Coastal foraging of Turbo sarmaticus at Klasies River Main Site: an experimental approach by Carl Louis Holmes (1447271) Dissertation Submitted in fulfilment of the requirements for the degree Master of Science by research only: in Archaeology In the Faculty of Science, University of the Witwatersrand, Johannesburg, South Africa, in fulfilment of the requirements for the degree of Master of Science. Supervisor: Prof. Sarah Wurz Co-supervisors: Dr Jerome Reynard and Dr Jan De Vynck 29 March 2023 ii Declaration I declare that this dissertation is my own, unaided work. It is being submitted for the Degree of Master of Science in Archaeology at the University of the Witwatersrand. It has not been submitted for any other degree or examination at any other University. (Signature of candidate) 29 day of March 20 23 at Edenvale iii Abstract An increase in shellfish exploitation and strong seasonal fluctuations in terrestrial foraging at ca. 100 000 years ago along the Cape coast has been noted. It is hypothesised that marine foraging provided a more reliable year-round source of nutrients. As Turbo sarmaticus (alikreukel) is one of the major prey species found at Middle Stone Age (MSA) sites along the southern Cape, and Klasies River Main Site (KRM) specifically, this species was targeted in an archaeological and actualistic study. T. sarmaticus specimens were collected and processed by two participants to examine the possible methods used by MSA foragers and the meat was extracted using three tool types (bone, lithics, wood). Some specimens were cooked whilst others were processed raw. An analysis of the lengths of T. sarmaticus opercula of the ca. 100 000 years ago Shell Midden One (SMONE) layer from KRM shows that the archaeological opercula are similar in size to the experimental sample. A larger proportion of the archaeological opercula, however, fall within the smaller size range. An age index was developed for T. sarmaticus by correlating operculum size to shell size. This suggested that the archaeological specimens were younger when collected, or that they foraged during spring or winter. The opercula from SMONE are markedly smaller than other MSA II opercula from the Cape coast but bigger than LSA opercula. This might be related to more intensive predation, seasonality and sea surface temperature. A taphonomic analysis of both the experimental and archaeological samples was undertaken. The experimental specimens that were processed raw indicated percussion marks typically placed in the 3rd and 4th wedges. A proportion of the archaeological sample shows this pattern as well. This shows that some of the archaeological samples may have been raw extracted. Furthermore, the breakage types are very similar in both samples, however post-depositional breakage only affected the archaeological sample. This study demonstrates that broader insight can be developed through an actualistic study combined with a taphonomic analysis. Keywords: Turbo sarmaticus, alikreukel, opercula, Klasies River Main Site, Middle Stone Age, MIS 5c, Experimental Archaeology, Actualistic study, Taphonomy, Southern Cape, Coastal Foraging. iv Acknowledgements I would like to acknowledge my supervisors, Prof. Sarah Wurz, Dr Jerome Reynard and Dr Jan De Vynck, for all their guidance and support during the process of my research. Their knowledge and expertise was greatly appreciated during the analysis and development of the taphonomic criteria, and for their patience during the entire process. I would not have been able to complete my thesis without their guidance. I would like to specifically thank Prof. Sarah Wurz for the out-of-context MSA stone tools. I would also like to acknowledge Dr Jan De Vynck for his knowledge and assistance during the running of the experimental process . His experience and knowledge about the intertidal zone are truly appreciated. I would like to acknowledge Dr Jan De Vynck’s family, for inviting me into their home during my actualistic study, allowing me the space to conduct the processing aspect of my experiment. Dr Jan De Vynck for arranging the Tragelaphus strepiceros (Greater Kudu) bones from the local butcher for the bone tool production and collecting the Olea europaea ssp. africana (African olive) branches before my arrival. Dr Jan De Vynck for providing me with data from his PhD thesis for comparative study, and for allowing me the time to understand and view the material that was used and collected. I would like to acknowledge Hlokwa Madithapa for making all travel arrangements required for my trip to Still Bay, especially for assisting and putting up with all the changes that were made due to the COVID–19 lockdowns that restricted my travels to and from the Western Cape. The Department of Environmental Affairs – specifically Zimasa Jika and Lieze Swart for their assistance in receiving my research permit (RES2021–85). I would like to thank my family for their assistance wherever possible during the process of completing my dissertation. My friends and colleagues for their support and guidance during the two years and their assistance, and their support of my research especially with the never-ending talking about my research and the assistance in gaining access to specific literature. I would like to graciously thank both Prof. Sarah Wurz (year 1) and Dr Jan De Vynck (year 2) for funding my research and my travels. The funding is greatly appreciated, and I am so very thankful for their assistance. This was through the National Research Foundation v (NRF), GENUS (CoE: Palaeoscience) and the African Centre for Coastal Palaeoscience, Nelson Mandela University. vi Table of Contents Declaration ....................................................................................................................................................... ii Abstract ............................................................................................................................................................. iii Acknowledgements ...................................................................................................................................... iv List of Figures ................................................................................................................................................... x List of Tables ................................................................................................................................................. xiv List of Appendix 1 Figures ....................................................................................................................... xvi List of Appendix 1 Tables ........................................................................................................................ xvii Chapter 1: Introduction ............................................................................................................................... 1 1.1. Research question ............................................................................................................................. 1 1.2. Research Aim and Objectives........................................................................................................ 1 1.3. Rationale ............................................................................................................................................... 2 1.4. Hypotheses .......................................................................................................................................... 3 1.5. Conclusion ............................................................................................................................................ 3 Chapter 2: Literature Review .................................................................................................................... 4 2.1. Site Context .......................................................................................................................................... 4 2.2. Coastal Intertidal Zone .................................................................................................................... 7 2.3. Habitat of Turbo sarmaticus .......................................................................................................... 8 2.4. MSA foraging of shellfish in the southern Cape ................................................................... 10 2.4.1. Theory .......................................................................................................................................... 10 2.4.2. Coastal adaptations and shellfish exploitation in the southern Cape ................. 12 2.4.3. Nutritional Value ..................................................................................................................... 13 2.5. Turbo sarmaticus operculum size and the ‘predation hypothesis’ .............................. 15 2.6. Turbo sarmaticus operculum size, growth patterns and the ‘climatic and environmental hypothesis’ .................................................................................................................. 15 2.7. Seasonality and Turbo sarmaticus ............................................................................................ 17 2.8. Previous work on experimental processing of Turbo sarmaticus ................................ 18 2.9. Turbo sarmaticus opercula and site formation .................................................................... 20 2.10. Conclusion ....................................................................................................................................... 21 Chapter 3: Experimental Materials and Methods ............................................................................ 22 vii 3.1. Introduction ...................................................................................................................................... 22 3.2. Foraging .............................................................................................................................................. 22 3.2.1. Participants and Prey ............................................................................................................. 24 3.2.2. Habitat ......................................................................................................................................... 25 3.2.3. Foraging Tools .......................................................................................................................... 26 3.2.4. Foraging Process...................................................................................................................... 33 3.2.5. Processing .................................................................................................................................. 34 3.3. Conclusion .......................................................................................................................................... 38 Chapter 4: Analytical Methods ................................................................................................................ 39 4.1. Introduction ...................................................................................................................................... 39 4.2. Opercula Size ..................................................................................................................................... 39 4.2.1. Opercula Length ....................................................................................................................... 39 4.2.2. Opercula Width ........................................................................................................................ 40 4.2.3. Opercula Thickness and Weight ........................................................................................ 40 4.3. Opercula Age ..................................................................................................................................... 41 4.4. Opercula Shape ................................................................................................................................. 43 4.5. Opercula Taphonomy Criteria .................................................................................................... 43 4.5.1. Completeness of Opercula Dome and Flange ............................................................... 44 4.5.2. Breakage Type shown on Opercula .................................................................................. 44 4.5.3. Incisions on Opercula ............................................................................................................ 44 4.5.4. Colouration of Opercula ........................................................................................................ 44 4.5.6. Percussion Marks on Opercula ........................................................................................... 44 4.5.7. Bore marks ................................................................................................................................. 46 4.5.8. Etching ......................................................................................................................................... 46 4.5.9. Encrustation of Opercula ...................................................................................................... 46 4.5.10. Surface Condition .................................................................................................................. 46 4.5.11. Abrasion ................................................................................................................................... 46 4.5.12. Preservation of Opercula ................................................................................................... 46 4.5.13. Perforations on Opercula ................................................................................................... 47 viii 4.5.14. Periostracum on Operculum ............................................................................................ 47 4.6. Opercula Placement Model .......................................................................................................... 47 4.7. Conclusion .......................................................................................................................................... 48 Chapter 5: Results ........................................................................................................................................ 49 5.1. Introduction ...................................................................................................................................... 49 5.2. Opercula size ..................................................................................................................................... 49 5.3. Opercula Shape ................................................................................................................................. 54 5.4. Opercula Age ..................................................................................................................................... 54 5.5. Opercula Taphonomy .................................................................................................................... 55 5.5.1. Incisions of opercula .............................................................................................................. 55 5.5.2. Completeness of Opercula ................................................................................................... 56 5.5.3. Breakage of Opercula ............................................................................................................. 57 5.5.4. Percussion marks on Opercula........................................................................................... 59 5.5.5. Placement of percussion marks ......................................................................................... 61 5.5.6. Bore marks ................................................................................................................................. 63 5.5.7. Opercula Colouration ............................................................................................................. 64 5.5.8. Etching of Opercula ................................................................................................................ 65 5.5.9. Encrustation of opercula ...................................................................................................... 66 5.5.10. Abrasion of Opercula ........................................................................................................... 67 5.5.11. Preservation of Opercula ................................................................................................... 68 5.6. Conclusion .......................................................................................................................................... 69 Chapter 6: Discussion ................................................................................................................................. 70 6.1. Introduction ...................................................................................................................................... 70 6.2. Opercula Size and shape ............................................................................................................... 70 6.3. Opercula Age ..................................................................................................................................... 72 6.4. Opercula Processing ....................................................................................................................... 73 6.5. Opercula Taphonomy .................................................................................................................... 75 Chapter 7: Conclusion ................................................................................................................................ 77 Future Avenues of Research .................................................................................................................... 77 References ...................................................................................................................................................... 79 ix Appendices ..................................................................................................................................................... 86 Appendix 1A: Modern-day experimental data ................................................................................. 86 Appendix 1A.1. Foraging Data ............................................................................................................ 86 Appendix 1A.1.1. Foraging Notes .................................................................................................. 86 Appendix 1A.2. Processing Data ........................................................................................................ 87 Appendix 1A.2.1. Processing Notes .............................................................................................. 87 Appendix 1B: Analytical Data .................................................................................................................. 91 Appendix 1B.1: Experimental Age Index ........................................................................................ 91 Appendix 1B.2: Opercula Shape ......................................................................................................... 95 Appendix 1B.3: Opercula Age ............................................................................................................. 95 Appendix 1B.4: Opercula Processing ............................................................................................... 96 Opercula Completeness .................................................................................................................... 96 Breakage Type ...................................................................................................................................... 96 Colouration of T. sarmaticus opercula ........................................................................................ 96 Percussion Marks on T. sarmaticus opercula ........................................................................... 97 Percussion Marks details.................................................................................................................. 97 Bore Marks on Opercula ................................................................................................................... 97 Appendix 1B.5: Opercula Taphonomy ............................................................................................. 98 Etching on Opercula ........................................................................................................................... 98 Encrustation on Opercula ................................................................................................................ 98 Encrustation percentage .................................................................................................................. 98 Abrasion on Opercula ........................................................................................................................ 98 Appendix 1C: Ocean Condition Rating ................................................................................................. 99 Appendix 1D: Foraging Permits .......................................................................................................... 100 Appendix 1D.1: Research Permit (RES2021–85) ..................................................................... 100 Appendix 1D.2: Additional Fishing Permit (1052834) .......................................................... 104 x List of Figures Figure 2. 1: 3D image plan of KRM Cave 1, Witness Baulk deposits (Wurz et al. 2018: Fig. 3, 105). ................................................................................................................................................................ 6 Figure 2. 2: KRM Cave 1 Witness Baulk south wall profile (Brenner et al. 2022: Figure 2, 7)........................................................................................................................................................................... 6 Figure 2. 3: Spirals on T. sarmaticus operculum (adapted from Galimberti et al. 2017: Fig.1, 63). ......................................................................................................................................................... 16 Figure 2. 4: Cooking of T. sarmaticus on the coals (De Vynck 2017: Fig. A 3.15a, Appendix 3)......................................................................................................................................................................... 19 Figure 2. 5: Raw extraction of the meat from T. sarmaticus shell with a bone tool (De Vynck 2017: Fig. A3.12, Appendix 3). ................................................................................................................ 20 Figure 3. 1: Jongensfontein, where the experimental foraging took place, in relation to the southern Cape coastal archaeological sites Klasies River main site (KRM), Blombos Cave (BBC) and Pinnacle Point (PP) (Google Earth Pro 2021a). .......................................................... 23 Figure 3. 2: Foraging locations for Still Bay foraging experiment, A: Test foraging day at Jongensfontein East, 1: Lithics foraging at Jongensfontein West, 2: Wooden tool and Bone tool foraging at Brandersfontein Farm (Google Earth Pro 2021b). ......................................... 23 Figure 3. 3: A, the dorsal side of a T. sarmaticus shell, B, the ventral side of the T. sarmaticus shell with operculum in place. C, of an intact P. perna shell. D, of the ventral side of the S. longicosta shell. E, of the ventral side of the C. oculus shell. F, of the ventral side of the C. miniata shell. ....................................................................................................................... 24 Figure 3. 4: A, dome (dorsal) surface of the T. sarmaticus opercula. B, the flange (ventral) surface of the T. sarmaticus opercula. .................................................................................................. 25 Figure 3. 5: Out of context Middle Stone Age lithics from the Klasies River main site spill heap after use, showing breakage on the lithics as a result of the experiment. .................. 26 Figure 3. 6: A, the measuring of the Olea europaea ssp. africana branches to 0.45 m. B, the hand saw used to cut the branches before the production of the tools, and C, the cut Olea europaea ssp. africana branches for both participants. ................................................................ 28 xi Figure 3. 7: Wooden tool production by P2 using the quartzite veins on the Table Mountain Sandstone rocky headlands. ................................................................................................ 28 Figure 3. 8: Worksite after the tool production, in the image the quartzite veins are protruding from the Table Mountain Sandstone exposed rocky headlands. ....................... 29 Figure 3. 9: Worksite after the tool production, in the image the Table Mountain Sandstone exposed rocky headland surface shows the site used to refine and sharpen the tool where no quartzite veins are present. ........................................................................................ 29 Figure 3. 10: Wooden tools made by P2 out of Olea europaea ssp. africana branches. .... 30 Figure 3. 11: 1, T. strepiceros humerus (baked). 2, T. strepiceros femur (raw). 3, T. strepiceros radius and ulna (baked). 4, T. strepiceros femur (raw). ......................................... 31 Figure 3. 12: Cobblestones used for the hammer and anvil technique to crack open the long bone. A is the Cobblestone used by P1, and B is the Cobblestone used by P2. ........... 32 Figure 3. 13: A, the area where P2 produced the bone tools. B, the remains left behind postproduction (bone marrow, skin, bone fragments). ................................................................ 32 Figure 3. 14: A, bone tools made by P2 out of T. strepiceros long bones. B, bone tools made by P1 out of T. strepiceros long bones. ................................................................................................. 33 Figure 3. 15: A, the tool produced by P2 from the baked long bone. B, the bone fragments produced by P2 from the raw long bone. ........................................................................................... 33 Figure 3. 16: Fire for cooking T. sarmaticus. ...................................................................................... 35 Figure 3. 17: A, the shell length measurement technique by P1. B, the shell width measurement technique by P1. .............................................................................................................. 37 Figure 3. 18: Labelled diagram of the internal portion of the T. sarmaticus. ........................ 37 Figure 3. 19: A, how the cooked T. sarmaticus individuals bubble when done. B, the cooking of the T. sarmaticus shells in batches of 15. ...................................................................... 38 Figure 4. 1: A, how the length and width are measured on the ventral (flange) of the T. sarmaticus operculum. B, how the length and width are measured on the dorsal (dome) of the T. sarmaticus operculum............................................................................................................... 40 Figure 4. 2: P1 measuring the length of the T. sarmaticus shell in the actualistic study. . 43 xii Figure 4. 3: A, an experimental T. sarmaticus opercula showing percussion notch marks (PM) on the dome surface with a break. B, bone percussion marks caused by hammer and anvil techniques (from Blumenschine et al. 1996: Figure 2, 498). C, an archaeological T. sarmaticus opercula showing percussion pit marks (PM) on the dorsal surface. .............. 45 Figure 4. 4: Wedge Placement Model (Sherwood et al. 2016). A, the dorsal side wedge model 1–8. B, the ventral side wedge model 1–8. ........................................................................... 47 Figure 4. 5: Spiral Placement Model. A, the dorsal spiral placement model A–D. B, the ventral spiral placement model A–D. ................................................................................................... 48 Figure 5. 1: T. sarmaticus opercula length of raw and cooked from all three tool types from the experimental sample and the archaeological sample. ........................................................... 51 Figure 5. 2: T. sarmaticus opercula width of raw and cooked from all three tool types from the experimental sample and the archaeological sample. ........................................................... 52 Figure 5. 3: A, weight of the T. sarmaticus opercula. B, thickness of the T. sarmaticus opercula of raw and cooked from all three tool types from the experimental sample and the archaeological sample. ....................................................................................................................... 53 Figure 5. 4: Age distribution for the T. sarmaticus based on the opercula measurements of the experimental and archaeological sample (see Table 4. 3). ............................................. 55 Figure 5. 5: A, complete dome (ventral surface) on T. sarmaticus opercula for combining the experimental sample and the archaeological sample. B, complete flange (dorsal surface) on T. sarmaticus opercula for combining the experimental sample and the archaeological sample. ............................................................................................................................... 56 Figure 5. 6: Breakage types seen on the SMONE samples: A, transverse break, B, eroded break, C, irregular break, D, Intact......................................................................................................... 58 Figure 5. 7: Breakage type on the archaeological and experimental T. sarmaticus opercula. ............................................................................................................................................................................. 59 Figure 5. 8: A, percussion marks on the ventral surface of the T. sarmaticus opercula, B, percussion mark details on the T. sarmaticus operculum. ........................................................... 60 Figure 5. 9: Wedge Placement Model (Sherwood et al. 2016). A, the dorsal side wedge model 1–8. B, the ventral side wedge model 1–8. ........................................................................... 62 xiii Figure 5. 10: Spiral Placement Model. A, the dorsal spiral placement model A–D. B, the ventral spiral placement model A–D. ................................................................................................... 62 Figure 5. 11: Bore Marks were seen on the ventral surface of the T. sarmaticus operculum. ............................................................................................................................................................................. 63 Figure 5. 12: A, bore mark seen on the experimental opercula. B, bore mark seen on the SMONE opercula. .......................................................................................................................................... 64 Figure 5. 13: Colouration on T. sarmaticus opercula. ..................................................................... 65 Figure 5. 14: Evidence of etching on the ventral surface of the T. sarmaticus operculum. ............................................................................................................................................................................. 65 Figure 5. 15: A, encrustation on the dome of the T. sarmaticus opercula. B, coverage of the encrusting on the T. sarmaticus opercula. .......................................................................................... 66 Figure 5. 16: Abrasions on the flange of the T. sarmaticus operculum. .................................. 68 Figure 5. 17: Preservation of the experimental and archaeological T. sarmaticus opercula. ............................................................................................................................................................................. 68 Figure 6. 1: T. sarmaticus operculum #41 from the SMONE layer at KRM, showing a charcoal line on the flange. ....................................................................................................................... 75 xiv List of Tables Table 2. 1: Intertidal Zones of Southern Africa and their placement within the intertidal and shallow subtidal zone (adapted from Langejans et al. 2012: Table 4, 84; Branch et al. 2022). .................................................................................................................................................................. 8 Table 2. 2: Common species found within the different Intertidal Zones (adapted from Langejans et al. 2012). ............................................................................................................................... 13 Table 3. 1: Ocean Condition rating scale, calibrated to the Beaufort scale, showing the optimal and suboptimal conditions for shellfish foraging. .......................................................... 34 Table 4. 1: Summary of statistics for the experimental T. sarmaticus shell and opercula length. ............................................................................................................................................................... 41 Table 4. 2: Relative age for T. sarmaticus shell length (from Yssel 1989: Table 11, 77). . 42 Table 4. 3: T. sarmaticus age index based on operculum length and inferred shell length (adapted from Yssel 1989: Table 11, 57). .......................................................................................... 42 Table 5. 1: A, Summary statistics for T. sarmaticus opercula with a combination of both P1 and P2, and the combination of all three tool types. B, Summary statistics for T. sarmaticus opercula from layer SMONE. ............................................................................................ 50 Table 5. 2: T. sarmaticus operculum shape for the experimental and archaeological sample. ............................................................................................................................................................. 54 Table 5. 3: T. sarmaticus opercula length and age index with experimental and SMONE n values. ............................................................................................................................................................... 54 Table 5. 4: Experimental T. sarmaticus opercula incision measurements, with tool type and participant. ............................................................................................................................................. 55 Table 5. 5: Instrument vs breakage type for the experimental sample with P1 and P2 combined. ........................................................................................................................................................ 57 Table 5. 6: Percussion marks on Experimental and SMONE T. sarmaticus opercula. ....... 59 Table 5. 7: Percussion mark details on Experimental and SMONE T. sarmaticus opercula. ............................................................................................................................................................................. 60 Table 5. 8: T. sarmaticus opercula percussion mark placement statistics. ............................ 61 xv Table 5. 9: T. sarmaticus operculum percussion marks localised statistics (Figure 5. 9 and Figure 5. 10). .................................................................................................................................................. 61 Table 5. 10: T. sarmaticus opercula encrustation placement statistics. .................................. 67 Table 5. 11: T. sarmaticus opercula encrustation localised numbers (Figure 4. 4 and Figure 4. 5). ................................................................................................................................................................... 67 Table 6. 1: Comparison of opercula length measurement (mm) statistics from various MSA II and LSA opercula from KRM and the experimental sample from this study. ........ 71 xvi List of Appendix 1 Figures Appendix 1A Figure 1: A, cobblestone used by P2 for Hammer and Anvil technique. B, lithic (J-LR 2) used by P1 that broke while raw processing the T. sarmaticus. .................... 88 Appendix 1A Figure 2: A, the foot of the T. sarmaticus. B, the muscle near the head sealing the operculum in place (Cold Water 2016). ...................................................................................... 88 Appendix 1A Figure 3: A, the workspace during raw processing using lithics. B, the operculum that fragmented and remained glued to the T. sarmaticus foot. ......................... 89 Appendix 1A Figure 4: The complete extraction of the T. sarmaticus meat from the shell due to cooked extraction. .......................................................................................................................... 89 Appendix 1A Figure 5: A, the foot-up technique used for the lithic raw processing. B, the new foot–down approach used for wooden and bone tool raw processing. ........................ 90 Appendix 1B Figure 1: Graph of the T. sarmaticus opercula shape for the experimental and archaeological samples. .................................................................................................................... 95 Appendix 1C Figure 1: Ocean Condition rating scale, showing the optimal and suboptimal conditions for shellfish foraging (pers. comm De Vynck). ........................................................... 99 xvii List of Appendix 1 Tables Appendix 1A Table 1: Experimental Opercula processing time for P1 and P2. ................... 87 Appendix 1A Table 2: Experimental Opercula cooking time for P1 and P2. ......................... 87 Appendix 1B Table 1: Experimental T. sarmaticus shell length age and opercula length age. ..................................................................................................................................................................... 91 Appendix 1B Table 2: T. sarmaticus opercula shape for both experimental and archaeological samples. ............................................................................................................................. 95 Appendix 1B Table 3: Age of T. sarmaticus for both experimental and archaeological samples. ........................................................................................................................................................... 95 Appendix 1B Table 4: Dome completeness for experimental and archaeological T. sarmaticus opercula. ................................................................................................................................... 96 Appendix 1B Table 5: Flange completeness for experimental and archaeological T. sarmaticus operculum. ............................................................................................................................... 96 Appendix 1B Table 6: Type of breakage seen in the experimental and archaeological T. sarmaticus opercula .................................................................................................................................... 96 Appendix 1B Table 7: Colouration seen on the experimental and archaeological T. sarmaticus operculum. ............................................................................................................................... 96 Appendix 1B Table 8: Percussion marks seen on T. sarmaticus opercula. ............................ 97 Appendix 1B Table 9: Percussion mark details for the experimental and archaeological T. sarmaticus opercula. ................................................................................................................................... 97 Appendix 1B Table 10: Bored marks on T. sarmaticus opercula. .............................................. 97 Appendix 1B Table 11: Etching seen on the dome of the experimental and archaeological T. sarmaticus opercula. .............................................................................................................................. 98 Appendix 1B Table 12: Encrustation of T. sarmaticus opercula. ............................................... 98 Appendix 1B Table 13: Percentage of T. sarmaticus opercula encrusted. ............................. 98 Appendix 1B Table 14: Abrasion on archaeological T. sarmaticus opercula. ....................... 98 1 Chapter 1: Introduction Turbo sarmaticus (alikreukel) is a species of aquatic Gastropoda, that is found along the Cape coastline of Southern Africa (Branch et al. 2022). This species of shellfish plays a vital role in understanding the foraging behaviour and coastal adaptation of Middle Stone Age (MSA) groups (e.g., Langejans et al. 2012). This study further investigates the role of T. sarmaticus in subsistence behaviour through an actualistic study. An actualistic study involves modern-day experimentation and interpretation in order to gain a deeper understanding of the archaeological material (Outram 2008; Toth and Schick 2009). As Outram (2008: 2) notes – “hypotheses can be tested with authentic materials and in a range of environmental conditions that aim to reflect more accurately ‘real life’ or ‘actualistic’ scenarios. Such experiments investigate activities that might have happened in the past using the methods and materials that would actually have been available.” T. sarmaticus and other shellfish specimens were collected during winter along the southern Cape coast, at Jongensfontein and they were processed in various ways. The size of the opercula is an important variable discussed in MSA coastal foraging studies and therefore experimental opercula were measured in addition to the archaeological opercula from ca. 100 000 years ago (ka) layer from Klasies River Main Site (KRM). A taphonomic investigation was undertaken to understand other processes that might have affected the surface conditions of the opercula. 1.1. Research question 1. What information on subsistence behaviour and taphonomy can be inferred from an actualistic study of T. sarmaticus? 2. Is the taphonomy of T. sarmaticus opercula from layer SMONE at KRM informative of MSA subsistence behaviour? 3. What are the possible site formation processes during this period at the site? 1.2. Research Aim and Objectives This dissertation aims to undertake an actualistic study through the foraging and processing of T. sarmaticus and to compare the foraged opercula with those from SMONE at KRM dating to ca.100 ka (MIS 5c). 2 Objectives: 1. To forage and process T. sarmaticus from the southern Cape coast (Jongensfontein). 2. To determine the taphonomic characteristics resulting from foraging and processing. 3. To analyse and examine the taphonomic conditions of the experimental and archaeological opercula. 4. To infer subsistence behaviour and possible site formation processes in the layer SMONE at KRM. 1.3. Rationale It is hypothesised that coastal foraging became more prominent around 100ka on the southern Cape coast (Will et al. 2019) and that this was associated with evolutionary advantages. Although the notion of shellfish gathering was initially seen as an inherently low-return activity (Meehan 1982), there is now an increased emphasis on the contribution of marine foraged proteins in carbohydrate-rich, plant-based diets (De Vynck et al. 2016: 102). Importantly, it is seen that intertidal foraging of shellfish provides a stable and relatively low depletion resource for human foragers. This would have provided a more reliable year-round resource base over the seasonally fluctuating terrestrial-based resources (De Vynck et al. 2020). Shellfish exploitation of the recently excavated layers at KRM has been described (Brenner et al. 2022; Wurz et al. 2022), and it has been shown that T. sarmaticus is one of the major prey species (Brenner et al. 2022). Notwithstanding its importance little is known about the foraging, processing and site formation processes that affect the opercula of this taxon. De Vynck (2017) undertook an actualistic study where the foraging of T. sarmaticus was done in a modern context with indigenous coastal foragers. An actualistic study combined with an archaeological analysis of MSA T. sarmaticus has not been undertaken yet, and this in-depth investigation will aid in a deeper understanding of the site formation processes and the shellfish foraging behaviour within this period. 3 1.4. Hypotheses Hypothesis 1 – An actualistic study of Turbo sarmaticus provides new information on Middle Stone Age coastal foraging. Hypothesis 2 – The taphonomic analysis of the T. sarmaticus opercula provides information on the processing. Hypothesis 3 – The taphonomic condition of the T. sarmaticus opercula, indicates that post-depositional processes affect the opercula in layer SMONE. 1.5. Conclusion This chapter introduces the areas of research that is investigated in the dissertations, and the rationale behind this particular study. Bringing forward the questions that will be answered in this research through the actualistic study and experimentation. In the next chapter the literature from previous studies is reviewed. 4 Chapter 2: Literature Review 2.1. Site Context Klasies River main site (KRM) is situated along the Tsitsikamma coast of the eastern Cape Province, South Africa. Turbo sarmaticus (alikreukel) is commonly found in archaeological sites along the southern Cape coastline, including at KRM (Langejans et al. 2012: 86). KRM has been recognised since the 1960s for its archaeological findings and importance. In March 2015 the site attained National Heritage Site status (Wurz et al. 2018: 103). KRM consists of four closely located caves and shelters (Caves 1, 1A, 1B and 2) (Brenner & Wurz 2019), with the samples analysed here originating from Cave 1, Witness Baulk (Wurz et al. 2018: 103). KRM was first excavated in the period between 1967 – 1968 by Ronald Singer and John Wymer (Singer & Wymer 1982). From 1984 to 1995, KRM was excavated by Hilary Deacon with the Witness Baulk (Figure 2. 1), being excavated between 1991 and 1995 (Deacon 1995; Wurz et al. 2018). The excavations were thereafter led by Sarah Wurz since 2015, continuing with the excavation of the Witness Baulk. The MSA II layers at KRM contain the SASU (Shell and Sand Upper) and the SASL (Shell and Sand Lower) sub-members that fall within the SAS (Shell and Sand) member containing MSA II/Mossel Bay stone artefacts (Wurz et al. 2018). The MSA II deposits also consists of large mammal remains, shellfish remains and a large rubble matrix (Wurz et al. 2018). The Wurz excavation started in the SASL sub-member of the Witness Baulk with the excavation of the HHH base and thereafter the SMONE (Shell Midden One) layer (Wurz et al. 2018), the layer investigated in this study. Singer and Wymer excavated the equivalent of the SASL member in layer 17 (Singer & Wymer 1982; Deacon 1995). The palaeoenvironmental conditions at KRM during the MSA II lower can be understood by investigating the link between fauna and the environment. Reynard & Wurz (2020) investigate the role that ungulate diversity at KRM play in understanding the environmental conditions during the MSA. Reynard & Wurz (2020) undertook a study to interpret the richness of ungulates in the deposit and the frequency of grazers. The results indicated a peak during the MSA I, with a substantial drop in the MSA II and a slight 5 increase again in the MSA III. This shows that changes in vegetation and the environment have an effect on the diversity of ungulates. It is suggested that the increase in ungulates, indicates a grassland environment. Therefore, during the MSA II, the environment around KRM was less grassland and more likely consisting of closed bushy vegetation (Reynard & Wurz 2020). It has been suggested that the faunal density is higher in the MSA II lower layers at KRM, in SMONE, during a period that is associated with MIS 5c. The shellfish density during that same period at KRM is also relatively high, suggesting some link between occupation intensity and coastal settlement (Reynard 2022). The layer investigated in this research is Shell Midden One (SMONE), which is situated in the Witness Baulk and is a ca. 12 m of baulk sediment in the central part of cave 1 (Figure 2. 1) (Deacon & Geleijnse 1988). Layer SMONE dates to ca. 100ka (Brenner & Wurz 2019). Figure 2. 2 shows the southern profile of the Witness Baulk. Layer SMONE is a sloped wedge deposit with a thickness of 5 cm in the western part, thickening towards the east to 15 cm. The layer consists of shellfish in a soil and rubble matrix with small charcoal pieces and small patches of leached ash preserved (Brenner & Wurz 2019; Morrissey et al. 2022). 6 Figure 2. 1: 3D image plan of KRM Cave 1, Witness Baulk deposits (Wurz et al. 2018: Fig. 3, 105). Figure 2. 2: KRM Cave 1 Witness Baulk south wall profile (Brenner et al. 2022: Figure 2, 7). 7 2.2. Coastal Intertidal Zone The South African coast experiences a semi-diurnal tidal system (the tides occur twice a day). This means that in a one-day cycle there will be two high tides and two low tides (Langejans et al. 2012), with the extent and peak of the tides depending on the lunar and solar cycles. The intertidal zone is the area where the water meets the shore between the high and low tide (Kvale 2006). International tidal zonation classification was determined by Stephenson & Stephenson (1949) and Stephenson (1939), but here the Langejans et al. (2012, Table 2.1) convention is followed. The intertidal zone of the southern Cape rocky shores is subdivided into six distinct zones (Table 2. 1). The Littorina Zone (Splash zone) is the highest zone remaining out of the water except during high spring tide (Stephenson 1939; Langejans et al. 2012; Branch et al. 2022). Just below the Littorina zone or splash zone, the upper Balanoid Zone occurs. This is the zone between the mean sea level (between low tide and high tide) and the sea level during high neap tide. Below this is the Lower Balanoid Zone which is the zone between the mean sea level (between low tide and high tide) and the sea level during low neap tide (Kvale 2006). The lowest intertidal zone is the Cochlear Zone. This is only exposed during low neap tide and low spring tide (Kvale 2006). The Sublittoral/Infra-tidal fringe is directly adjacent to the Cochlear zone and is mostly submerged, except for Spring equinox cycles (a period of a full moon that can be seen during the day and at night) (Da Silva 2004), and during high-pressure cell phenomena (Fu & Pihos 1994; Mather et al. 2009; Chandler & Merry 2010). The Subtidal/Infratidal Zone is the zone which is always submerged (Stephenson 1939; Kilburn & Rippey 1982; Langejans et al. 2012). However, along the southern Cape coast there are 14-days out of the 28-day lunar cycle where the intertidal zone offers high return rates, and apart from nutritional diversity requirements, terrestrial resources are important to sustain good cost to benefit ratios over one month (Langejans et al. 2012). Along the southern Cape intertidal zone, species distribution is rigidly zoned within these areas (Langejans et al. 2012). In the context of this study five different species were collected. T. sarmaticus is found within the Lower Balanoid zone to the Subtidal/Infratidal zone (Yssel 1989; Pulfrich & Branch 2002). However, they are known to migrate between the Balanoid and Cochlear zones (see De Vynck et al. 2020). Perna perna (brown mussel) can be found between the Upper Balanoid and the Subtidal/Infratidal zone, Cymbula 8 oculus (goats eye limpet) is found between the upper and Lower Balanoid, Cymbula miniata (pinkray limpet) is found in the Lower Balanoid, and Scutellastra longicosta (long-spined limpet) is found in the Lower Balanoid (Kilburn & Rippey 1982; Branch et al. 2010: 168; Langejans et al. 2012: Table 5, 85). In this study, as mentioned above, I refer to the South African coastal intertidal zones set out by Langejans et al. (2012) (Table 2. 1). Table 2. 1: Intertidal Zones of Southern Africa and their placement within the intertidal and shallow subtidal zone (adapted from Langejans et al. 2012: Table 4, 84; Branch et al. 2022). 2.3. Habitat of Turbo sarmaticus T. sarmaticus occurs in both warm and cool water (catholic species) (Langejans et al. 2017), and along the southern Cape along the rocky coastlines between the Lower Balanoid and the Cochlear zones to between 8 – 12 m depth in the Subtidal zone (Langejans et al. 2012; De Vynck 2017), with a max of 30 m depth (Yssel 1989; Pulfrich & Branch 2002). As this giant turban snail is abundant, accessible by hand, has a high calorific yield, is easily replenished from the shallow Subtidal/Infratidal zone by Intertidal Zone Placement Tide Levels Littorina Zone Highest Zone High Spring Tide Upper Balanoid Zone High Neap Tide Lower Balanoid Zone Low Neap Tide Cochlear Zone Sublittoral/ Infratidal Fringe Low Spring tide Subtidal/ Infratidal Zone Lowest Zone 9 migration, their occurrence and foraging are associated with the “pantry hypothesis” (De Vynck et al. 2020: 11). The Agulhas Current, originating from the sub-tropical Indian ocean flows south-westerly along the eastern coast of southern Africa and carries warm, nutrient-poor waters from the Equator to the Southern Ocean (Kyriacou 2017). The Agulhas Current results in high species diversity for shellfish, but with a low within- species abundance. The Benguela Current, originates in the Antarctic region of the Atlantic Ocean, flowing in a north-westerly direction along the west coast of southern Africa (Kyriacou 2017). The Benguela current carries cold nutrient-rich waters from the Antarctic to the subtropics (Garzoli & Gordon 1996; Hutchings et al. 2009), with the current resulting in a low diversity in shellfish species, but with high abundance within the individual species (Garzoli & Gordon 1996; Hutchings et al. 2009). On the south and southeast Cape coasts the Agulhas and Benguela confluence results in species diversity and within-species abundance. This is due to the seasonal upwelling of cold water from the Benguela Current, which provides the southern Cape coastline with nutrient-rich waters (Bustamante & Branch 1996; Hutchings et al. 2009; Branch et al. 2022). KRM is situated in the marine province known as the Algoa province. The Algoa province is influenced by the warm Agulhas current and the unpredictable upwelling of cold water of the Benguela current caused by northward winds (Langejans et al. 2017: 62). This is a suitable habitat for a catholic species like T. sarmaticus (Langejans et al. 2017). Along the southern Cape coast, the geological formation of aeolianites along the coast, (Helm et al. 2018) provides a suitable habitat for T. sarmaticus due to the lithified beach sand and is a softer substratum (De Vynck et al. 2020). The aeolianites in modern studies show that there is easier access for foragers to T. sarmaticus in this context (De Vynck et al. 2020). This is because most of the aeolianite shores consist of wide, flat, wave-cut platforms with shallow pools, and are often scattered with large slabs of loose aeolianite (Nami et al. 2016). This causes the Intertidal zones to be less defined as the Cochlear to Balanoid zones are wide, with a narrow or no upper Littorina zone because of the presence of aeolianite cliffs (De Vynck et al. 2020). Table Mountain Group Sandstones (TMS) are found in the three sub-categories: exposed rocky headlands, wave–cut platforms and boulders (De Vynck et al. 2020). TMS has a higher diversity of species targeted by humans. However, the abundance of T. sarmaticus is lower than in aeolianite 10 reefs (De Vynck et al. 2020). The remaining marine habitat types are sandy/beach and estuarine. However, these do not harbour T. sarmaticus (De Vynck et al. 2020). Along the southern Cape coastline between Cape Town and Gqeberha (formerly Port Elizabeth), there is a large record of sea-level change that is associated with the formation of coastal dunes during the Plio-Pleistocene (Carr et al. 2019). These coastal dunes later formed a carbonate-cemented aeolianite due to the shifts in temperature and other possible variables which are still under investigation (Carr et al. 2019; Morrissey et al. 2022). The surrounding coastal shelf is associated with Table Mountain Sandstone (Brenner & Wurz 2019). During glacial cycles, the now–submerged Palaeo–Agulhas Plain (PAP), directly south of the southern Cape coast, was partially to fully exposed. Cawthra et al. (2020) concluded that the PAP intertidal zone was mostly comprised of aeolianite reefs which possibly increased the calorific productivity for humans. De Vynck et al. (2016) showed that the existing aeolianite reefs revealed the highest productivity of all marine habitat types. An actualistic study by De Vynck was conducted to understand the depletion of T. sarmaticus in the intertidal zone, accessible to human foragers in the Still Bay area, 300 km west of Klasies River (De Vynck et al. 2020: 3). The study showed that over a period of 10 months there was no significant depletion of this species due to their ability to migrate from adjacent areas to the intertidal zone. Migration was tested and it was found that replenishment of the intertidal zone did not come from long-shore migration but from the subtidal zone. This means that there is a perpendicular migration to the coast as they migrate from the Subtidal ‘pantry’ (De Vynck et al. 2020), behind the tides up to the Balanoid zone. Ninety-five percent of T. sarmaticus live in the subtidal zone, but the species strives to live in the Cochlear zone (De Vynck et al. 2020). These results were interpreted to indicate a high resilience to persistent foraging of this species during lower pre-historic population densities (De Vynck et al. 2020: 10). 2.4. MSA foraging of shellfish in the southern Cape 2.4.1. Theory For this research, Optimal Foraging Theory (OFT) is used as a broad theoretical background to contextualise foraging patterns. OFT is discussed by Langejans et al. 11 (2012) in the context of shellfish return rates (see also MacArthur & Pianka (1966) and Hawkes et al. (1982) for a broader understanding of the theory). OFT for Middle Stone Age (MSA) shellfish collecting relates to foraging strategies and technological developments that involves the handling costs of specific resources, leading to predictions according to the model (Langejans et al. 2012: 81). In short, OFT investigates caloric cost to benefit ratios. Langejans et al. (2012: 81–82), discuss planning needed for optimal shellfish collecting by MSA foragers. When foraging trips coincided with low tides allowing for a higher return rate due to the most productive intertidal zone being exposed. OFT is used to hypothesise a “pre-historic state” associated with relatively little planning, where shellfish species from higher zones are overrepresented; and for an “evolved state” species from lower zones are well represented (Langejans et al. 2012: 81– 82). Blombos Cave, Pinnacle Point and KRM show Cochlear zone foraging in some Marine Isotope Stage 5 periods which indicates an evolved understanding of the coastal environment (Langejans et al. 2012), as T. sarmaticus (alikreukel) are foraged from the lower zones, the Lower Balanoid and the Cochlear zones (Langejans et al. 2012). OFT in relation to shellfish exploitation further involves return rates, handling, and search costs. The profitability of a species that is exploited depends on two variables: The first variable is the return rate, the absolute energetic value of the shellfish. This can be modelled by ranking the species according to their raw meat weight (kilo-calorific values) (Kyriacou et al. 2014; Kyriacou 2017). The return rate is the rate at which calories are returned by food, per hour including all caloric costs involved in procurement, processing, and transport (the optimal cost-to-benefit ratio) (Kyriacou et al. 2014). Return rate helps with understanding if there was an in-depth understanding of the coastal environment by the foragers (Langejans et al. 2012). The second variable is the handling cost (cost to benefit ratio) (MacArthur & Pianka 1966; Langejans et al. 2012: 82). Handling cost is a composite variable, dependant on the species’ defence mechanisms to avoid predation and on the capabilities of predators concentrated on the habitat along the shore. It is also dependant on the species' abundance and caloric gain (Langejans et al. 2012). These species are assumed to have a higher handling cost (MacArthur & Pianka 1966; Langejans et al. 2012: 82). Search cost looks at the amount of time the forager spends looking for suitable prey, meaning that if a species is rare, it will only be encountered sporadically and will be uncommon in excavated assemblages, even if it is favoured by foragers (Langejans et al. 2012). This means if a species inhabits the 12 lower intertidal zones of the shoreline, they are difficult to obtain because they are predominantly submerged. The correlation between search cost and handling cost looks at the moment at which the forager decides whether the prey is ignored, and the search continues looking for higher-ranked prey. This is according to OFT and the Diet Breadth Model (MacArthur & Pianka 1966). For example, encounter rates with high-ranked prey species are less likely, whereas more low-ranked prey species are likely to be exploited. This is due to the higher-ranking species according to yield being more difficult to forage. However, this situation does not always apply to the collection of shellfish especially T. sarmaticus due to their abundance and are considered a high-ranked species (Langejans et al. 2012). Species are not encountered randomly along the shore but occur in specific zones (Langejans et al. 2012: 82). 2.4.2. Coastal adaptations and shellfish exploitation in the southern Cape A definition of coastal adaption according to Will et al. (2019: 2) is the behaviour associated with hunter-gatherers where the regular use of coastal resources and the formation of settlement along the shoreline is established. The onset of coastal adaption is first suggested at 164 ka at Pinnacle Point (Will et al. 2016, 2019). However, coastal adaptation is fully established along the southern Cape coast between 110 and 100 ka with all the major coastal sites showing coastal foraging on a regular basis within the Cochlear zone and fewer Balanoid species added (Will et al. 2016, 2019), and this occurs also between 130 – 115ka along the western Cape coast (Kyriacou et al. 2015). Coastal adaption according to the definition above is very different to maritime adaptation. Maritime adaptation is a much more recent adaptation occurring during the Neolithic (Lidour et al. 2021: 1). It is the development and diversifying of fishing techniques including the exploitation of marine animal materials that would be harder to forage for (e.g., shark teeth, string ray barbs) for the purpose of both tools and personal adornments, as well as the advancement of seafaring, and colonization of offshore islands due to the increase in the number of people along the coast (Will et al. 2016, 2019; Lidour et al. 2021). The Cape south coast has distinct, biotic (tidal) zones characterised by distinctive fauna, some of which are sought by human foragers (see Table 2. 2). The lowest tides generally expose the larger species such as Haliotis midae (abalone) which has a higher meat yield but low abundance. However, foraging in the intertidal zone near the low-water mark can 13 be dangerous and needs to be timed to coincide with the most favourable part of the daily tidal cycle and the monthly lunar cycle (De Vynck et al. 2020). In the southern Cape, at Pinnacle Point, in the 164ka layers, evidence suggests people foraged along the higher, less productive Lower to Upper Balanoid zones for Donax serra (white mussel), Choromytilus meridionalis (black mussel), Perna perna (brown mussel) and Dinoplax gigas (giant chiton) (Jerardino & Marean 2010; De Vynck et al. 2020: 2; Jerardino & Navarro 2021). By 100ka, prey species such as T. sarmaticus, P. perna, Haliotis spp., and Burnupena spp. from the lowest tidal zone, the Cochlear zone, show up in MSA sites such as Blombos Cave, Pinnacle Point and KRM, signifying ongoing systematic exploitation of shellfish (Jerardino & Marean 2010; Wurz et al. 2022). Apart from the shift in prey species corresponding to deeper infiltration of tidal-zone habitats, the remains of shells increase in density, indicating that MSA people increasingly depended on these resources as a component of their diet from 110ka (Will et al. 2019; De Vynck et al. 2020: 2). Table 2. 2: Common species found within the different Intertidal Zones (adapted from Langejans et al. 2012). Tidal Zone Species Littorina Zone Oxystele spp. Upper Balanoid Zone Cymbula oculus Lower Balanoid Zone Scutellastra longicosta Cochlear Zone Turbo sarmaticus Sublittoral/Infratidal Fringe Perna perna Subtidal/Infratidal Zone Haliotis midae 2.4.3. Nutritional Value Shellfish are nutritionally beneficial and fit the category of so-called functional foods (De Vynck et al. 2020: 14). However, if shellfish lack resilience in the face of predation, their value as a primary food resource is minimal as ongoing foraging could decrease productivity (De Vynck et al. 2020: 2). Resilience is therefore seen as dependence or high productivity over a long period of time. Calorific gain is commonly measured through ethnographic observations combined with a quantitative assessment of the associated harvest. The results are usually expressed in terms of hourly caloric benefit (kcal/h-1) 14 from pursuing various resources (De Vynck et al. 2020: 2). In De Vynck et al. (2020: Table 3, 7) the total calories for all the T. sarmaticus that were collected was 31 148.7 kcal which consisted of 81.9% of the total calorific percentage for the entire collection. This means that one individual T. sarmaticus contributed 19.22 kcal (De Vynck et al. 2020: 7). Langejans et al. (2012: 84) looked at the ranking of species by focusing on the shellfish species that were exploited for nutrition. By combining the data on yield and handling cost, the assumption was made that the shellfish with the highest meat yield would have been preferred during collection (Langejans et al. 2012: 83). A quantitative framework has been set up for broader discussions on the importance and role of shellfish in the diets of MSA hunter-gatherers in two disparate coastal regions (southwestern Cape and Kwazulu-Natal) (Kyriacou 2017: 31). The two coastal regions differ in shellfish diversity due to the difference in ocean currents, nutrients in the water and sea-surface temperature (SST), with the southwestern Cape having a higher species diversity due to the cold waters upwelled by the Benguela current (Kyriacou 2017). The most energy received from shellfish meat derives from the protein and, to a lesser extent, the lipid components of edible soft tissue (Kyriacou 2017: 34). T. sarmaticus is seen to have the highest caloric harvest values compared to all other species harvested (De Vynck et al. 2016: Figure 3, 106). Experiments on intertidal foraging along the southern Cape coast of South Africa suggest that foragers with simple tools could extract sufficient nutrients along stretches of the coast at rates that compare favourably with the returns from terrestrial hunting and gathering observed in modern hunter-gatherer societies (De Vynck et al. 2016: 109). The marine intertidal foraging habitat would often provide higher energetic returns than other foraging activities. Additionally, marine invertebrates yield protein content that is comparable (in weight ratio) to protein from terrestrial fauna. For example, the gonad of a single Scutellastra granularis (granular limpet) yields 8.1 g/100 g of protein, compared to the muscle of a single Kudu which yields 12.9 g/100 g of protein (Kyriacou et al. 2014: Table 3, 68). This means that while meeting energetic needs, marine intertidal foragers could also easily deliver sufficient protein in the diet for good health, and may need to complement the lean protein shellfish diet with carbohydrate or lipid-rich resources to avoid negative consequences of a too-high protein diet (De Vynck et al. 2016: 109). 15 2.5. Turbo sarmaticus operculum size and the ‘predation hypothesis’ Turbo sarmaticus operculum size gives insight into the life cycle of this species (Steele & Klein 2008). T. sarmaticus grows in two stages. The first stage is the first few years of the species' lifecycle where it grows rapidly in size and the second stage is where the growth rate of the species slows down (Yssel 1989). For example, the shell length will be between 24 mm – 72 mm for the first three years and will be more than 112 mm after nine years (Yssel 1989). In a study on various coastal sites along the Cape coastline comparing the T. sarmaticus opercula size of MSA to Later Stone Age (LSA) coastal occupation sites, Steele & Klein (2008, Klein & Steele 2013) showed that T. sarmaticus opercula differ in size between the LSA (median 26.3 mm) and MSA (MSA I median 37.2 mm, MSA II median 39.4 mm, and MSA III median 42.3 mm) periods at KRM. (Klein & Steele 2013: 10913). The median for T. sarmaticus opercula length during MSA II, the period during which layer SMONE was formed was thus 39.3 mm (see also Steele & Klein 2008). Steele & Klein (2008) argue that the larger opercula from the MSA sites are due to MSA humans not exploiting shellfish as intensively as LSA humans and therefore, the shellfish were able to develop to full size. LSA humans however are thought to have a higher population density and individuals who understood the coastal environment better and were, therefore, able to exploit more of the shellfish continuously, and as a result the shellfish were unable to develop to full size (Steele & Klein 2008; Klein & Steele 2013; Klein & Bird 2016). 2.6. Turbo sarmaticus operculum size, growth patterns and the ‘climatic and environmental hypothesis’ The following section discusses factors other than predation that might influence opercula size. An isotopic study looked at the increments of the complex spirals seen on T. sarmaticus opercula (Figure 2. 3), to understand the growth patterns of the individuals (Galimberti et al. 2017). The opercula did not show any clear annual or sub-annual bands (Sealy & Galimberti 2011; Galimberti et al. 2017), as the opercula grow at a slow rate of around ca. 9 mm per/year during the first three years, slowing to ca. 2 mm per/year thereafter for the remaining years of their 10–year lifespan (Foster 1997; Galimberti et al. 2017). T. sarmaticus growth is slowed during periods of extreme temperatures (Galimberti et al. 2017: 65–66). During the months of winter, when the water 16 temperature is colder (drops to lower than 14⁰C in the Mossel Bay region) the growth of the operculum and shell is slowed. The same is seen in the hottest months when the temperature of the water is greater than 23,5⁰C (Sealy & Galimberti 2011; Galimberti et al. 2017). Therefore, it is believed that the sea surface temperature (SST) would have a direct effect on the T. sarmaticus growth and would consequentially affect the size of the operculum, showing that the climate and environment play a role in their growth. The size of T. sarmaticus and thus the size of the opercula is dependent on the environmental conditions and larger T. sarmaticus are generally associated with increased nutrient availability (Sealy & Galimberti 2011). An increase in SST would have a slowing effect on T. sarmaticus, as the warm water increases the metabolic rates of the mollusc (Sealy & Galimberti 2011; Langejans et al. 2017), resulting in a decrease in the size of the opercula due to the slowed growth rate. There is an increase in productivity in colder SST due to the increase in nutrients from algae and plankton (Langejans et al. 2017). Figure 2. 3: Spirals on T. sarmaticus operculum (adapted from Galimberti et al. 2017: Fig.1, 63). A study was conducted that documented the difference between the growth rates and reproductive fitness of T. sarmaticus which were kept in a controlled tank environment (Foster et al. 1999). The oceanic productivity, of organic matter by phytoplankton, helps 17 generate food for shellfish, and variations in oceanic productivity affect the marine food chain and may affect the size and distributions of shellfish species (Sealy & Galimberti 2011). The results showed that there could be a regional difference in the growth of T. sarmaticus due to the different algae species affecting the growth rate (Foster et al. 1999). 2.7. Seasonality and Turbo sarmaticus Seasonality is a key dimension of subsistence behaviour based on the acquisition of resources and the occupation of sites (Loftus et al. 2019). In southern African archaeology, a framework on seasonal mobility has been used to explore whether hunter- gatherers exploited between coastal and inland sites due to fluctuations in seasonality and understand the different resources found between the different ecozones (Parkington 1976). The seasonal occupation of coastal sites compared to year around occupation. Loftus et al. (2019) provided a perspective of shellfish foraging seasonally along the southern Cape coastline in the context of existing detailed coastal archaeological sites and records. Oxygen isotopes from marine mollusc shells can be used to reconstruct the seasonality of the sea surface temperature (SST) during the period of shell growth and seasonal climate changes to the environment (Loftus et al. 2019). Loftus et al. (2019) conducted the oxygen isotope study using T. sarmaticus opercula from five archaeological sites (Nelson Bay Cave, Robberg Cave, Byneskranskop 1, Klasies River Main Site and Pinnacle Point) along the southern Cape coastline within the last interglacial cycle. The T. sarmaticus opercula from the five assemblages showed visible growth increments on the flange surface, due to the species location in the intertidal zone, it can be used as a good indicator for SST and for palaeotemperature reconstruction as they are mostly always submerged (Loftus et al. 2019). Semi-annual foraging patterns based on δ18O analysis for KRM (MSA II Lower) T. sarmaticus opercula showed more warm water (summer and autumn) (n=7) foraging patterns but not a big difference between the cool water (winter and spring) (n=5) with the quarterly season mainly showing autumn (n=4), spring (n=3) and summer (n=3) (Loftus et al. 2019: 9). The majority of the MSA coastal sites showed evidence of foraging of T. sarmaticus during summer and spring compared to the LSA coastal sites where spring and winter were much more evident (Loftus et al. 2019). 18 The analysed assemblage from KRM during Loftus et al. (2019) study showed the fewest T. sarmaticus were harvested during winter, which is the least optimal foraging season. The KRM assemblage showed a strong seasonal signal with nearly 50% of the assemblage being foraged during spring and notably less during summer, the optimal foraging period (Loftus et al. 2019). 2.8. Previous work on experimental processing of Turbo sarmaticus De Vynck et al. (2016: 102), described caloric return rates of intertidal foraging within an actualistic context along the Still Bay coastline. They found that the variables that could affect foraging returns include, the individual forager’s ability, and tidal level (meters). The tide is a synthetic controlling variable, which can affect the foraging success, like an ocean swell, wind, and rain (De Vynck et al. 2020: 5). Statistical analysis was used to investigate the effect of tidal level, habitat type and weather patterns, as well as age and sex of the forager to measure shellfish return rates (De Vynck et al. 2016: 102). Overall, the harvest was mainly comprised of T. sarmaticus (56.1%) (De Vynck et al. 2016: 102). In the experiments undertaken by De Vynck (2017), non–metal tools (bone tools and wooden sticks) were created to forage and process shellfish by indigenous coastal foragers. The 2017 methodology, similar to those used in this study and described in Chapter 3, was informed by the indigenous coastal foragers, due to their knowledge and experience with coastal foraging and processing of the shellfish species, as well as their experience with being in the intertidal zones and producing tools that would be successful during the process. The collection of material for tool production was done before the experiments, as bone and wood used needed time to dry before tool production could occur. The production of the tools was done along the protruding rocks along the Still Bay coast and only once the tools had been produced, the foraging could take place. The foraging for the shellfish took place at spring low tide and was done by experienced foragers from the local indigenous communities. After the foraging had taken place, the processing occurred in a similar style as discussed in De Vynck et al. (2016: 109). Both the raw and cooked processing was done by the same foragers after foraging was completed, and extraction of the edible portion was done after cooking and for some, the extraction was done raw (De Vynck 2017: Appendix 3). T. sarmaticus were usually cooked in batches of 15–20 shells placed directly on the coals (Figure 2. 4). The 19 cooking occurred in batches, as this way, the mean cooking time required was only 27 seconds per item. Extracting the meat from the shell (Figure 2. 5) and removing the guts before consumption took 18 seconds per item on average (no variance was measured because numbers are calculated from a single large batch) (De Vynck et al. 2016: 109). T. sarmaticus did not require constant attention but cooking instead took place while the foragers were conversing, socialising and repairing tools (De Vynck et al. 2016: 109). In this study, these experiments are expanded on, and the tools produced for the De Vynck (2017) experiments are replicated. Figure 2. 4: Cooking of T. sarmaticus on the coals (De Vynck 2017: Fig. A 3.15a, Appendix 3). 20 Figure 2. 5: Raw extraction of the meat from T. sarmaticus shell with a bone tool (De Vynck 2017: Fig. A3.12, Appendix 3). 2.9. Turbo sarmaticus opercula and site formation Taphonomy is the study of the biological, physical and chemical processes that change organisms after death (Lyman 1994). As there are no published papers that have studied taphonomic aspects of T. sarmaticus opercula, literature relating to Turbo undulatus (Common warrener) was used to help understand the taphonomic characteristics of shellfish within the same genus (Sherwood et al. 2016; see also Holmes 2020). T. undulatus opercula are found along the southern Australian coastline in Holocene- occupied sites. Sherwood et al. (2016) investigated the differences between anthropogenic, naturally occurring, and coastal gull middens. Anthropogenic middens [or in the case of Sherwood et al. (2016) Aboriginal middens] are defined as middens that have been created by human collection and consumption of shellfish and feature a bias towards larger shells found in the midden due to the prey selection. This means that larger opercula will be found at the site predominantly due to the selection of larger individuals (Sherwood et al. 2016). Naturally occurring middens are beach deposits that occur due to storm surges that wash the opercula to the shore. In beach deposits, a wide variety of sizes are seen with a higher proportion of smaller opercula. The opercula from a beach deposit will show evidence of water-wear or rounding of the edges due to the 21 abrasion of the sand and waves (see Chapter 4: Analytical Methods) (Sherwood et al. 2016). Human depositional middens or cultural middens are formed at occupation sites where the processing of shellfish took place (in the case of KRM, the at the caves). In South Africa, few researchers have looked at understanding how cultural middens form and the possible taphonomy that goes along with the human interaction with shellfish and opercula (cf. Sherwood et al. 2016; 2018). A study done in southern Australian on early and late Holocene Aboriginal coastal occupation sites looked at understanding the taphonomy left on the shells of T. undulatus. However, there were no conclusive differences between the cultural middens’ taphonomy and that of a coastal gull’s midden. Burning was the only attribute that distinguished between the human and natural midden (Sherwood et al. 2016; Sherwood et al. 2018). 2.10. Conclusion In this chapter, the literature on KRM and the context of the material analysed have been discussed. The literature on previous actualistic work associated with T. sarmaticus, the understanding of the influences of exploitation, coastal adaptations, the environment and the factors of SST and the presence or absence of nutrients that can affect the growth of T. sarmaticus and their operculum were addressed. In the next chapter the methodology and materials used in the actualistic study conducted for this paper is discussed. 22 Chapter 3: Experimental Materials and Methods 3.1. Introduction An actualistic study was undertaken to investigate the subsistence behaviour and taphonomic marks that are left on T. sarmaticus from the MSA II layer SMONE at KRM. In this chapter, the experimental methodology is described involving the foraging practice and methods used for processing. 3.2. Foraging The foraging experiment to collect and process Turbo sarmaticus (alikreukel), and three other shell species, took place at Jongensfontein, Still Bay, Western Cape, South Africa (Figure 3. 1). Three different locations were visited (Figure 3. 2). Experimentation took place over a period of seven days, between the 18 August 2021 and the 25 August 2021 (late winter), during a period of spring tide (full moon). Tidal variation occurs at several levels based on the monthly lunar cycles (spring and neap tides) and daily lunar cycles (low and high tides). The maximum monthly lunar cycle variation is classified into spring and neap tides, which are driven by a relative lunar position to the sun (Marean 2010: 432). Field research commenced after the habitat, the correct ethnographic replication of tool production and foraging, and the materials required were studied. To conduct the foraging research, a Research fishing permit was obtained from the South African Department of Environmental Affairs [Appendix 1D.1: Research Permit (RES2021–85)] which allowed the collection of 150 individual T. sarmaticus that were greater than 55.5 mm in shell length. In addition, an annual Fishing licence [Appendix 1D.2: Additional Fishing Permit (1052834)] was bought from the Still Bay Post Office, allowing for the collection of additional mollusc species and an additional five individual T. sarmaticus. 23 Figure 3. 1: Jongensfontein, where the experimental foraging took place, in relation to the southern Cape coastal archaeological sites Klasies River main site (KRM), Blombos Cave (BBC) and Pinnacle Point (PP) (Google Earth Pro 2021a). Figure 3. 2: Foraging locations for Still Bay foraging experiment, A: Test foraging day at Jongensfontein East, 1: Lithics foraging at Jongensfontein West, 2: Wooden tool and Bone tool foraging at Brandersfontein Farm (Google Earth Pro 2021b). 24 3.2.1. Participants and Prey For the purposes of this study, Dr Jan De Vynck is referred to as participant 1 (P1) and I, Carl Holmes, as participant 2 (P2). P1 has experience through observations of local coastal foragers and had learnt foraging and tool–making techniques from them (De Vynck 2017). Before the experiment P2 had no foraging experience or experience making tools. P2 had to learn while foraging and could be considered to have the same skills as an adult with no foraging experience in a Middle Stone Age (MSA) context. For the foraging, we targeted five intertidal species commonly found within the archaeological record along the southern Cape coast. The five species were targeted as they probably would have foraged at the same time in the past: T. sarmaticus (Figure 3. 3A and B), Perna perna (brown mussel) (Figure 3. 3C), Scutellastra longicosta (long- spined limpet) (Figure 3. 3D), Cymbula oculus (goat’s eye limpet) (Figure 3. 3E) and Cymbula miniata (pink ray limpet) (Figure 3. 3F), with the processing focusing mainly on the T. sarmaticus. Figure 3. 3: A, the dorsal side of a T. sarmaticus shell, B, the ventral side of the T. sarmaticus shell with operculum in place. C, of an intact P. perna shell. D, of the ventral side of the S. longicosta shell. E, of the ventral side of the C. oculus shell. F, of the ventral side of the C. miniata shell. 25 The opercula from the T. sarmaticus are the focus of the study. The opercula are hard round/oval calcium carbonate disks carried on the foot of the species. The T. sarmaticus opercula have a rigiclaudent spiral with multiple spirals and the shape of the opercula fitting the opening (Checa & Jiménez-Jiménez 1998), where the spirals rotate in a clockwise direction. The opercula serve the purpose of protecting the molluscs head near the shell aperture (Checa & Jiménez-Jiménez 1998). The T. sarmaticus opercula consist of two sides, the dome (ventral) and the flange (dorsal) (see Figure 3. 4). Figure 3. 4: A, dome (dorsal) surface of the T. sarmaticus opercula. B, the flange (ventral) surface of the T. sarmaticus opercula. 3.2.2. Habitat The foraging of the shellfish took place between the Lower Balanoid and the Cochlear zones during low tide (Table 2. 1) (Stephenson 1939), corresponding to the accessibility of the tidal zones during the spring low tide (Marean 2010: 432). The geology of the Still Bay coastal landscape is dominated by Table Mountain Sandstone (TMS) consisting of boulders, exposed rocky headlands, and wave-cut platforms (De Vynck 2017: 7; Cawthra et al. 2020). 26 3.2.3. Foraging Tools Three different types of tools were used over three days of foraging during low tide. One tool type was used per day: MSA lithics from the spill heap at KRM (Day 1), experimentally made wooden tools (Day 2), and experimentally made bone tools (Day 3). Production of foraging tools Lithics The lithics used for the foraging experiment were MSA lithics recovered out of context from the spill heap at Klasies River Main Site (Figure 3. 5). The lithics were described by Sarah Wurz to provide a clear understanding of the types of lithics that were used and selected by each participant. The lithics in this experiment consisted of MSA points, blades, and core edge (débordant) flakes and blades, with the raw material being quartzite for all the lithics (pers. comm Wurz). Figure 3. 5: Out of context Middle Stone Age lithics from the Klasies River main site spill heap after use, showing breakage on the lithics as a result of the experiment. 27 Wooden tools The Wooden tools were produced by P1 and P2 from Olea europaea ssp. africana (Wild Olive) branches which were collected from the Still Bay area. This species is the wood of choice used by modern day local communities along the southern Cape coast for the making of wooden tools, due to its accessibility and density of the wood allowing it to be reused (De Vynck 2017, pers. comm De Vynck). The production of the wooden tools took place at Morris Point above the Still Bay Harbour (34⁰25’13,22”S, 21⁰21’49,40”E). The type of wooden and bone tools was based on replicated tools that P1 observed local community members make during his PhD research (De Vynck 2017). This was the basis for the production of the tools in this dissertation. Before the production of the tools, the branches were measured (using a standard 3m tape measure) to equal 0.45 m (Figure 3. 6A) and were then cut using a hand saw (Figure 3. 6B). Pre-historic tool producers would have cut the lengths with lithics or placed lengths in a fire to shorten the branches. We felt that there was no need to replicate either cutting method as cutting with a saw had zero effect on the outcome of the experiment. The wooden tools were produced by scraping the wood on TMS rocks with quartz veins running through the dominating quartzite base geology (Figure 3. 7, Figure 3. 8 and Figure 3. 9). Three wooden tools (Figure 3. 10) were produced by each of the participants, with the time being recorded for how long the participants took to produce each tool. 28 Figure 3. 6: A, the measuring of the Olea europaea ssp. africana branches to 0.45 m. B, the hand saw used to cut the branches before the production of the tools, and C, the cut Olea europaea ssp. africana branches for both participants. Figure 3. 7: Wooden tool production by P2 using the quartzite veins on the Table Mountain Sandstone rocky headlands. 29 Figure 3. 8: Worksite after the tool production, in the image the quartzite veins are protruding from the Table Mountain Sandstone exposed rocky headlands. Figure 3. 9: Worksite after the tool production, in the image the Table Mountain Sandstone exposed rocky headland surface shows the site used to refine and sharpen the tool where no quartzite veins are present. 30 Figure 3. 10: Wooden tools made by P2 out of Olea europaea ssp. africana branches. Bone tools Four Tragelaphus strepiceros (Greater Kudu) long bones (1: humerus, 2: femur, 3: radius and ulna, 4: femur, see Figure 3. 11) from the local butcher in Still Bay were used to construct the bone tools. Two long bones were baked in the oven for 60 minutes at 200⁰C, and the remaining two bones were kept raw. Cooking bone before production was conducted under the assumption that marrow was important for MSA foragers to add to the optimal caloric gain per energy package (individual animals) (Hawkes et al. 1982; Hawkes & O’Connell 1992). Also, heat-treated bone is potentially denser and more 31 effective as a tool than raw bone (Bradfield 2015). The bones were then placed on the roof to dry for a week. Figure 3. 11: 1, T. strepiceros humerus (baked). 2, T. strepiceros femur (raw). 3, T. strepiceros radius and ulna (baked). 4, T. strepiceros femur (raw). The production of the bone tools took place near Still Bay in a pristine intertidal area (34°24'43.80"S, 21°22'49.40"E), along TMS exposed rocky headlands. Each participant received one cooked T. strepiceros long bone and 1 raw T. strepiceros long bone to produce as many bone tools as they could (Figure 3. 11). The hammer and anvil technique were used by both participants, with small cobbles and rocks found near the production site, to crack open the long bone (Figure 3. 12) (De Vynck 2017), so the marrow could be removed. Then the participants used the fragment from the long bone to produce the tools, by rubbing the edge of the bone fragment on the TMS rocks until the participant was satisfied with the tool produced and they were similar to the ethnographic observations (point, blade, or scraper) (Figure 3. 13) (De Vynck 2017). Tool production time, per tool and per participant, was recorded. Most of the bone tools were produced from the baked bone (Figure 3. 14), as the raw bone fragmented into small unusable fragments (Figure 3. 15B). 32 Figure 3. 12: Cobblestones used for the hammer and anvil technique to crack open the long bone. A is the Cobblestone used by P1, and B is the Cobblestone used by P2. Figure 3. 13: A, the area where P2 produced the bone tools. B, the remains left behind postproduction (bone marrow, skin, bone fragments). 33 Figure 3. 14: A, bone tools made by P2 out of T. strepiceros long bones. B, bone tools made by P1 out of T. strepiceros long bones. Figure 3. 15: A, the tool produced by P2 from the baked long bone. B, the bone fragments produced by P2 from the raw long bone. 3.2.4. Foraging Process Before the foraging took place each day, the following data were recorded as affecting variables (Appendix 1A Table 1), the Geographical Positioning System (GPS) coordinates for the start and stop location of the foraging patch (using a Garmin eTrex®10 handheld GPS), the habitat type, the swell size, period and direction (Windy.com 2011), the low tide and high tide times and tidal heights (tides4fishing.com 2021), the phase of the moon, the 34 wind direction and speed (Windy.com 2011), the aspect of the shoreline, the ocean condition rating (Table 3. 1), and the water temperature (Windy.com 2011). Table 3. 1: Ocean Condition rating scale, calibrated to the Beaufort scale, showing the optimal and suboptimal conditions for shellfish foraging. Ocean Condition Rating Beaufort Scale Wind (kph) Swell (m) Status 1 0–2 0–11 0–0.3 Good 2 3 12–19 0.4–1 Good 3 4–5 20–38 1.1–2.4 Average 4 6 39–50 2.5–5.8 Poor 5 7–12 51+ 5.9+ Unsafe The Ocean Condition rating (Table 3. 1) was correlated to the Beaufort scale rating system (De Vynck 2017), based on wind speed, and swell size. These conditions affect the condition of the sea and the ability of a forager to successfully forage in the intertidal zones. The Ocean Condition Rating is ranked on a score of 1 (the best) to 5 (the worst). Foraging time was recorded to measure caloric return rates (Kcal/Hr-1) related to the effectiveness and the effects on the shell damage of the various foraging tool media. 30 individuals of T. sarmaticus were collected per participant, 10 of each of the limpet species (S. longicosta, C. oculus, and C. miniata) and 20 P. perna. Each participant used ethnographic observations (De Vynck 2017), to conduct foraging methods within the intertidal zone and foraging time elapsed when the above volumes per species were collected per foraging event. The foraging tools were used to extract the prey species from the intertidal zone as per ethnographic observations. 3.2.5. Processing Processing methods of T. sarmaticus were consistent throughout the research experiment. Each day processing took place immediately after foraging was completed. The foraging tools were used during the processing with the aim to measure the operculum damage variations between each tool. T. sarmaticus operculum damage was investigated subsequently in the laboratory. The first step in the processing commenced by weighing out 10 kg of Acacia cyclops (Rooikrans) firewood (using a Pelouze® Rubbermaid 45 kg x 50 g digital hanging scale) and lighting the fire within a container (one fire was made per day, and the fire was shared by both participants) (Figure 3. 16). 35 Pre-historic replications of cooking shellfish directly on the fire embers (De Vynck 2017). However, T. sarmaticus were placed on a grid above the fire to ensure practical field research methods for allometric conversions and did not result in a variation from pre- historic cooking results (pers. comm De Vynck). Total cooking time was recorded for each fire each day from when the fire was started to when the fire was ready for cooking and cooking time was recorded as a variable affecting the total cost–benefit ratios of caloric gain (De Vynck 2017). Figure 3. 16: Fire for cooking T. sarmaticus. Raw and Cooked Processing Processing T. sarmaticus was twofold: raw extraction and cooked extraction of edible portions using the three types of tools. The processing started by recording the following information for both participants: the total weight of the combined foraged shellfish (using an AWS-SR-20 20 000 g x 10 g digital hanging scale), along with the total count for each species individually (T. sarmaticus, P. perna, S. longicosta, C. oculus and C. miniata), and the weight and count for the raw and cooked T. sarmaticus edible portions (Appendix 1A.2. Processing Data). Raw Processing For the processing of the raw T. sarmaticus, the following was recorded during the processing for each shell by each participant: the shell length in millimetres (Figure 3. 36 17A), the shell width in millimetres (Figure 3. 17B), the operculum length in millimetres (mm), the operculum width in millimetres, the gross weight in grams (g) (using an AWS- SR-5 5 000 g x 5 g digital hanging scale), the edible portion raw weight in grams (using an AWS-70 70 g x 0.01 g digital pocket scale), the gonad weight in grams (using an AWS- 70 70 g x 0.01 g digital pocket scale), the gonad colour, and the shell colour (pos