith and Gordon 2007). The ALBS tracks most c losely to the unbiased (following it nearly exactly), suggesting occupants of PP5 6 were transporting skeletal remains in direct proportion to their utility. T he SADBS closely aligns with the bulk strategy . The food utility lines for both the LBSR and the SADBS are brought down by lower than expected proportions of femora. This may be due to difficulties in assigning skeletal element because of the fragmentation of bones and extensive post depositional damage, though it may also be related to butchering str ategy. The low levels of femora present at PP5 6 may be explained by ethnographic study conducted by Binford (1978:54) who noted that Nuna mi butchering process and thus never transported these bones ba ck to site. A similar patterning of missing femora has been noted in other archaeological assemblages, suggesting this maybe be a consistent explanation applicable to numerous contexts ( Costamagno et al. 2006 , Hodgkins 2012).

PAGE 69

58 Table 5 . Minimum number of elements, MAU, SFUI and normed MAU values for each StratAgg, used to create Figure 2. Created using SFUI, as calculated by Metcalfe and Jones (1981). Element MNE L MNE R Side N/A MAU SFUI %MAU SADBS Cranial 2 1 9.1 27.9 Mandible 2 1 11.5 27.9 Humerus 1 1 1 36.8 27.9 Radius 1 5 3 25.8 85.7 Metacarpal 2 2 2 5.2 57.1 Femur 1 0.5 100 14.3 Tibia 5 2 3.5 62.8 100 ALBS Cranial 3 1.5 9.1 11.1 Mandible 3 1.5 11.5 11.1 Humerus 1 8 4.5 36.8 100 Radius 4 3 3.5 25.8 77.8 Metacarpal 4 2 5.2 44.4 Femur 2 1 100 22.2 Tibia 3 5 4 62.8 88.9 Metatarsal 1 3 2 37 44.4 LBSR Cranial N/A N/A 9 4.5 9.1 100 Mandible 1 0.5 11.5 11.1 Humerus 2 1 36.8 22.2 Femur 1 0.5 100 11.1 Tibia 2 1 1.5 62.8 33.3 Metatarsal 1 0.5 37 11.1

PAGE 70

59 Figure 14 . Distance weighted mean squares curve fitted to each StratAgg and plotted , Faith and Gordon 2007 ).

PAGE 71

60 Surface Modification Cut and percussion mark frequency was recorded for all levels. Only faunal remains exhibiting surface visibility greater than 20% were included in this analysis. The frequency of cutmarked bones through the transition decreases sign ificantly but modestly ( = 21.51 , df=2, p<0.0001 , v=0.106 ) from the LBSR (12.1%) to the ALBS (8.7%) to the SADBS (7.3%) . Percussion marks do not exhibit a significant difference between levels. Approximately 5.1% of the LBSR assemblage exhibits percussion damage. Roughly 3.8% of the ALBS has percussion damage, and the SADBS shows approximately 3.2%. An analysis of the number of cutmarks ( Figure 15 ) and percussion marks ( Figure 16 ) per fragment ( controlling for surface area, using the geometric mean ) wa s also undertaken to determine whether bones were being more intensely processed across StratAggs . The analysis of cutmarks per fragment shows a statistically significant difference between levels Figure 15 ). The p valu e for this test (p=0.049) is close to the boundary of significance; however, this may indicate a larger pattern of behavioral change. A Tukey s post hoc test was performed to determine where the significant difference is located in the transition. This sh owed the change in cutmarks per fragment is significant (p=0.016, df=1) between the ALBS and the SADBS, with cutmarks per fragment increasing 0.06 marks per fragment. The analysis on the number of percussion marks per fragment was not significant across le vels Figure 16 ).

PAGE 72

61 Figure 15 . The results of the AN C OVA with Poisson distribution to determine the least squares mean number of cut marks per fragmen t , controlled for surface area (calculated using the geometric mean). The error bars show the 95% confidence intervals for each StratAgg.

PAGE 73

62 Figure 16 . The results of the AN C OVA with Poisson distribution to determine the least squares mean number of percussion marks per fragmen t , controlled for surface area (calculated using the geometric mean). The error bars show the 95% confidence intervals for each StratAgg. 3D GIS The spatial relationship between artifacts was mapped using Arc Scene 10.3.1 software. First, a 3D visualization of the three different StratAggs was undertaken , and points that were wrongly input into the Total Station were removed . These were identified by th eir location outside excavated areas and verified against the zooarchaeological data. This visualization show ed that while the faunal remains analyzed account for a large proportion of the continuous transition, some gaps remain in the transitional data. W hile a complete analysis of the PP5 6 fauna is outside the scope of this project, these data suggest

PAGE 74

63 these findings may be strengthened with further analysis of missing portions of the sequence, as illustrated by the GIS data. The spatial relationship b etween the largest proportions of taxonomic identifications was undertaken ( Figure 17 ). For ease of visualization, only the two highest proportion of taxon were mapped : Family Bovidae and Family Testitudidae . This showed a normal distribution of Family Bovidae and Family Testitudidae, with identifiable remains clustering closely together. The LBSR shows a concentration of tortoise remains in the lower portion of the StratAgg, suggesting a relatively higher reliance on tortoise during MIS 5, which gives way to a more varied sample of both tortoises and Bovidae as the LBSR gives way to MIS 4. The GIS data also show that the tortoise remains from the ALBS are, in general, situated more closely to the LBSR while the later portion of the ALBS is dominated by Bovidae. This suggests that the lower proportion of tortoise remains in the ALBS is a function of the changing environment, which is later mitigated as MIS 4 we ars on in the SADBS, where tortoise and Bovidae remains are equally distributed. To examine the relationship of faunal remains to hearth features, the assemblage was then mapped by bones recorded from a combustion feature versus those coded from all other contexts ( Figure 18 ). This shows the concentration of hearth features (and bones either discarded in the fire or used as fuel) for each level. To see if the degree of bu rning would concord with the combustion feature data, the levels were mapped according to the highest observed level of burning ( Figure 20 ). Despite the location of hearths within the transition, burned bones are dispersed throughout the sequence. There is some aggregation of burned bone at the locations of the hearths; however, the location of burned

PAGE 75

64 bones is not confined to them. This sugg ests the bones were being discarded throughout the cave after consumption and do not appear to be either systematically used for fuel or discarded in known refuse piles. With regard to the lower proportion of burned versus unburned bone in the ALBS, the GI S data does not offer valuable explication. There is a concentration of unburned bone in the ALBS that appears to account of a large number of the coded specimens. This cluster is located outside of known hearth features and very near the onset of MIS 4, s uggesting the lack of preparation with fire might be the result of clima c tic change. The exact date of this grouping of unburned bones is not known, but it may represent an effect of the Toba Eruption (e.g. animals reacting to effects of the eruption seek shelter in the rockshelter ). Another possible explanation is that the bones were coded as unburned due to a coding error; whole lots were often undertaken by the same analyst therefore if an error was not corrected quickly it may have not been noted. This analytical error is unlikely as all coding was confirmed and overseen by a lead zooarchaeologist. Finally, the location of bone marked by cut or percussion marks were visualized across the transition. This shows an even distribution across all three StratAggs; the anthropogenically modified bones do not appear especially aggregated between the levels. Within each level, however, the bones are situated near or within the combustion features illustrated in Figure 18 . This supports the use of the cave by humans as a food processing and preparation site. This spatial relationship is an obvious one; bones percussed for marrow or cut to access meat were subsequently heated for consumption.

PAGE 76

65 Figure 17 . Analyzed fauna from the transition, identified by taxon .

PAGE 77

66 Figure 18 . Visualization of faunal remains located within combustion features versus those located in other contexts .

PAGE 78

67 Figure 19 . Burned bones, displayed by level of burning across the transition.

PAGE 79

68 Figure 20 . Anthropogenically modified bones shown in association with bones recovered from hearth contexts (in black).

PAGE 80

69 CHAPTER V D I SCUSSION AND CONCLUS ION The results from the zooarchaeological anal ysis are discussed within each StratAgg, then broad trends or changes are identified and discussed. LBSR Discussion The LBSR, a relatively static layer of moderate anthropogenic input dating to MIS 5, serves as the layer against which the MIS 4 levels can be compared to determine changes in behavior related to climate. During this level, the coast was very near the shor e (Fisher et al. 2010) with the environment typified by a fynbos r enosterveld mosaic ( Braun et al. 2018, Esteban et al. 2019). Though the intertidal zone was close and easily accessible, the LBSR sediments display only moderate exploitation of coastal reso urces such as mollusks (Karkanas et al. 2015, Wilkins et al. 2017). The faunal analysis suggests that during the LBSR, occupants were exploiting small bodied mammals, primarily under 250 kg . According to the fauna identifiable to taxon and body size, these animals are comprised primarily of Bovidae (grysbok/steenbok/springbok) and Testudinidae (tortoise). A majority of fragments from this level were not assignable to taxon (66.8% of the total LBSR assemblage) or body size classes (27.5% indeterminate), due to fragmentation and lack of identifiable features. Fragment size and non nutritive breakage rates suggest preservation in the LBSR is poor, though the weathering rates show bones likely were not exposed on the surface long before burial. Burning data indi cates approximately 56.1% of the bones from the LBSR were heated ; however, only 9.3% of the bones are calcined to some degree. These calcined remains were in direct contact with the hearth (Stiner et al. 1995), possibly

PAGE 81

70 indicating their use as fuel. The re maining heated fragments (46.8%) are likely the result of cooking or processing of nutrients. Unburned bones do not necessarily indicate a lack of processing, as burning signatures on bone can be obscured by tissue. During the LBSR, humans at PP5 6 were em ploying a bulk transport strategy , suggesting PP5 6 occupant s during this time were choosing to transport whole or nearly whole carcasses back to the site . Schoville and Otorola Castillo (2014) have demonstrated that transport strategies calculated with as semblages comprised primarily of body sizes 1 and 2 are not reliable indicators of distance to kill sites, as proposed and used by Binford. Instead, this transport strategy is a better indication of the decision by PP5 6 occupants to transport whole carcas ses back to the site for processing. The number of nutritional breaks and percentage of cutmarked bones (12.1%) further indicate fauna were being processed more heavily than in later levels. A L B S Discussion The ALBS is dated to the onset of MIS 4 and con tains relatively frequent hearths interspersed with dense shell middens (Karkanas et al. 2017). During the ALBS, the coast was located a moderate distance away (Fisher et al. 2015), and the environment shows an increase in C4 grasses though the impact of t his increase has not been well demonstrated (Braun et al. 2018, Esteban et al. 2019). It is also during the ALBS that the Toba eruption occurs in Sumatra (Chesner et al. 1991, Buhring and Sarnthein 2000), potentially plunging the area into more extreme gla cial conditions (Ambrose 1998, Ambrose 2000, Ambrose 200 0 , Rampino and Self 1992 ).

PAGE 82

71 Though fragment size remains constant in the ALBS (from the LBSR), non nutritive breakage, weathering, and surface obstruction increase, possibly signaling poorer preservation than in the LBSR. Surface visibility may be affected by leeching of calcium carbonate from shells into the sediment (Claassen 1990) . Faunal analysis shows nearly the same amount of remains identifiable to taxon as in the LBSR; however, proport ions of Testudinidae decrease to 15.2%. This might indicate either an availability disruption or procurement difficulty during the ALBS. None of the Testudinidae specimens were identifiable to species; however, faunal analysis conducted at nearby sites sug gests the sample would include the angulate tortoise ( Chersina angulata ), and the pancake tortoise ( Homopus arelatus ) ( Henshilwood et al. 2001, Halkett et al. 2003, Klein and Cruz Uribe 2000, Klein et al. 2004, Thompson 2010). These species of tortoise are common to modern fynbos and thrive in temperate, coastal , semi arid environments (Branch 1984, Thompson 2010). This suggests the change in Testudinidae during the ALBS is likely attributable to a decrease in tortoise availability during the initial coo ling of MIS 4 . Additionally, the ALBS shows an increase in cutmarks per fragment (0.31 per fragment) and nutritive breakage s from the LBSR, both of which indicate increasing processing intensity (Hodgkins et al. 2016). This seems to track well with the ch ange to an unbiased food transport strategy, where occupants transport food back in direct relation to its nutritive value (Binford 1978, Faith and Gordon 2007). This level has the lowest proportion of total burned bone across the transition (30.3%), which may be partially explained by the decreased surface visibility. However, the rate of decrease in burned bone cannot be explained entirely by obscured surfaces but may instead indicat e nutritional stress. The lower proportion of burned bone

PAGE 83

72 could illustrat e less time spent cooking or processing the meat before consumption ; d ecreasing the cooking time of meat to the shortest time is a possible tactic for minimizing the time between hunting and nutrient acquisition. Another explanation may be related to the r elationship between AMH and fire European assemblages show that evidence of fire during glacial periods is low (Sandgathe et al. 2011a, Sandgathe et al. 2011b). Sandgathe et al. (2011a) propose that this is due to an opportunistic (rather than controlled) use of fire. The lower proportion of burned bone seems to roughly accord with these data, possibly suggesting a more limited availability of natural fire during the initial onset of MIS 4 . In general, the ALBS might be marked by increasing nutritional str ess coupled with decreased preservation of faunal remains. SADBS Discussion The SADBS is squarely situated in MIS 4, when the coast has retreated to an average distance of 15.1 km (Fisher et al. 201 0 ). The occupation of the cave is denser, evidence by thi ck hearths and increased sediment trampling (Karkanas et al. 2017). As with the ALBS, the environment shows a continuing increase in C4 plants and shrubby vegetation (Braun et al. 2018, Esteban et al. 2019). Though fragment size, weathering, and surface vi sibility in the SADBS closely tracks with that of the LBSR, non nutritive breakage is at the highest level throughout the transition. This is likely due to the increased trampling of sediments caused by longer occupation times or larger group sizes (Karkan as et al. 2017). In many ways, the SADBS is very similar to the LBSR; tax a , body size, cut marks per fragment, proportion of burned bones, and transport strategy are much the same. However, the proportion of nutritive breakage in the SADBS is the highest a cross the transition. A

PAGE 84

73 possible explanation for this increase could be the proposed larger population sizes of the SADBS; more occupants would require all possible nutrients to be extracted from fauna. The decrease in cut marks per fragment from the ALBS to the SADBS, however, seems to indicate an increase in nutrient acquisition only as it relates to shaft breakage. Three explanations may account for the decrease in cutmarks per fragment. The advent of microlithic technology occurring the SADB S (Wilkins et al. 2017) may have affected the way meat was removed from bones such that cut marks on bone are not as frequently left behind as in earlier levels. Alternatively, a socially or culturally determined change in butchering technique could account for the decrease in marks. Finally, there may have been an increased reliance on marrow extraction but not on flesh; nutritive breakage may have increased to access the fatty, nutrient dense marrow while no change in defleshing occurred. It is difficult to determi ne which of these explanations best fits the scenario at PP5 6, though it is clear a change in behavior did occur during the SADBS. Conclusions The results of the zooarchaeological analysis show that there was likely some behavioral mitigation occurring d uring the transition from interglacial to glacial climate at PP5 6. The data show several trends relating to subsistence strategies during the change from interglacial to glacial climate the key differences are summarized in Table 6 and visualized in Figure 21 . There is an overall trending decre ase throughout the transition to a lower proportion of body size 3 fauna, suggesting a change in reliance on larger fauna by the occupants. This is counter to the changing environment that would have produced more C4 grasses; a change which, coupled with a more exposed Agulhas plain, should have

PAGE 85

74 supported larger bodied mammals. Though teeth were not included in this analysis, they have been recovered and identified as large bodied ungulates through the transition, suggesting the occupants of PP5 6 were expl oiting body sizes 3 6. This is likely explained, in part, by the high level of fragmentation caused by increased trampling and post depositional damage of faunal remains through the transition, which obscure the identification of larger body sizes. Addit ionally, there is a steady decrease in both nutritive breakage and percentage of cut marked bones, suggesting less nutritional stress as the transition carried on. Backed microlith tools, with a superior ability to hunt large game may explain the decrease in cut marked bone, either through an increased success rate of hunters or by changing the way butchering is undertaken (i.e. new strategies could result in fewer cuts on bone surfaces). Though the overall proportion of cut marked bones decreases steadily through time, the ALBS shows a significant increase in number of cutmarks per fragment (controlled for fragment size), relative to both the LBSR and SADBS. This indicates fragments were being more intensively processed during the ALBS, relative to both th e LBSR and SADBS. As Pinnacle Point became more mosaic, the PP5 6 occupants may have experienced an initial increase in nutritional stress that was mitigated during the ALBS. The occupants appear to have returned to the same strategies used during MIS 5 d uring the SADBS, either by changing subsistence strategies or a return to steadier climatic conditions. Increased nutritional stress is also visible in the total percentage of burned bone, which experiences a statistically significant decrease in the ALBS. Cooking time may have been abbreviated

PAGE 86

75 of the glacial period may have altered the patterning of burning during the ALBS. Table 6 . Summ ary table of the key differences between StratAggs during the MIS 5 4 transition. LBSR ALBS SADBS MIS 5 (81 ± 4) MIS 5 (71 ± 3) MIS 5 (72 ± 3) Environment Primarily C3 vegetation with bimodal rain Increase in C4 vegetation and summer rains Highest proportion of C4 grasses and summer rains Shell exploitation Thin lenses of shell exploitation Dense, shell supported matrices Abundant shell but less than in ALBS Average Distance to Coast (Fisher et al. 2010) 1.1 km 10.5 km 15.1 km Body Size 3 16% 5% 3% 100% Surface Visibility 35% 25% 50% Non nutritive Breakage 22% 24% 32% Nutritive Breakage 78% 74% 55% Cutmarked bone 12.10% 0.1 7.30% Cutmarks per fragment 0.26 0.32 (Statistically significant difference) 0.27 Heated Bone 56.1% 30.0% (Statistically significant drop in overall proportion of burned bones) 56.3% Transport Strategy Bulk Unbiased Bulk This analysis suggests humans were transporting food back to PP5 6 differently across the transition. Similar transport strategies were being undertaken in both the LBSR and the SABDS, while during the ALBS occupants are exhibiting preferential transport o f higher utility elements. This may suggest an energy saving measure intended to reduce

PAGE 87

76 expended energy on the lowest utility elements, possibly to mitigate changing fauna availability or increased nutritional stress (Schoville and Otorola Castillo 2014). Based on the Food Utility Curves, the data show there may have been at least some change in the ability of the occupants to access nutrients, causing modified behavior at the onset of MIS 4. Figure 21 . Summary figure of key diffe rences between StratAggs. Dashed lines indicate usage of the secondary axis. T he most tumultuous time of change during the transition occurs in the ALBS, where the environment was likely the most unpredictable. This accords well with the proposed glacial winter spurned by the Toba Caldera eruption (Ambrose 1998, 2000). The subsistence behaviors and environmental indicators revealed during the ALBS suggest the glacial winter hypothesis should be further investigated. T aken in sum, the results of this study indicate there was a temporary change in subsistence strategies occurring at the

PAGE 88

77 onset of glacial MIS 4, followed by a return to the behaviors of MIS 5 as the climate settled into the glacial period. The implications of these findings are, as yet, prelimi nary but suggest the population continuously occupying PP5 6 was sufficiently behaviorally elastic as to withstand environmental changes related to resource availability. There is no evidence for gaps in the occupation at PP5 6 (Karkanas et al. 2015, Wilki ns et al. 2017), suggesting one group or related groups were occupying the cave, rather than representing unlinked bands of AMH. Further work is required at this site before meaningful conclusions can be drawn on the multitude of ways that humans at PP5 6 modified their behavior. Despite distant coastlines, changing flora, and climate, the population at this site was not diminished. These results indicate humans on the southern coast of South Africa had the cognitive ability to adapt their resource procure ment (e.g. transport strategies), cooking (e.g. burning), and nutrient acquisition (e.g. increasing nutritive breakage and cut marks) to an environment of shifting resource availability. This suggests this group of early anatomically modern humans possess all of the necessary cognition and behavioral elasticity to survive varied climates and niches. While this study alone cannot prove AMH possessed superior adaptability, it can serve as a useful case study of human resilience in the face of rapidly changing climate. The adaptability of AMH, as demonstrated by this thesis, can be compared to other sites, time periods, and hominid occupations to further

PAGE 89

78 REFERENCES Ambrose, S.H., 1998. Late Pleistocene human population bottlenecks, volcanic winter, and differentiation of modern humans. Journal of human evolution , 34 (6), pp.623 651. Ambrose, S.H., 2000. Volcanic winter in the Garden of Eden: the Toba supereruption and the late Pleistocene human population crash. Volcanic hazards and disasters in human antiquity , 345 , p.71. Ambrose, S.H., 2003. Did the super eruption of Toba cause a human population bottleneck? Reply to Gathorne Hardy and Harcourt Smith. Journal of Human Evoluti on , 45 (3), pp.231 237. Andrus, C.F.T., 2011. Shell midden sclerochronology. Quaternary Science Reviews , 30 (21 22), pp.2892 2905. Bar Matthews, M., Marean, C.W., Jacobs, Z., Karkanas, P., Fisher, E.C., Herries, A.I., Brown, K., Williams, H.M., Bernatchez, J., Ayalon, A. and Nilssen, P.J., 2010. A high resolution and continuous isotopic speleothem record of paleoclimate and paleoenvironment from 90 to 53 ka from Pinnacle Point on the south coast of South Africa. Quaternary Science Reviews , 29 (17), pp.2131 2 145. Bernatchez, J. and Marean, C.W., 2011. Total station archaeology and the use of digital photography. SAA Archaeological Record , 11 (3), pp.16 21. Behar, D.M., Villems, R., Soodyall, H., Blue Smith, J., Pereira, L., Metspalu, E., Scozzari, R., Makkan, H., Tzur, S., Comas, D. and Bertranpetit, J., 2008. The dawn of human matrilineal diversity. The American Journal of Human Genetics , 82 (5), pp.1130 1140. Behrensmeyer, A.K., 1978. Taphonomic and ecologic information from bone weathering. Paleobiology , 4 ( 2), pp.150 162. Behrensmeyer, A.K., Dechant Boaz, D.E. and Hill, A.P., 1980. The recent bones of Amboseli Park, Kenya, in relation to East African paleoecology. Fossils in the making: Vertebrate taphonomy and paleoecology , pp.72 92. Bergh, N.G. , Verboom, G.A. , Rouget, M. , Cowling, R.M. , 2014 . Vegetation types of the Greater Cape Floristic Region . In: Allsop, N., Colville, J.F., Verboom, G.A. (Eds.), Fynbos: Ecology, Evolution, and Conservation of a Megadiverse Region . Oxford University Press , Oxfo rd , pp. 1 25 .

PAGE 90

79 Binford, L.R., 1978. Nunamiut: Ethnoarchaeology . New York: Academic Press. Binford, L.R., 1962. Archaeology as anthropology. American antiquity , 28 (2), pp.217 225. Binford, L.R., 1964. A consideration of archaeological research design. American antiquity , 29 (4), pp.425 441. Binford, L.R., 1965. Archaeological Systematics and the Study of Culture Process 1. American antiquity , 31 (2Part1), pp.203 210. Binford, L.R., 1989. Isolating the transition to cultural adaptations: an organizationa l approach. The emergence of modern humans: biocultural adaptations in the later Pleistocene , pp.18 41. Binford, L.R., 1991. When the going gets tough, the tough get going: Nunamiut local groups, camping patterns and economic organization. Ethnoarchaeolog ical Approaches to Mobile Campsites. International Monographs in Prehistory, Ann Arbor , pp.25 137. Blumenschine, R.J., 1988. An experimental model of the timing of hominid and carnivore influence on archaeological bone assemblages. Journal of Archaeologic al Science , 15 (5), pp.483 502. Blumenschine, R.J., 1995. Percussion marks, tooth marks, and experimental determinations of the timing of hominid and carnivore access to long bones at FLK Zinjanthropus, Olduvai Gorge, Tanzania. Journal of Human Evolution , 29 (1), pp.21 51. Blumenschine and Marean 1993 Blumenschine, R.J. and Selvaggio, M.M., 1988. Percussion marks on bone surfaces as a new diagnostic of hominid behaviour. Nature , 333 (6175), pp.763 765. Blumenschine, R.J. and Selvaggio, M.M., 1991. On the mar ks of marrow bone processing by hammerstones and hyenas: their anatomical patterning and archaeological implications. Cultural Beginnings: Approaches to understanding early hominid life ways in the African savanna. Bonn , pp.17 32. Blumenschine, R.J., Mare an, C.W. and Capaldo, S.D., 1996. Blind tests of inter analyst correspondence and accuracy in the identification of cut marks, percussion marks, and carnivore tooth marks on bone surfaces. Journal of Archaeological Science , 23 (4), pp.493 507.

PAGE 91

80 Braun, K., Bar Matthews, M., Matthews, A., Ayalon, A., Cowling, R.M., Karkanas, P., Fisher, E.C., Dyez, K., Zilberman, T. and Marean, C.W., 2019. Late Pleistocene records of speleothem stable isotopic compositions from Pinnacle Point on the South African south coast. Quaternary Research , 91 (1), pp.265 288. Brain, C.K., 1981. The hunters or the hunted? . University of Chicago Press. Brown, K.S., Marean, C.W., Jacobs, Z., Schoville, B.J., Oestmo, S., Fisher, E.C., Bernatchez, J., Karkanas, P. and Matthews, T., 2012. An early and enduring advanced technology originating 71,000 years ago in South Africa. Nature , 491 (7425), pp.590 593. Buhring and Sarnthein 2000 Bunn, H.T. 1981. Archaeological evidence for meat eating by Plio Pleistocene hominids from Koobi Fora and Olduv ai Gorge. Nature 291: 574 577. Bunn, H.T., 1983. Comparative analysis of modern bone assemblages from a San hunter gatherer camp in the Kalahari Desert, Botsuwana, and from a spotted hyena den near Nairobi, Kenya. Animal and Archaeology, 1. Hunters and Th eir Prey; BAR International Series , 163 , pp.141 148. Cain, C.R., 2005. Using burned animal bone to look at Middle Stone Age occupation and behavior. Journal of Archaeological Science , 32 (6), pp.873 884. Capaldo, S.D. and Blumenschine, R.J., 1994. A quantitative diagnosis of notches made by hammerstone percussion and carnivore gnawing on bovid long bones. American Antiquity , 59 (4), pp.724 748. Chesner, C., Rose, W.I., Deino, A.L., Drake, R. and Westgate, J.A., 1991. Eruptive history of Earth's largest Quaternary caldera (Toba, Indonesia) clarified. Geology , 19 (3), pp.200 203. Claassen, C., 1998. Shells . Cambridge University Press. Cowling, R.M. and Lombard, A.T., 2002. Heterogeneity, speciation/extinction history and climate: explaining regional plan t diversity patterns in the Cape Floristic Region. Diversity and Distributions , 8 (3), pp.163 179. Dibble, Harold L., Curtis W. Marean, and Shannon P. McPherron. "The use of barcodes in excavation projects: examples from Mossel Bay (South Africa) and Roc d e Marsal (France)." The SAA Archaeological Record 7.1 (2007): 33 38. Erlandson, J.M., 2001. The archaeology of aquatic adaptations: paradigms for a new millennium. Journal of Archaeological Research , 9 (4), pp.287 350.

PAGE 92

81 Esteban, I., Marean, C.W., Cowling, R .M., Fisher, E.C., Cabanes, D. and Albert, R.M., 2019. Palaeoenvironments and plant availability during MIS 6 to MIS 3 on the edge of the Palaeo Agulhas Plain (south coast, South Africa) as indicated by phytolith analysis at Pinnacle Point. Quaternary Scie nce Reviews . Fagundes, N.J., Ray, N., Beaumont, M., Neuenschwander, S., Salzano, F.M., Bonatto, S.L. and Excoffier, L., 2007. Statistical evaluation of alternative models of human evolution. Proceedings of the National Academy of Sciences , 104 (45), pp.17614 17619. Faith, J.T. and Gordon, A.D., 2007. Skeletal element abundances in archaeofaunal assemblages: economic utility, sample size, and assessment of carcass transport strategies. Journal of Archaeological Science , 34 (6), pp.872 882. Faith, J.T. , Marean, C.W. and Behrensmeyer, A.K., 2007. Carnivore competition, bone destruction, and bone density. Journal of Archaeological Science , 34 (12), pp.2025 2034. Fisher, J.W., 1995. Bone surface modifications in zooarchaeology. Journal of Archaeological met hod and theory , 2 (1), pp.7 68. Fisher, E.C., Bar Matthews, M., Jerardino, A. and Marean, C.W., 2010. Middle and Late Pleistocene paleoscape modeling along the southern coast of South Africa. Quaternary Science Reviews , 29 (11), pp.1382 1398. Galán, A.B., Rodríguez, M., De Juana, S. and Domínguez Rodrigo, M., 2009. A new experimental study on percussion marks and notches and their bearing on the interpretation of hammerstone broken faunal assemblages. Journal of Archaeological Science , 36 (3), pp.776 784. Ga thorne Hardy, F.J. and Harcourt Smith, W.E.H., 2003. The super eruption of Toba, did it cause a human bottleneck?. Journal of Human Evolution , 45 (3), pp.227 230. Goldblatt, P., 1997. Floristic diversity in the Cape flora of South Africa. Biodiversity & Co nservation , 6 (3), pp.359 377. Hammer, M.F., Karafet, T., Rasanayagam, A., Wood, E.T., Altheide, T.K., Jenkins, T., Griffiths, R.C., Templeton, A.R. and Zegura, S.L., 1998. Out of Africa and back again: nested cladistic analysis of human Y chromosome varia tion. Molecular biology and evolution, 15(4), pp.427 441. Harpending, H.C., Batzer, M.A., Gurven, M., Jorde, L.B., Rogers, A.R. and Sherry, S.T., 1998. Genetic traces of ancient demography. Proceedings of the National Academy of Sciences , 95 (4), pp.1961 1 967.

PAGE 93

82 Hawks, J., Hunley, K., Lee, S.H. and Wolpoff, M., 2000. Population bottlenecks and Pleistocene human evolution. Molecular biology and evolution, 17(1), pp.2 22. Higham, T., Douka, K., Wood, R., Ramsey, C.B., Brock, F., Basell, L., Camps, M., Arrizab alaga, A., Baena, J., Barroso Ruíz, C. and Bergman, C., 2014. The timing and spatiotemporal patterning of Neanderthal disappearance. Nature, 512(7514), p.306. Hodgkins J , Marean CW, Turq A, Sandgathe D, McPherron SP, Dibble HL, 2016. Climate mediated (Dordogne Valley, France) Journal of Human Evolution. 96:1 18. Jacobs, Z., 2010. An OSL chronology for the sedimentary deposits from Pinnacle Point Cave 13B a punctuated pres ence. Journal of Human Evolution , 59 (3), pp.289 305. Johnson, E., 1985. Current developments in bone technology. In Advances in archaeological method and theory (pp. 157 235). Academic Press. Kaplan, J., 1997. Archaeological study: proposed Pinnacle Point development. Riebeek West, South Africa: Agency for Cultural Resource Management . Karkanas, P. and Goldberg, P., 2010. Site formation processes at Pinnacle point Cave 13B (Mossel Bay, Western Cape Province, South Africa): resolving stratigraphic and depositional complexities with micromorphology. Journal of Human Evolution , 59 (3), pp.256 273. Karkanas, P., Brown, K.S., Fisher, E.C., Jacobs, Z. and Marean, C.W., 2015. Interpreting human behavior from depositional rates and combustion features through the study of sedimentary microfacies at site Pinnacle Point 5 6, South Africa. Journal of human evolution , 85 , pp.1 21. Kelly, R.L., 1995. The foraging spectrum. Klein, R.G., 1983. Palaeoenvironmental implications of Quaternary large mammals in the fynb os region. Fynbos palaeoecology: a preliminary synthesis , pp.116 138. Klein, R.G. and Cruz Uribe, K., 1984. The analysis of animal bones from archeological sites . University of Chicago press. Kosso, P., 1991. Method in archaeology: Middle range theory as hermeneutics. American Antiquity , 56 (4), pp.621 627.

PAGE 94

83 Lam, Y.M., Chen, X., Marean, C.W. and Frey, C.J., 1998. Bone density and long bone representation in archaeological faunas: comparing results from CT and photon densitometry. Journal of Archaeological Science , 25 (6), pp.559 570. Levin, M.G., Potapov, L.P. and Dunn, S.P., 1964. The peoples of Siberia . University of Chicago Press. Lewontin, R.C., 1972. The apportionment of human diversity. In Evolutionary biology (pp. 381 398). Springer, New York, NY. L Loftus, E., Lee Thorp, J., Leng, M., Marean, C. and Sealy, J., 2019. Seasonal scheduling of shellfish collection in the Middle and Later Stone Ages of southern Africa. Journal of human evolution, 128, pp.1 16. Lyman, R.L., 1985. Bone frequencies: differential transport, in situ destruction, and the MGUI. Journal of Archaeological Science, 12(3), pp.221 236. Marean, CW., (Ed.). 2010a. The Middle Stone Age at Pinnacle Point Site 13B, a Coastal Cave near Mossel Bay (Western Cape Province, South Africa) [Special Issue]. Journal of Human Evolution, 59, (3 4). Marean, C.W., 2010b. Introduction to the Special Issue The Mid dle Stone Age at Pinnacle Point Site 13B, a Coastal Cave near Mossel Bay (Western Cape Province, South Africa). Marean, C.W., 2010c. Pinnacle Point Cave 13B (Western Cape Province, South Africa) in context: the Cape floral kingdom, shellfish, and modern h uman origins. Journal of Human Evolution , 59 (3), pp.425 443. Marean, C.W., 2011. Coastal South Africa and the coevolution of the modern human lineage and the coastal adaptation. In Trekking the shore (pp. 421 440). Springer, New York, NY. Marean, C.W., 2 014. The origins and significance of coastal resource use in Africa and Western Eurasia. Journal of Human Evolution, 77, pp.17 40. Marean, C.W., 2015. The most invasive species of all. Scientific American, 313(2), pp.32 39. Marean, C.W. and Assefa, Z., 1 999. Zooarcheological evidence for the faunal exploitation behavior of Neandertals and early modern humans. Evolutionary Anthropology: Issues, News, and Reviews: Issues, News, and Reviews , 8 (1), pp.22 37.

PAGE 95

84 Marean, C.W. and Frey, C.J., 1997. Animal bones fr om caves to cities: reverse utility curves as methodological artifacts. American Antiquity , 62 (4), pp.698 711. Marean, C.W. and Kim, S.Y., 1998. Mousterian large mammal remains from Kobeh Cave behavioral implications for Neanderthals and early modern huma ns. Current anthropology , 39 (S1), pp.S79 S114. Marean, C.W., Abe, Y., Frey, C.J. and Randall, R.C., 2000. Zooarchaeological and taphonomic analysis of the Die Kelders Cave 1 Layers 10 and 11 Middle Stone Age larger mammal fauna. Journal of Human Evolution , 38 (1), pp.197 233. Marean, C.W., Bar Matthews, M., Bernatchez, J., Fisher, E., Goldberg, P., Herries, A.I., Jacobs, Z., Jerardino, A., Karkanas, P., Minichillo, T. and Nilssen, P.J., 2007. Early human use of marine resources and pigment in South Africa d uring the Middle Pleistocene. Nature , 449 (7164), pp.905 908. Marean, Curtis W., Nilssen, P.J., and Kyle Brown, Antonieta Jerardino, and Deano Stynder. 2004. Paleoanthropological investigations of Middle Stone Age sites at Pinnacle Point, Mossel Bay (South Africa): archaeology and hominid remains from the 2000 field season. Paleoanthropology 2, no. 1. Martinson, D.G., Pisias, N.G., Hays, J.D., Imbrie, J., Moore, T.C. and Shackleton, N.J., 1987. Age dating and the orbital theory of the ice ages: development of a high resolution 0 to 300,000 year chronostratigraphy. Quaternary research , 27 (1), pp.1 29. Mellars, P., 1996. Symbolism, language, and the Neanderthal mind. Modelling the early human mind , pp.15 32. Metcalfe, D. and Jones, K.T., 1988. A reconsideration of animal body part utility indices. American Antiquity , 53 (3), pp.486 504. Morlan, R.E., 1984. Toward the Definition of Criteria for the Recognition of Artificial Bone Alterations 1. Quaternary Research , 22 (2), pp.160 171. O'Connell, J.F., Hawkes, K. and Jones, N.B., 1988. Hadza hunting, butchering, and bon e transport and their archaeological implications. Journal of Anthropological research , 44 (2), pp.113 161. O'Connell, J.F. and Marshall, B., 1989. Analysis of kangaroo body part transport among the Alyawara of Central Australia. Journal of Archaeological Science , 16 (4), pp.393 405.

PAGE 96

85 Oestmo, S. and Marean, Curtis W., 2014. Pinnacle Point: Excavation and Survey Methods. In Smith C. (ed.) Encyclopedia of Global Archaeology, New York: Springer. Outram, A.K., 2001. A new approach to identifying bone marrow and grease exploitation: Journal of archaeological science, 28(4), pp.401 410. Pickering, T.R., 2002. Reconsideration of criteria for differentiating faunal assemblages accumulated by hyenas and hom inids. International Journal of Osteoarchaeology, 12(2), pp.127 141. Potts, R., 2017. Early Hominid Activities at Olduvai: Foundations of Human Behaviour. Routledge. Potts, R. and Shipman, P., 1981. Cutmarks made by stone tools on bones from Olduvai Gorg e, Tanzania. Nature , 291 (5816), p.577. Potts, R., Shipman, P. and Ingall, E., 1988. Taphonomy, paleoecology, and hominids of Lainyamok, Kenya. Journal of Human Evolution , 17 (6), pp.597 614. Rampino, M.R., Ambrose, S.H., McCoy, F. and Heiken, G., 2000. Volcanic hazards and disasters in human antiquity. Geological Society of America Special Paper, 345, pp.71 82. Rampino, M.R. and Self, S., 1992. Volcanic winter and accelerated glaciation following the Toba super eruption. Nature , 359 (6390), p.50. Rector , A.L. and Reed, K.E., 2010. Middle and late Pleistocene faunas of Pinnacle Point and their paleoecological implications. Journal of Human Evolution , 59 (3), pp.340 357. Relethford, J.H., 1995. Genetics and modern human origins. Evolutionary Anthropology: Issues, News, and Reviews , 4 (2), pp.53 63. Renfrew, C. and Zubrow, E.B.W., 1994. Towards a cognitive archaeology. South African Coast Paleoclimate Paleoenvironment Paleoecology Paleoanthropology Project, 2017. [Pinnacle Point Excavation Database]. Unpublished Raw Data. Schoville, B.J. and Otárola Castillo, E., 2014. A model of hunter gatherer skeletal element transp ort: The effect of prey body size, carriers, and distance. Journal of human evolution, 73, pp.1 14.

PAGE 97

86 Selvaggio, M.M., 1998. Evidence for a three stage sequence of hominid and carnivore involvement with long bones at FLKZinjanthropus, Olduvai Gorge, Tanzani a. Journal of Archaeological Science , 25 (3), pp.191 202. Shackleton, N.J. and Opdyke, N.D., 1973. Oxygen isotope and palaeomagnetic stratigraphy of Equatorial Pacific core V28 238: Oxygen isotope temperatures and ice volumes on a 10 5 year and 10 6 year sca le. Quaternary research, 3(1), pp.39 55. Shipman, P., 1981. Applications of scanning electron microscopy to taphonomic problems. Annals of the New York Academy of Sciences , 376 (1), pp.357 385. Shipman, P. and Rose, J., 1983. Early hominid hunting, butche ring, and carcass processing behaviors: approaches to the fossil record. Journal of anthropological Archaeology , 2 (1), pp.57 98. Smith, E.I., Jacobs, Z., Johnsen, R., Ren, M., Fisher, E.C., Oestmo, S., Wilkins, J., Harris, J.A., Karkanas, P., Fitch, S. an d Ciravolo, A., 2018. Humans thrived in South Africa through the Toba eruption about 74,000 years ago. Nature , 555 (7697), p.511. Stiner, M.C., Kuhn, S.L., Weiner, S. and Bar Yosef, O., 1995. Differential burning, recrystallization, and fragmentation of archaeological bone. Journal of Archaeological Science , 22 (2), pp.223 237. Thompson, J.C., 2010. Taphonomic analysis of the Middle Stone Age faunal assemblage from Pinnacle Point Cave 13B, Western Cape, South Africa. Journal of Human Evolution , 59 (3), pp.321 339. Thompson, E., Williams, H.M. and Minichillo, T., 2010. Middle and late Pleistocene middle stone age lithic technology from Pinnacle point 13B (Mossel Bay, western Cape province, South Africa). Journal of Human Evolution , 59 (3), pp.358 377. Vi lla, P. and Mahieu, E., 1991. Breakage patterns of human long bones. Journal of human evolution , 21 (1), pp.27 48. Walter, R.C., Buffler, R.T., Bruggemann, J.H., Guillaume, M.M., Berhe, S.M., Negassi, B., Libsekal, Y., Cheng, H., Edwards, R.L., von Cosel, R. and Néraudeau, D., 2000. Early human occupation of the Red Sea coast of Eritrea during the last interglacial. Nature, 405(6782), p.65. Watts, I., 2010. The pigments from Pinnacle Point Cave 13B, Western Cape, South Africa. Journal of Human Evolution , 5 9 (3), pp.392 411.

PAGE 98

87 Wilkins, J., Brown, K.S., Oestmo, S., Pereira, T., Ranhorn, K.L., Schoville, B.J. and Marean, C.W., 2017. Lithic technological responses to Late Pleistocene glacial cycling at Pinnacle Point Site 5 6, South Africa. PloS one , 12 (3), p.e01 74051. Yellen, J.E., 1991. Small mammals: !Kung San utilization and the production of faunal assemblages. Journal of Anthropological Archaeology , 10 (1), pp.1 26. Zielinski, G.A., 2000. Use of paleo records in determining variability within the volcanism climate system. Quaternary Science Reviews, 19(1 5), pp.417 438. Zielinski, G.A., Mayewski, P.A., Meeker, L.D., Whitlow, S., Twickler, M.S. and Taylor, K., 1996. 71,000 years ago. Geophysical Research Letters , 23 (8), pp.837 840.