- Email: [email protected]
SW Ethiopia
Quaternary International xxx (2017) 1e10
Contents lists available at ScienceDirect
Quaternary International journal homepage: www.elsevier.com/locate/quaint
Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia Ralf Vogelsang*, Karl Peter Wendt Institute of Prehistoric Archaeology, University of Cologne, Bernhardt-Feilchenfeldstr. 11, 50969, Germany
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 January 2017 Received in revised form 19 June 2017 Accepted 28 June 2017 Available online xxx
Although high-altitude mountain habitats are often regarded as unfavorable for human occupation (e.g. Aldenderfer 2014); on the other hand tropical highlands in Africa are suggested as potential refugia during times of environmental stress (e.g. Basell 2008; Brandt et al. 2012). Archaeological investigations on Mount Damota (2908 m a.s.l.), located on the boundary between the Southwest Ethiopian Highlands to the west and the southern Main Ethiopian Rift valley to the east, yielded a large number of archaeological sites from the Middle Stone Age period until historical times. In this paper we try to reconstruct settlement models for the late Pleistocene and Holocene occupation in this area and speculate about potential land use patterns. Such complex topics demand a landscape archaeological approach that includes open-air sites and rock-shelters. The results from our excavations at Mochena Borago Rock-shelter and evidence from open-air-sites that were recorded during intensive surveys on the slopes and plateau of the mountain, allow a first reconstruction of the settlement history of the area. © 2017 Elsevier Ltd and INQUA. All rights reserved.
Keywords: Ethiopia Mount damota Middle stone age (MSA) Later stone age (LSA) Settlement pattern
1. Study background 1.1. Setting Although high-altitude mountain habitats are often regarded as unfavorable for human occupation (e.g. Aldenderfer 2014); on the other hand tropical highlands in Africa are suggested as potential refugia during times of environmental stress (e.g. Basell 2008; Brandt et al. 2012). With a maximal altitude of 2908 m a.s.l, Mt. Damota is a perfect case study for the verification of these contradictory conceptions. The study area is located 320 km south of Addis Ababa, where the southwest Ethiopian highlands intersect the Main Ethiopian Rift Valley near the town of Sodo (Fig. 1). Mt. Damota's steep upper slopes, deep gorges, gentle lower flanks and other physiographic features contribute to variation in vegetation within short distances. As a result, numerous vertical ecozones are compressed within a small geographic area on and around Mt. Damota. Like other regions of the Southwest Ethiopian Highlands (Hildebrand et al., 2010), the area around Mochena Borago is characterized by
* Corresponding author. E-mail addresses: [email protected] (R. Vogelsang), [email protected] (K.P. Wendt).
high biodiversity and an abundant supply of natural and domesticated faunal and floral resources for exploitation by contemporary and past human populations (Lesur et al., 2007; Brandt et al., 2012). This tight variation in altitude e and consequently in rainfall and vegetation e make the study area an excellent setting in which to examine changing human landscape preferences related to elevation. 1.2. Landforms Mount Damota is a dormant trachytic volcano comprised mainly of Pliocene porphyritic, anorthoclase-phenocryst trachytes and was active from at least 2.94 Ma until the Late Quaternary (Woldegabriel et al., 1990; Chernet, 2011; Corti et al., 2013). Rising steeply from its surrounding flanks (1300e1600 m a.s.l.), Mt Damota's a summit (2908 m a.s.l.; Fig. 2A) offers striking views of the central Main Ethiopian Rift Valley lakes to the distant north, the Bilate River and the southern Main Ethiopian Rift Valley to the east, Lake Abaya to the south, the Gibe and Omo River valleys to the southwest and the Wolayta/Hadiya Highlands to the West. The topography in the study area is dominated by steep slopes and rocky gorges in the mountain ranges (Abbate et al., 2015) and by lower relief topography in the transitional zone descending to the Rift Valley floor (Fig. 2C). Due to orographic effects, mean annual rainfall is significantly
http://dx.doi.org/10.1016/j.quaint.2017.06.061 1040-6182/© 2017 Elsevier Ltd and INQUA. All rights reserved.
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
2
R. Vogelsang, K.P. Wendt / Quaternary International xxx (2017) 1e10
Fig. 1. Location of Mount Damota, Mochena Borago Rockshelter (MB), the two known Humbo area obsidian sources and the survey area (detailed map shown in Fig. 4).
higher on Mt. Damota compared to the surrounding plains and the southern Main Ethiopian Rift Valley. Precipitation amounts to 2000 mm p.a. with half of the annual precipitation falling from June to September (Fisher, 2010; Viste and Sorteberg, 2013).
1.3. Ecological zones Population pressure, intensive farming and heavy erosion have impacted Mount Damota's natural landscape such that natural vegetation is only found in and near ravines, rocky outcrops, steep slopes and other areas too difficult to settle, plow or hoe. It is therefore difficult to make direct observations of ecological variation. Three of the major agro-ecological zones recognized by Ethiopian farmers on the basis of altitudinal changes in precipitation, temperature, soils and crop suitability (Hurni, 1998, pp. 18e19) are found on and around Mt. Damota: dega, woyna dega and kola (Fig. 3AeC). Natural floral distributions seem to parallel these
traditional agro-ecological zones. At ~2200 m asl, Mochena Borago Rockshelter is within the upper elevations of woyna dega, with bamboo growing nearby (Brandt et al., 2012). Thirty km SE of Mt. Damota, Ethiopia's largest rift lake, Lake Abaya, lies at 1169 m a.s.l. (Fig. 3 D). Ringed by freshwater marshes, Lake Abaya covers more than 1100 km2 and forms another significant regional biotope (Awulachew, 2006).
2. Archaeological research This study incorporates data from both open-air surface sites and rockshelters, which provide complementary records of human behavior and settlement patterns. Open-air surface sites are often numerous and can represent many distinct specific activities (e.g. settlement, hunting/butchering, raw material procurement). Rockshelters, while much rarer, carry several advantages. Their sealed contexts, while not assuring undisturbed contexts for finds, have much lower probability of mixed assemblages than surface
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
R. Vogelsang, K.P. Wendt / Quaternary International xxx (2017) 1e10
3
Fig. 2. View of Mount Damota (A), Mochena Borago Rockshelter (B), view from Mount Damota on the transitional zone descending to the Rift Valley floor (C).
sites. Also, because rockshelters offer protection from natural forces, they have often repeated episodes of use and sedimentation, leading to preservation of stratified sequences. Consistent low temperatures and protection from rainfall aid the preservation of bones, macrobotanical remains and charcoal, which are essential for both radiocarbon dating and paleoenvironmental reconstruction. Rockshelter sequences therefore provide crucial benchmarks for the chronological classification and environmental contextualization of assemblages from open-air sites. 2.1. History of research Three decades of archaeological survey and excavation have laid the groundwork for assessing human landscape use in the study area. In 1995, GEPCA (Groupe pour l’Etude de la Protohistoire de la Corne de l’Afrique) began a survey of the Wolayta region of SW Ethiopia directed by R. Joussaume. Between 1998 and 2001, GEPCA's extensive excavation of Mochena Borago Rockshelter (Fig. 2 B), directed by X. Gutherz, focused on mid to late Holocene
occupations but also revealed ~0.8 m of Later Pleistocene deposits in a 1.5 m2 test unit (Gutherz, 2000; Gutherz et al., 2002). In 2006, the Southwest Ethiopian Archaeological Project (SWEAP), directed by S. Brandt and E. Hildebrand, initiated Pleistocene-focused research at Mochena Borago. Since 2009, this work has continued under the co-direction of S. Brandt and R. Vogelsang within the framework of the University of Cologne's Collaborative Research Center 806 (CRC 806, http://www.sfb806.uni-koeln.de). CRC 806/ SWEAP carried out five additional fieldwork seasons between 2010 and 2014 (Brandt et al., 2012), including excavations of new areas within Mochena Borago. 2.2. Mochena Borago Rockshelter Mochena Borago rockshelter is situated on the southwestern flanks of Mount Damota at ~2200 m asl. (Fig. 2 B). Excavations of five different trenches yielded complex stratigraphies. Geoarchaeological analyses indicate a multiphase, polygenetic accumulation and erosion of sediments deposited through natural and
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
4
R. Vogelsang, K.P. Wendt / Quaternary International xxx (2017) 1e10
Fig. 3. Ecological zones on and around Mount Damota.
anthropogenic processes. The sedimentological evidence points to multiple occupations of the shelter mainly during rather humid conditions (Meyer et al., submitted). The Late Pleistocene layers span a period from >49 ka to ~36 ka. Fifty-nine radiocarbon dates on charcoal make Mochena Borago one of the best-dated archaeological sites from this period in eastern and northeastern Africa (Brandt et al., 2017). Pleistocene deposits in the main excavation area (BXA) consist of a sequence of six major lithostratigraphic groups incorporating four occupational episodes (Lower T-Group, Upper T-Group, SGroup, and R-Group). Preliminary descriptions of assemblages from T-Group, S-Group (Brandt et al., 2012) and R-Group (Brandt et al., 2017) are briefly summarized here, because the sequence is an
important reference point for a detailed chronological classification that cannot be achieved by surface assemblages found during survey. Remarkable is the continuity of technological and typological attributes of the lithic assemblages dating between ~49.4 and ~36.6 ka (Upper T-Group, S-Group and R-Group). Archaeological assemblages consist mainly of obsidian flaked stone artefacts (>96%). Other rocks, such as chalcedony, rhyolite and different cryptocrystaline silicates are represented only by single pieces. Such less frequently used rocks, mainly rhyolite, are with 7.5% more prominent only in the Lower T-Group assemblage that is older than 50 ka. The dominance of obsidian in the raw material spectrum is not surprising seeing that the Humbo area, a huge obsidian outcrop, is
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
R. Vogelsang, K.P. Wendt / Quaternary International xxx (2017) 1e10
5
Fig. 4. Survey-routes with the location of survey sites. MSA ¼ red, LSA ¼ yellow, historical ¼ orange.
situated only 20 km southeast from Mount Damota. First results of XRF analysis of obsidian artefacts and raw material pieces indicate that more than 90% of the obsidian artefacts are produced from this source. Although raw material availability seems not to have been a delimiting factor, artefacts in all assemblages are very small. Remarkable is the longstanding maintenance of many core reduction strategies, such as centripetal cores (Levallois- and radial cores) and a heterogeneous spectrum of other flake- and blade cores (single-, double- and multiple platform cores, bipolar cores, irregular cores) that coexist in all assemblages. However, in many cases the extreme reduction of the cores complicates the determination of their former shape. Characteristic retouched artefacts are facially retouched tools and backed pieces that are present in all inventories. However there is a clear change in proportion from a predominance of facial retouch in the older assemblages (Upper and Lower T-Group) to a predominance of backed pieces in the two younger strat-units. Unifacial and bifacial points are typical, but not all facially retouched tools can be classified as points. Especially in the TGroup assemblages there is a broad spectrum of facially retouched tool types that needs further investigation. The discovery of backed pieces in the deepest levels of Lower T-Group's undated, but older than 50 ka sediments suggests that this hafting strategy was practiced in SW Ethiopia when hunter-gatherers began their early MIS 3 range expansion (regarding the significance of backed tools for hafting and composite technology, and as an argument for symbolic behavior see e.g. Wurz, 1999; Barham, 2002; Villa and Soriano, 2010; McBrearty, 2012). In general, a longstanding continuity in the technological and typological attributes of all assemblages is remarkable considering the considerable climate oscillations of MIS 3 (e.g. Tiermey et al., 2008; Foerster et al., 2012), punctuated by major, possibly
catastrophic volcanic events (Abebe et al., 2007; Corti et al., 2013; Benvenuti and Carnicelli, 2015). 2.3. Open-air surface sites Extensive archaeological surveys in the surroundings of Mochena Borago were conducted concomitant to the excavations in the rockshelter. The survey strategy is best described as a “stratified random sampling” (Shafer, 2016). Pedestrian surveys attempted to include all different landforms of the area, such as slopes, steep valleys and plateaus, in different altitudes from the foot of the mountain (~1900 m a.s.l.) to the summit (2900 m a.s.l). For this reason, survey routes followed not only currently used pathways that are often along ridges. Some routes also crossed the flanks of the mountain horizontally, thus covering steep slopes and gorges. Surveys between 2010 and 2014 summed up to a total distance of 130 km. Sixty three archaeological sites were located (Fig. 4). Diagnostic finds e which range from “Middle Stone Age” (MSA) and “Later Stone Age” (LSA) technological elements to historic artefacts e were collected and suggest recurring use of this area over the past 50,000 years. Forms allowed collection of standardized survey data for sites and their surrounding areas (SOM Fig. 1). Data on topographic features and archaeological site attributes aimed to assist reconstruction of settlement systems for different time periods. The intensification of farming and settlement activities strongly affected most archaeological sites and many of them will be destroyed in the near future. For this reason most of the recorded information will soon be lost. 2.3.1. Topographical distribution Orographically induced high precipitation at Mount Damota is predominantly stored in the clastic volcanic sediments on its
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
6
R. Vogelsang, K.P. Wendt / Quaternary International xxx (2017) 1e10
Fig. 5. Mount Damota survey form.
slopes. Together with the actual surface runoff this provides water supply through all seasons. This territorial advantage is mirrored also in the distance of the archaeological sites to the nearest water source. No site is > 1000 m from a permanent or at least seasonal water source, with an average distance of ~275 m. This is surely a major factor why the mountain has been a favorable area for settlement in past and present times. Other environmental factors, such as the natural vegetation in the surroundings of the sites are nearly completely reshaped by intensive farming activities. Today most parts of the mountain are a cultural landscape formed by the cultivation of land and most sites are located in currently or previously ploughed or fallow fields with isolated trees. For this reason, virtually all surfaces are more or less disturbed and affected by erosion. Given the steep topography of Mt. Damota, it is not surprising that nearly all sites were found on hill slopes. The number of archaeological sites in relation to elevation shows a quite consistent distribution between 2000 and 2400 m a.s.l. for both the “MSA”and the “LSA” periods. There is a slight concentration between 2000 and 2150 m a.s.l. (n ¼ 25 out of a total of 63) and most of the large surface sites are also located within this altitudinal belt. In some areas the dense scatter of archaeological finds makes it difficult to separate individual concentrations. The virtual absence
of sites above 2400 m a.s.l. during the first years of prospections seemed to point to an elevation dependent settlement restriction. However, surveys in 2014 resulted in the recording of sites on the plateau on top of the mountain. The explanation for the scarcity of sites at altitudes >2400 m seems to be the steepness of the slopes with a mean value of 19.49 in areas with absence of finds, compared to mean values of 13.79 for MSA sites and 14.39 for LSA sites (calculated on the basis of DLR SRTM-data with 30 m resolution). Most probably these areas were simply avoided for any human activities due to their steepness. However, the conditions for the preservation of sites are also poor in these areas, because the higher gradient intensified erosional processes that might have destroyed all former evidence. 2.3.2. Chronological distribution The chronological classification of the sites is solely based on characteristic stone tool types and potsherds (type fossils). However, the stratified assemblages from Mochena Borago showed that the general classification as “Middle Stone Age” or “Late Stone Age” stone tools is a simplification (Brandt et al., 2012; Pleurdeau et al., 2014). Keeping this in mind, the significance of single “type forms” is limited and only the overall impression of an assemblage is of informative value. In our study not the existence of backed
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
7
d2
d1
R. Vogelsang, K.P. Wendt / Quaternary International xxx (2017) 1e10
d3
A
B
C
Fig. 6. Construction scheme for of Delauney triangulations and Thiessen-polygons. Legend: black points ¼ archaeological site; dotted line ¼ Delauney triangles; d1 - d3 ¼ flanks of a triangle; open circles ¼ LECs; white squares ¼ node of a Thiessen-polygon/midpoint of circle; dark continuous lines ¼ borderlines of valid Thiessen-polygons; grey lines ¼ arbitrary border lines of Thiessen-polygons.
artefacts in general, but of backed microliths was regarded as specific of LSA assemblages. Regarding core reduction strategies, Levallois cores and flakes characterize MSA assemblages, whereas bladelet cores and their end-products occur during both periods. Unifacial and bifacial points are regarded as characteristic MSA tool types. According to their characteristic finds, 33 sites were classified as “Middle Stone Age” and 39 sites were classified as “Later Stone Age”. Disregarded for the purpose of the following reconstruction of settlement patterns were 12 sites classified as “historical” and 11 sites that could not be classified to a specific period. The total number of 63 sites arises from the multiple occupations of some localities during different time periods (e.g. 20 sites were classified as MSA and LSA). Up to now, not a single Early Stone Age type form, such as hand-axes or cleavers, was found. According to this missing evidence the occupation of the mountain did not start before 200 ka. A special feature of some of the surface MSA sites is the occurrence of large facially retouched points (Fig. 5, 7e10) beside the smaller pieces that characterize the excavated assemblage in the “Block Excavation Area” from Mochena Borago (Fig. 5, 1e6). They might represent older phases of the MSA with less standardized size ranges (Douze and Delagnes, 2016; Tables S1eS3) that are missing in this trench of the rock shelter. The observed gradual technological and typological changes that characterize the different Late Pleistocene cultural groups at Mochena Borago can only be differentiated in a stratigraphic context but cannot be transferred to surface assemblages. Stone tool manufacturing continued in the area up until recent times as evidenced by hide-working scrapers found at settlements abandoned in the last decade. 3. Reconstruction of settlement pattern Methods used for a spatial analysis of the sites are Delaunay triangulation and Thiessen-polygons (or Voronoi diagrams). Both are methodically close related and are based on the position of the sites in a defined area. They proved to be very successful not only in archaeology, but also in diverse subjects such as geography,
astronomy, crystallography, computational geometry, algorithm design, scientific computing, and optimization (Liebling and Pournin, 2012, p. 419). These methods are effective tools to describe archaeological settlement patterns. The quality of all referring analyses is dependent on the quality of the underlying archaeological survey-data. 3.1. Creating settlement areas The method of the Delaunay triangulation is based on a set of three adjacent points (sites) which are linked via the form of a triangle, excluding any nearby sites (Fig. 6 A). Sites connected in this way are also located on the border of the so-called “Largest Empty Circle” (LEC) (Fig. 6 B). The Euclidean distance (the length of the flanks of the triangle) is often used in archaeology to find regular (standardized) distances between prehistoric sites (Zimmermann, 1992). The distances not only describe a dense or loose settlement pattern, but can also provide indications for underlying networks of communication and infrastructure (Claßen and Zimmermann, 2004). The Thiessen-polygons were introduced to archaeological science in the 1970s by the British “New Archaeology” or “Processual Archaeology” (e.g. Clarke, 1977; Kienlin, 1998; Nakoinz, 2009, p. 361). The method is used in Archaeology to create territories (political or economic), especially if the landscape proves to be an inappropriate instrument for spatializing the recognized area. The construction of these territories is based on the principle of the LEC. The center of a LEC is the position of a territorial node (Fig. 6 B). The territory of one site can be constructed by linking the nodal points around the nearest site through a borderline, (Fig. 6 C). This method is inaccurate on the outskirts of a settlement, because the neighboring sites that are necessary to create a valid territory, are missing. Besides relating a site with surrounding space, the sizes of the polygons are also used to measure site densities. 3.2. “MSA” and “LSA” settlement clusters on Mount Damota A methodological problem is the time-depth of our chronological groups that are only roughly dated by surface artefacts into the
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
8
R. Vogelsang, K.P. Wendt / Quaternary International xxx (2017) 1e10
Fig. 7. MSA and LSA settlement pattern.
MSA and LSA period. For the purpose of our analysis we postulate that all sites classified as LSA and all sites classified as MSA are contemporaneous. Keeping this in mind, the triangulation and Thiessen-polygon based interpretations give only some first ideas,
which have to be proven by further research. For the 34 MSA sites 162 triangulation-distances can be provided with an average of half a kilometer (km) and most of the neighboring sites lying within a distance of 1.1 km (Fig. 7, Table 1).
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
R. Vogelsang, K.P. Wendt / Quaternary International xxx (2017) 1e10 Table 1 Statistical data of the length of the measured triangulation distances and of the sizes of the site territories determined by Thiessen-polygons.
n min. mean median max. stand.dev.
Triangulation (in m)
Thiessen-polygons (in ha)
MSA
LSA
MSA
LSA
162 38.22 563.05 502,26 2987.52 417.67
201 45.00 603.51 564.00 1804.00 358.14
18 2.60 20.21 14.51 46.09 13.56
23 4.55 20.42 18.00 46.61 10.88
The significant decrease in the site distribution beyond the 300 m mark might indicate that all cluster of archaeological finds within a shorter distance are part of one single site or that they represent a periodic resettlement at the same location. The size of the MSA site territories range from 3 to 46 ha (ha) within the 18 valid Thiessenpolygons (Sum ¼ 363 ha, mean ¼ 20.2 ha). Four separated settlement cluster can be reconstructed for the MSA. Except an isolated northern site concentration (C1), which is more or less west-east orientated, the other three settlement cluster (C2eC4) are stretching mainly from south-west to northeast (Fig. 7). This linear, vertical orientation includes different altitudinal belts. The pattern points to small, relatively isolated, hunter-gatherer groups moving up and down on one ridge of the mountain. The large size of some of these MSA sites might be interpreted as the result of numerous repeated stays at the same spot during annual migration cycles. The subsequent LSA provides some more sites (n ¼ 39) and triangulation distances (n ¼ 201 (Fig. 7, Table 1). The average distance is slightly higher (600 m) and the standard deviation less than the one for the MSA-period. The related histogram is more homogenous and shows no breaks, more or less following the normal distribution curvature. Visible is a trend to prefer shorter distances up to 500 m. The 23 Thiessen-polygons of the LSA enclose a space of 471 ha, in which their sizes range from 5 to 47 ha (mean ¼ 20.5 ha). With 18 ha the median of the LSA polygons is bigger than the one for the MSA (14.5 ha). New sites filled up the empty spaces that existed between the settlement areas during the MSA period and created a more regular structure. This indicates an intensification of settlement activities from the MSA to the LSA, especially if the different duration of both periods is taken into account. With the exception of the isolated cluster C1, two west-east orientated sub-groups are visible that are connected with each other through north-south contacts and thus form one large settlement cluster (C2). This network ensured (direct or indirect) access to different, elevation bound eco-zones. 3.3. Specialized “MSA” mountain dwellers? Of special interest for our research is the MSA settlement pattern that might indicate a land-use model that is known from historical hunter-gatherer groups. Such an example are the Okiek, who lived in the forested highlands of the Mau escarpment in west central Kenya until deforestation and the change of the land tenure system by government intervention over the last decades changed the demographics of the area (Huntingford, 1929; Blackburn, 1982; Kratz, 1999; Muchemi and Ehrensperger, 2011). Their residence groups used to be small extended families practicing a way of subsistence of non-specialized, opportunistic hunting. A specific feature of the Okiek subsistence is their specialization as honeygatherers. The settlement area of an Okiek group consisted of one
9
mountain ridge, on which they move to different forest zones according to three main honey seasons. Even though this is a specific subsistence strategy, the basic concept of taking advantage of different altitudinal eco-zones to exploit a wide spectrum of resources might explain the MSA settlement patterns at Mount Damota as periodically vertical movements of small huntergatherer groups. The compression of numerous vertical eco-zones within a small distance renders long distance annual migrations unnecessary. In this respect, tropical highland zones in Africa, such as Mount Damota are an attractive environment. However, it seems that the system fails during long-term climatic deterioration when even such favorable regions were abandoned (Foerster et al., 2015). Acknowledgements We thank the Ethiopian Authority for Research and Conservation of Cultural Heritage, the Southern Nations, Nationalities and Peoples Regional Government Bureau of Culture and Tourism, the Wolaita Zone Bureau of Culture and Tourism and Wolaita Sodo University for facilitating fieldwork and museum studies. We also thank the many Ethiopian, German and American students and colleagues who participated in the fieldwork or provided assistance in laboratory work. Funding was provided by the Deutsche Forschungsgemeinschaft (CRC 806), US National Science Foundation (BCS #0553371) and the University of Florida International Center. Finally we thank the people of Wolaita, and in particular the local residents of Mt. Damota, for their participation and support in field research. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.quaint.2017.06.061. References Abebe, B., Acocella, V., Korme, T., Ayalew, D., 2007. Quaternary faulting and volcanism in the main Ethiopian rift. J. Afr. Earth Sci. 48, 115e124. Abbate, E., Bruni, P., Sagri, M., 2015. Geology of Ethiopia: a review and geomorphological perspectives. In: Billi, P. (Ed.), Landscapes and Landforms of Ethiopia. Springer ScienceþBusiness Media, Dortrecht. Aldenderfer, M., 2014. Altitude environments in archaeology. In: Smith, C. (Ed.), Encyclopedia of Global Archaeology. Springer, New York, pp. 163e168. Awulachew, S.B., 2006. Investigation of physical and bathymetric characteristics of Lakes Abaya and Chamo, Ethiopia, and their management implications. Lakes Reservoirs Res. Manag. 11 (3), 133e140. Barham, L., 2002. Backed tools in Middle Pleistocene central Africa and their evolutionary significance. J. Hum. Evol. 43, 585e603. Basell, L.S., 2008. Middle stone age (MSA) site distributions in eastern Africa and their relationship to quaternary environmental change, refugia and the evolution of homo sapiens. Quat. Sci. Rev. 27, 2484e2498. Benvenuti, M., Carnicelli, S., 2015. The geomorphology of the lake region (Main Ethiopian Rift): the record of paleohydrological and paleoclimatic events in an active volcano-tectonic setting. In: Billi, P. (Ed.), Landscapes and Landforms of Ethiopia. Springer ScienceþBusiness Media, Dortrecht. Blackburn, R.H., 1982. In the land of milk and honey: okiek adaptations to their forests and neighbours. In: Leacock, E., Lee, R.B. (Eds.), Politics and History in Band Societies. Cambridge University Press, Cambridge, pp. 283e305. Brandt, S.A., Fisher, E.C., Hildebrand, E.A., Vogelsang, R., Ambrose, S.H., Lesur, J., Wang, H., 2012. Early MIS 3 occupation of Mochena Borago Rockshelter, southwest Ethiopian highlands: implications for late pleistocene archaeology, paleoenvironments and modern human dispersals. Quat. Int. 274, 38e54. Brandt, S.A., Hildebrand, E., Vogelsang, R., Wolfhagen, J., Wang, H., 2017. A new MIS 3 radiocarbon chronology for Mochena Borago Rockshelter, SW Ethiopia: implications for the interpretation of late pleistocene chronostratigraphy and human behavior. J. Archaeol. Sci. Rep. 11, 352e369. Chernet, T., 2011. Geology and hydrothermal resources in the northern lake Abaya area (Ethiopia). J. Afr. Earth Sci. 61 (2), 129e141. Clarke, D.L., 1977. Spatial information in archaeology. In: Clark, D.L. (Ed.), Spatial Archaeology. Academic Press, London, pp. 1e32. Claßen, E., Zimmermann, A., 2004. Tessellations and triangulations e understanding early neolithic social networks. In: Magistrat der Stadt Wien (Ed.), Enter the
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
10
R. Vogelsang, K.P. Wendt / Quaternary International xxx (2017) 1e10
Past. The E-way into the Four Dimensions of Cultural Heritage. CAA 2003, Computer Applications and Quantitative Methods in Archaeology, Proceedings of the 31st Conference, Vienna, Austria, April 2003, 467e471. Corti, G., Sani, F., Philippon, M., Sokoutis, D., Willingshofer, E., Molin, P., 2013. Quaternary volcano-tectonic activity in the Soddo region, western margin of the southern main Ethiopian rift. Tectonics 32 (4), 861e879. Douze, K., Delagnes, A., 2016. The pattern of emergence of a Middle Stone Age tradition at Gademotta and Kulkuletti (Ethiopia) through convergent tool and point technologies. J. Hum. Evol. 91, 93e121. Fisher, E., 2010. Late Pleistocene Technological Change and Hunter Gatherer Behavior at Moche Borago Rockshelter, Sodo-wolayta, Ethiopia: Flaked Stone Artifacts from the Early OIS 3 (60e43 Ka) Deposits. Unpublished PhD thesis. University of Florida, Gainesville. Foerster, V., Junginger, A., Langkamp, O., Gebru, T., Asrat, A., Umer, M., Lamb, H.F., Wennrich, V., Rethemeyer, J., Nowaczyk, N., Trauth, M., Schaebitz, F., 2012. Climatic change recorded in the sediments of the Chew Bahir basin, southern Ethiopia, during the last 45,000 years. Quat. Int. 274, 25e37. Foerster, V., Vogelsang, R., Junginger, A., Asrat, A., Lamb, H.F., Schaebitz, F., Trauth, M.H., 2015. Environmental change and human occupation of southern Ethiopia and northern Kenya during the last 20,000 years. Quat. Sci. Rev. 129, 333e340. Gutherz, X., 2000. Sondages dans l’abri sous-roche de Moche Borago Gongolo dans le Wolayta (Ethiopie). Ann. d'Ethiopie XVI 35e38. Gutherz, X., Jallot, L., Lesur, J., Pouzolles, G., Sordoillet, D., 2002. Les fouilles de l’abrisous-roche de Moche Borago (Soddo, Wolyata), Premier bilan. Ann. d'Ethiopie XVIII 181e190. Hildebrand, E., Brandt, S., Lesur, J., 2010. The Holocene archaeology of southwest Ethiopia: new insights from the kafa archaeological Project. Afr. Archaeol. Rev. 27, 255e289. Huntingford, G.W.B., 1929. Some account of the kamelilo-kapchepkendi dorobo (Okiek) of Kenya colony. J. R. Anthropol. Inst. G. B. Irel. 59, 333e378. Hurni, H., 1998. Agroecological Belts of Ethiopia: Explanatory Notes on Three Maps at a Scale of 1:1,000,000. Research Report for the Soil Conservation Research Programme, Ethiopia. Centre for Development and Environment. University of Bern, Switzerland, in association with The Ministry of Agriculture, Ethiopia. Kienlin, T. L., 1998. Die britische Processual Archaeology und die Rolle David L. Clarkes und Colin Renfrews: Herausbildung, Struktur, Einfluß. In: Eggert, M.K.H., Veit, U. (Eds.), Theorie in der Arch€ aologie. Zur englischsprachigen Diskussion. Tübinger Arch€ aologische Taschenbücher 1. Waxmann, Münster,
67e113. Kratz, C.A., 1999. The okiek of Kenya. In: Lee, R.B., Daly, R. (Eds.), The Cambridge Encyclopedia of Hunters and Gatherers. Cambridge University Press, Cambridge, pp. 220e224. Lesur, J., Vigne, J.-D., Gutherz, X., 2007. Exploitation of wild animals in south-west Ethiopia during the Holocene (4000 BC-500 AD): the finds from Moche Borago shelter (Wolayta). Environ. Archaeol. 12 (2), 139e159. Liebling, ThM., Pournin, L., 2012. Voronoi diagrams and Delaunay triangulations: ubiquitous siamese twins. Doc. Math. Extra Vol. ISMP 419e431. McBrearty, S., 2012. Sharpening the mind. Nature 491, 531e532. €deker, O., Fisher, E.C., Kehl, M., Brandt, S.A., Goldberg, P., Vogelsang, R., Meyer, S., Bo Zinaye, B., Bubenzer, O., 2017. Deciphering rockshelter evolution and sediment deposition in volcanic, mountainous, tropical humid landscapes: the case of Mochena Borago (Ethiopian Highlands). Geoarchaeology (Submitted). Muchemi, J.G., Ehrensperger, A., 2011. Ogiek Peoples Ancestral Territories Atlas: Safeguarding Territories, Cultures and Natural Resources of Ogiek Indigenous People in the Eastern Mau Forest, Kenya. ERMIS Africa & CDE, Nairobi/Bern. €rtliche Theorie. ArchNakoinz, O., 2009. Zentralortforschung und zentralo €ologisches Korresp. 39, 361e380. a Pleurdeau, D., Hovers, E., Assefa, E., Asrat, A., Pearson, O., Bahain, J.-J., Lam, Y.M., 2014. Cultural change or continuity in the late MSA/Early LSA of southeastern Ethiopia? The site of Goda Buticha, Dire Dawa area. Quat. Int. 343, 117e135. Shafer, H.J., 2016. Research design and sampling techniques. In: Hester, T.R., Shafer, H.J., Feder, K.L. (Eds.) Field Methods in Archaeology. Routledge, Oxford. , J.S., Hopmans, E.C., Tiermey, J.E., Russel, J.M., Huang, Y., Sinnighe Damste Cohen, A.S., 2008. Northern hemisphere controls on tropical southeast African climate during the past 60,000 years. Science 322, 252e255. Villa, P., Soriano, S., 2010. Hunting weapons of neanderthals and early modern humans in South Africa: similarities and differences. J. Anthropol. Res. 66 (1), 5e38. Viste, E., Sorteberg, A., 2013. Moisture transport into the Ethiopian highlands. Int. J. Climatol. 33, 249e263. Woldegabriel, G., Aronson, J.L., Walter, R.C., 1990. Geology, geochronology, and rift basin development in the central sector of the main ethiopia rift. Geol. Soc. Am. Bull. 102 (4), 439e458. Wurz, S., 1999. The howiesons poort backed artefacts from Klasies river: an argument for symbolic behaviour. South Afr. Archaeol. Bull. 54 (No. 169), 38e50. Zimmermann, A., 1992. Tesselierung und Triangulationals Techniken zur Bes€ologischer Funddichten. Archa €ologische Inf. 15, 107e112. timmung archa
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
Contents lists available at ScienceDirect
Quaternary International journal homepage: www.elsevier.com/locate/quaint
Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia Ralf Vogelsang*, Karl Peter Wendt Institute of Prehistoric Archaeology, University of Cologne, Bernhardt-Feilchenfeldstr. 11, 50969, Germany
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 January 2017 Received in revised form 19 June 2017 Accepted 28 June 2017 Available online xxx
Although high-altitude mountain habitats are often regarded as unfavorable for human occupation (e.g. Aldenderfer 2014); on the other hand tropical highlands in Africa are suggested as potential refugia during times of environmental stress (e.g. Basell 2008; Brandt et al. 2012). Archaeological investigations on Mount Damota (2908 m a.s.l.), located on the boundary between the Southwest Ethiopian Highlands to the west and the southern Main Ethiopian Rift valley to the east, yielded a large number of archaeological sites from the Middle Stone Age period until historical times. In this paper we try to reconstruct settlement models for the late Pleistocene and Holocene occupation in this area and speculate about potential land use patterns. Such complex topics demand a landscape archaeological approach that includes open-air sites and rock-shelters. The results from our excavations at Mochena Borago Rock-shelter and evidence from open-air-sites that were recorded during intensive surveys on the slopes and plateau of the mountain, allow a first reconstruction of the settlement history of the area. © 2017 Elsevier Ltd and INQUA. All rights reserved.
Keywords: Ethiopia Mount damota Middle stone age (MSA) Later stone age (LSA) Settlement pattern
1. Study background 1.1. Setting Although high-altitude mountain habitats are often regarded as unfavorable for human occupation (e.g. Aldenderfer 2014); on the other hand tropical highlands in Africa are suggested as potential refugia during times of environmental stress (e.g. Basell 2008; Brandt et al. 2012). With a maximal altitude of 2908 m a.s.l, Mt. Damota is a perfect case study for the verification of these contradictory conceptions. The study area is located 320 km south of Addis Ababa, where the southwest Ethiopian highlands intersect the Main Ethiopian Rift Valley near the town of Sodo (Fig. 1). Mt. Damota's steep upper slopes, deep gorges, gentle lower flanks and other physiographic features contribute to variation in vegetation within short distances. As a result, numerous vertical ecozones are compressed within a small geographic area on and around Mt. Damota. Like other regions of the Southwest Ethiopian Highlands (Hildebrand et al., 2010), the area around Mochena Borago is characterized by
* Corresponding author. E-mail addresses: [email protected] (R. Vogelsang), [email protected] (K.P. Wendt).
high biodiversity and an abundant supply of natural and domesticated faunal and floral resources for exploitation by contemporary and past human populations (Lesur et al., 2007; Brandt et al., 2012). This tight variation in altitude e and consequently in rainfall and vegetation e make the study area an excellent setting in which to examine changing human landscape preferences related to elevation. 1.2. Landforms Mount Damota is a dormant trachytic volcano comprised mainly of Pliocene porphyritic, anorthoclase-phenocryst trachytes and was active from at least 2.94 Ma until the Late Quaternary (Woldegabriel et al., 1990; Chernet, 2011; Corti et al., 2013). Rising steeply from its surrounding flanks (1300e1600 m a.s.l.), Mt Damota's a summit (2908 m a.s.l.; Fig. 2A) offers striking views of the central Main Ethiopian Rift Valley lakes to the distant north, the Bilate River and the southern Main Ethiopian Rift Valley to the east, Lake Abaya to the south, the Gibe and Omo River valleys to the southwest and the Wolayta/Hadiya Highlands to the West. The topography in the study area is dominated by steep slopes and rocky gorges in the mountain ranges (Abbate et al., 2015) and by lower relief topography in the transitional zone descending to the Rift Valley floor (Fig. 2C). Due to orographic effects, mean annual rainfall is significantly
http://dx.doi.org/10.1016/j.quaint.2017.06.061 1040-6182/© 2017 Elsevier Ltd and INQUA. All rights reserved.
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
2
R. Vogelsang, K.P. Wendt / Quaternary International xxx (2017) 1e10
Fig. 1. Location of Mount Damota, Mochena Borago Rockshelter (MB), the two known Humbo area obsidian sources and the survey area (detailed map shown in Fig. 4).
higher on Mt. Damota compared to the surrounding plains and the southern Main Ethiopian Rift Valley. Precipitation amounts to 2000 mm p.a. with half of the annual precipitation falling from June to September (Fisher, 2010; Viste and Sorteberg, 2013).
1.3. Ecological zones Population pressure, intensive farming and heavy erosion have impacted Mount Damota's natural landscape such that natural vegetation is only found in and near ravines, rocky outcrops, steep slopes and other areas too difficult to settle, plow or hoe. It is therefore difficult to make direct observations of ecological variation. Three of the major agro-ecological zones recognized by Ethiopian farmers on the basis of altitudinal changes in precipitation, temperature, soils and crop suitability (Hurni, 1998, pp. 18e19) are found on and around Mt. Damota: dega, woyna dega and kola (Fig. 3AeC). Natural floral distributions seem to parallel these
traditional agro-ecological zones. At ~2200 m asl, Mochena Borago Rockshelter is within the upper elevations of woyna dega, with bamboo growing nearby (Brandt et al., 2012). Thirty km SE of Mt. Damota, Ethiopia's largest rift lake, Lake Abaya, lies at 1169 m a.s.l. (Fig. 3 D). Ringed by freshwater marshes, Lake Abaya covers more than 1100 km2 and forms another significant regional biotope (Awulachew, 2006).
2. Archaeological research This study incorporates data from both open-air surface sites and rockshelters, which provide complementary records of human behavior and settlement patterns. Open-air surface sites are often numerous and can represent many distinct specific activities (e.g. settlement, hunting/butchering, raw material procurement). Rockshelters, while much rarer, carry several advantages. Their sealed contexts, while not assuring undisturbed contexts for finds, have much lower probability of mixed assemblages than surface
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
R. Vogelsang, K.P. Wendt / Quaternary International xxx (2017) 1e10
3
Fig. 2. View of Mount Damota (A), Mochena Borago Rockshelter (B), view from Mount Damota on the transitional zone descending to the Rift Valley floor (C).
sites. Also, because rockshelters offer protection from natural forces, they have often repeated episodes of use and sedimentation, leading to preservation of stratified sequences. Consistent low temperatures and protection from rainfall aid the preservation of bones, macrobotanical remains and charcoal, which are essential for both radiocarbon dating and paleoenvironmental reconstruction. Rockshelter sequences therefore provide crucial benchmarks for the chronological classification and environmental contextualization of assemblages from open-air sites. 2.1. History of research Three decades of archaeological survey and excavation have laid the groundwork for assessing human landscape use in the study area. In 1995, GEPCA (Groupe pour l’Etude de la Protohistoire de la Corne de l’Afrique) began a survey of the Wolayta region of SW Ethiopia directed by R. Joussaume. Between 1998 and 2001, GEPCA's extensive excavation of Mochena Borago Rockshelter (Fig. 2 B), directed by X. Gutherz, focused on mid to late Holocene
occupations but also revealed ~0.8 m of Later Pleistocene deposits in a 1.5 m2 test unit (Gutherz, 2000; Gutherz et al., 2002). In 2006, the Southwest Ethiopian Archaeological Project (SWEAP), directed by S. Brandt and E. Hildebrand, initiated Pleistocene-focused research at Mochena Borago. Since 2009, this work has continued under the co-direction of S. Brandt and R. Vogelsang within the framework of the University of Cologne's Collaborative Research Center 806 (CRC 806, http://www.sfb806.uni-koeln.de). CRC 806/ SWEAP carried out five additional fieldwork seasons between 2010 and 2014 (Brandt et al., 2012), including excavations of new areas within Mochena Borago. 2.2. Mochena Borago Rockshelter Mochena Borago rockshelter is situated on the southwestern flanks of Mount Damota at ~2200 m asl. (Fig. 2 B). Excavations of five different trenches yielded complex stratigraphies. Geoarchaeological analyses indicate a multiphase, polygenetic accumulation and erosion of sediments deposited through natural and
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
4
R. Vogelsang, K.P. Wendt / Quaternary International xxx (2017) 1e10
Fig. 3. Ecological zones on and around Mount Damota.
anthropogenic processes. The sedimentological evidence points to multiple occupations of the shelter mainly during rather humid conditions (Meyer et al., submitted). The Late Pleistocene layers span a period from >49 ka to ~36 ka. Fifty-nine radiocarbon dates on charcoal make Mochena Borago one of the best-dated archaeological sites from this period in eastern and northeastern Africa (Brandt et al., 2017). Pleistocene deposits in the main excavation area (BXA) consist of a sequence of six major lithostratigraphic groups incorporating four occupational episodes (Lower T-Group, Upper T-Group, SGroup, and R-Group). Preliminary descriptions of assemblages from T-Group, S-Group (Brandt et al., 2012) and R-Group (Brandt et al., 2017) are briefly summarized here, because the sequence is an
important reference point for a detailed chronological classification that cannot be achieved by surface assemblages found during survey. Remarkable is the continuity of technological and typological attributes of the lithic assemblages dating between ~49.4 and ~36.6 ka (Upper T-Group, S-Group and R-Group). Archaeological assemblages consist mainly of obsidian flaked stone artefacts (>96%). Other rocks, such as chalcedony, rhyolite and different cryptocrystaline silicates are represented only by single pieces. Such less frequently used rocks, mainly rhyolite, are with 7.5% more prominent only in the Lower T-Group assemblage that is older than 50 ka. The dominance of obsidian in the raw material spectrum is not surprising seeing that the Humbo area, a huge obsidian outcrop, is
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
R. Vogelsang, K.P. Wendt / Quaternary International xxx (2017) 1e10
5
Fig. 4. Survey-routes with the location of survey sites. MSA ¼ red, LSA ¼ yellow, historical ¼ orange.
situated only 20 km southeast from Mount Damota. First results of XRF analysis of obsidian artefacts and raw material pieces indicate that more than 90% of the obsidian artefacts are produced from this source. Although raw material availability seems not to have been a delimiting factor, artefacts in all assemblages are very small. Remarkable is the longstanding maintenance of many core reduction strategies, such as centripetal cores (Levallois- and radial cores) and a heterogeneous spectrum of other flake- and blade cores (single-, double- and multiple platform cores, bipolar cores, irregular cores) that coexist in all assemblages. However, in many cases the extreme reduction of the cores complicates the determination of their former shape. Characteristic retouched artefacts are facially retouched tools and backed pieces that are present in all inventories. However there is a clear change in proportion from a predominance of facial retouch in the older assemblages (Upper and Lower T-Group) to a predominance of backed pieces in the two younger strat-units. Unifacial and bifacial points are typical, but not all facially retouched tools can be classified as points. Especially in the TGroup assemblages there is a broad spectrum of facially retouched tool types that needs further investigation. The discovery of backed pieces in the deepest levels of Lower T-Group's undated, but older than 50 ka sediments suggests that this hafting strategy was practiced in SW Ethiopia when hunter-gatherers began their early MIS 3 range expansion (regarding the significance of backed tools for hafting and composite technology, and as an argument for symbolic behavior see e.g. Wurz, 1999; Barham, 2002; Villa and Soriano, 2010; McBrearty, 2012). In general, a longstanding continuity in the technological and typological attributes of all assemblages is remarkable considering the considerable climate oscillations of MIS 3 (e.g. Tiermey et al., 2008; Foerster et al., 2012), punctuated by major, possibly
catastrophic volcanic events (Abebe et al., 2007; Corti et al., 2013; Benvenuti and Carnicelli, 2015). 2.3. Open-air surface sites Extensive archaeological surveys in the surroundings of Mochena Borago were conducted concomitant to the excavations in the rockshelter. The survey strategy is best described as a “stratified random sampling” (Shafer, 2016). Pedestrian surveys attempted to include all different landforms of the area, such as slopes, steep valleys and plateaus, in different altitudes from the foot of the mountain (~1900 m a.s.l.) to the summit (2900 m a.s.l). For this reason, survey routes followed not only currently used pathways that are often along ridges. Some routes also crossed the flanks of the mountain horizontally, thus covering steep slopes and gorges. Surveys between 2010 and 2014 summed up to a total distance of 130 km. Sixty three archaeological sites were located (Fig. 4). Diagnostic finds e which range from “Middle Stone Age” (MSA) and “Later Stone Age” (LSA) technological elements to historic artefacts e were collected and suggest recurring use of this area over the past 50,000 years. Forms allowed collection of standardized survey data for sites and their surrounding areas (SOM Fig. 1). Data on topographic features and archaeological site attributes aimed to assist reconstruction of settlement systems for different time periods. The intensification of farming and settlement activities strongly affected most archaeological sites and many of them will be destroyed in the near future. For this reason most of the recorded information will soon be lost. 2.3.1. Topographical distribution Orographically induced high precipitation at Mount Damota is predominantly stored in the clastic volcanic sediments on its
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
6
R. Vogelsang, K.P. Wendt / Quaternary International xxx (2017) 1e10
Fig. 5. Mount Damota survey form.
slopes. Together with the actual surface runoff this provides water supply through all seasons. This territorial advantage is mirrored also in the distance of the archaeological sites to the nearest water source. No site is > 1000 m from a permanent or at least seasonal water source, with an average distance of ~275 m. This is surely a major factor why the mountain has been a favorable area for settlement in past and present times. Other environmental factors, such as the natural vegetation in the surroundings of the sites are nearly completely reshaped by intensive farming activities. Today most parts of the mountain are a cultural landscape formed by the cultivation of land and most sites are located in currently or previously ploughed or fallow fields with isolated trees. For this reason, virtually all surfaces are more or less disturbed and affected by erosion. Given the steep topography of Mt. Damota, it is not surprising that nearly all sites were found on hill slopes. The number of archaeological sites in relation to elevation shows a quite consistent distribution between 2000 and 2400 m a.s.l. for both the “MSA”and the “LSA” periods. There is a slight concentration between 2000 and 2150 m a.s.l. (n ¼ 25 out of a total of 63) and most of the large surface sites are also located within this altitudinal belt. In some areas the dense scatter of archaeological finds makes it difficult to separate individual concentrations. The virtual absence
of sites above 2400 m a.s.l. during the first years of prospections seemed to point to an elevation dependent settlement restriction. However, surveys in 2014 resulted in the recording of sites on the plateau on top of the mountain. The explanation for the scarcity of sites at altitudes >2400 m seems to be the steepness of the slopes with a mean value of 19.49 in areas with absence of finds, compared to mean values of 13.79 for MSA sites and 14.39 for LSA sites (calculated on the basis of DLR SRTM-data with 30 m resolution). Most probably these areas were simply avoided for any human activities due to their steepness. However, the conditions for the preservation of sites are also poor in these areas, because the higher gradient intensified erosional processes that might have destroyed all former evidence. 2.3.2. Chronological distribution The chronological classification of the sites is solely based on characteristic stone tool types and potsherds (type fossils). However, the stratified assemblages from Mochena Borago showed that the general classification as “Middle Stone Age” or “Late Stone Age” stone tools is a simplification (Brandt et al., 2012; Pleurdeau et al., 2014). Keeping this in mind, the significance of single “type forms” is limited and only the overall impression of an assemblage is of informative value. In our study not the existence of backed
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
7
d2
d1
R. Vogelsang, K.P. Wendt / Quaternary International xxx (2017) 1e10
d3
A
B
C
Fig. 6. Construction scheme for of Delauney triangulations and Thiessen-polygons. Legend: black points ¼ archaeological site; dotted line ¼ Delauney triangles; d1 - d3 ¼ flanks of a triangle; open circles ¼ LECs; white squares ¼ node of a Thiessen-polygon/midpoint of circle; dark continuous lines ¼ borderlines of valid Thiessen-polygons; grey lines ¼ arbitrary border lines of Thiessen-polygons.
artefacts in general, but of backed microliths was regarded as specific of LSA assemblages. Regarding core reduction strategies, Levallois cores and flakes characterize MSA assemblages, whereas bladelet cores and their end-products occur during both periods. Unifacial and bifacial points are regarded as characteristic MSA tool types. According to their characteristic finds, 33 sites were classified as “Middle Stone Age” and 39 sites were classified as “Later Stone Age”. Disregarded for the purpose of the following reconstruction of settlement patterns were 12 sites classified as “historical” and 11 sites that could not be classified to a specific period. The total number of 63 sites arises from the multiple occupations of some localities during different time periods (e.g. 20 sites were classified as MSA and LSA). Up to now, not a single Early Stone Age type form, such as hand-axes or cleavers, was found. According to this missing evidence the occupation of the mountain did not start before 200 ka. A special feature of some of the surface MSA sites is the occurrence of large facially retouched points (Fig. 5, 7e10) beside the smaller pieces that characterize the excavated assemblage in the “Block Excavation Area” from Mochena Borago (Fig. 5, 1e6). They might represent older phases of the MSA with less standardized size ranges (Douze and Delagnes, 2016; Tables S1eS3) that are missing in this trench of the rock shelter. The observed gradual technological and typological changes that characterize the different Late Pleistocene cultural groups at Mochena Borago can only be differentiated in a stratigraphic context but cannot be transferred to surface assemblages. Stone tool manufacturing continued in the area up until recent times as evidenced by hide-working scrapers found at settlements abandoned in the last decade. 3. Reconstruction of settlement pattern Methods used for a spatial analysis of the sites are Delaunay triangulation and Thiessen-polygons (or Voronoi diagrams). Both are methodically close related and are based on the position of the sites in a defined area. They proved to be very successful not only in archaeology, but also in diverse subjects such as geography,
astronomy, crystallography, computational geometry, algorithm design, scientific computing, and optimization (Liebling and Pournin, 2012, p. 419). These methods are effective tools to describe archaeological settlement patterns. The quality of all referring analyses is dependent on the quality of the underlying archaeological survey-data. 3.1. Creating settlement areas The method of the Delaunay triangulation is based on a set of three adjacent points (sites) which are linked via the form of a triangle, excluding any nearby sites (Fig. 6 A). Sites connected in this way are also located on the border of the so-called “Largest Empty Circle” (LEC) (Fig. 6 B). The Euclidean distance (the length of the flanks of the triangle) is often used in archaeology to find regular (standardized) distances between prehistoric sites (Zimmermann, 1992). The distances not only describe a dense or loose settlement pattern, but can also provide indications for underlying networks of communication and infrastructure (Claßen and Zimmermann, 2004). The Thiessen-polygons were introduced to archaeological science in the 1970s by the British “New Archaeology” or “Processual Archaeology” (e.g. Clarke, 1977; Kienlin, 1998; Nakoinz, 2009, p. 361). The method is used in Archaeology to create territories (political or economic), especially if the landscape proves to be an inappropriate instrument for spatializing the recognized area. The construction of these territories is based on the principle of the LEC. The center of a LEC is the position of a territorial node (Fig. 6 B). The territory of one site can be constructed by linking the nodal points around the nearest site through a borderline, (Fig. 6 C). This method is inaccurate on the outskirts of a settlement, because the neighboring sites that are necessary to create a valid territory, are missing. Besides relating a site with surrounding space, the sizes of the polygons are also used to measure site densities. 3.2. “MSA” and “LSA” settlement clusters on Mount Damota A methodological problem is the time-depth of our chronological groups that are only roughly dated by surface artefacts into the
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
8
R. Vogelsang, K.P. Wendt / Quaternary International xxx (2017) 1e10
Fig. 7. MSA and LSA settlement pattern.
MSA and LSA period. For the purpose of our analysis we postulate that all sites classified as LSA and all sites classified as MSA are contemporaneous. Keeping this in mind, the triangulation and Thiessen-polygon based interpretations give only some first ideas,
which have to be proven by further research. For the 34 MSA sites 162 triangulation-distances can be provided with an average of half a kilometer (km) and most of the neighboring sites lying within a distance of 1.1 km (Fig. 7, Table 1).
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
R. Vogelsang, K.P. Wendt / Quaternary International xxx (2017) 1e10 Table 1 Statistical data of the length of the measured triangulation distances and of the sizes of the site territories determined by Thiessen-polygons.
n min. mean median max. stand.dev.
Triangulation (in m)
Thiessen-polygons (in ha)
MSA
LSA
MSA
LSA
162 38.22 563.05 502,26 2987.52 417.67
201 45.00 603.51 564.00 1804.00 358.14
18 2.60 20.21 14.51 46.09 13.56
23 4.55 20.42 18.00 46.61 10.88
The significant decrease in the site distribution beyond the 300 m mark might indicate that all cluster of archaeological finds within a shorter distance are part of one single site or that they represent a periodic resettlement at the same location. The size of the MSA site territories range from 3 to 46 ha (ha) within the 18 valid Thiessenpolygons (Sum ¼ 363 ha, mean ¼ 20.2 ha). Four separated settlement cluster can be reconstructed for the MSA. Except an isolated northern site concentration (C1), which is more or less west-east orientated, the other three settlement cluster (C2eC4) are stretching mainly from south-west to northeast (Fig. 7). This linear, vertical orientation includes different altitudinal belts. The pattern points to small, relatively isolated, hunter-gatherer groups moving up and down on one ridge of the mountain. The large size of some of these MSA sites might be interpreted as the result of numerous repeated stays at the same spot during annual migration cycles. The subsequent LSA provides some more sites (n ¼ 39) and triangulation distances (n ¼ 201 (Fig. 7, Table 1). The average distance is slightly higher (600 m) and the standard deviation less than the one for the MSA-period. The related histogram is more homogenous and shows no breaks, more or less following the normal distribution curvature. Visible is a trend to prefer shorter distances up to 500 m. The 23 Thiessen-polygons of the LSA enclose a space of 471 ha, in which their sizes range from 5 to 47 ha (mean ¼ 20.5 ha). With 18 ha the median of the LSA polygons is bigger than the one for the MSA (14.5 ha). New sites filled up the empty spaces that existed between the settlement areas during the MSA period and created a more regular structure. This indicates an intensification of settlement activities from the MSA to the LSA, especially if the different duration of both periods is taken into account. With the exception of the isolated cluster C1, two west-east orientated sub-groups are visible that are connected with each other through north-south contacts and thus form one large settlement cluster (C2). This network ensured (direct or indirect) access to different, elevation bound eco-zones. 3.3. Specialized “MSA” mountain dwellers? Of special interest for our research is the MSA settlement pattern that might indicate a land-use model that is known from historical hunter-gatherer groups. Such an example are the Okiek, who lived in the forested highlands of the Mau escarpment in west central Kenya until deforestation and the change of the land tenure system by government intervention over the last decades changed the demographics of the area (Huntingford, 1929; Blackburn, 1982; Kratz, 1999; Muchemi and Ehrensperger, 2011). Their residence groups used to be small extended families practicing a way of subsistence of non-specialized, opportunistic hunting. A specific feature of the Okiek subsistence is their specialization as honeygatherers. The settlement area of an Okiek group consisted of one
9
mountain ridge, on which they move to different forest zones according to three main honey seasons. Even though this is a specific subsistence strategy, the basic concept of taking advantage of different altitudinal eco-zones to exploit a wide spectrum of resources might explain the MSA settlement patterns at Mount Damota as periodically vertical movements of small huntergatherer groups. The compression of numerous vertical eco-zones within a small distance renders long distance annual migrations unnecessary. In this respect, tropical highland zones in Africa, such as Mount Damota are an attractive environment. However, it seems that the system fails during long-term climatic deterioration when even such favorable regions were abandoned (Foerster et al., 2015). Acknowledgements We thank the Ethiopian Authority for Research and Conservation of Cultural Heritage, the Southern Nations, Nationalities and Peoples Regional Government Bureau of Culture and Tourism, the Wolaita Zone Bureau of Culture and Tourism and Wolaita Sodo University for facilitating fieldwork and museum studies. We also thank the many Ethiopian, German and American students and colleagues who participated in the fieldwork or provided assistance in laboratory work. Funding was provided by the Deutsche Forschungsgemeinschaft (CRC 806), US National Science Foundation (BCS #0553371) and the University of Florida International Center. Finally we thank the people of Wolaita, and in particular the local residents of Mt. Damota, for their participation and support in field research. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.quaint.2017.06.061. References Abebe, B., Acocella, V., Korme, T., Ayalew, D., 2007. Quaternary faulting and volcanism in the main Ethiopian rift. J. Afr. Earth Sci. 48, 115e124. Abbate, E., Bruni, P., Sagri, M., 2015. Geology of Ethiopia: a review and geomorphological perspectives. In: Billi, P. (Ed.), Landscapes and Landforms of Ethiopia. Springer ScienceþBusiness Media, Dortrecht. Aldenderfer, M., 2014. Altitude environments in archaeology. In: Smith, C. (Ed.), Encyclopedia of Global Archaeology. Springer, New York, pp. 163e168. Awulachew, S.B., 2006. Investigation of physical and bathymetric characteristics of Lakes Abaya and Chamo, Ethiopia, and their management implications. Lakes Reservoirs Res. Manag. 11 (3), 133e140. Barham, L., 2002. Backed tools in Middle Pleistocene central Africa and their evolutionary significance. J. Hum. Evol. 43, 585e603. Basell, L.S., 2008. Middle stone age (MSA) site distributions in eastern Africa and their relationship to quaternary environmental change, refugia and the evolution of homo sapiens. Quat. Sci. Rev. 27, 2484e2498. Benvenuti, M., Carnicelli, S., 2015. The geomorphology of the lake region (Main Ethiopian Rift): the record of paleohydrological and paleoclimatic events in an active volcano-tectonic setting. In: Billi, P. (Ed.), Landscapes and Landforms of Ethiopia. Springer ScienceþBusiness Media, Dortrecht. Blackburn, R.H., 1982. In the land of milk and honey: okiek adaptations to their forests and neighbours. In: Leacock, E., Lee, R.B. (Eds.), Politics and History in Band Societies. Cambridge University Press, Cambridge, pp. 283e305. Brandt, S.A., Fisher, E.C., Hildebrand, E.A., Vogelsang, R., Ambrose, S.H., Lesur, J., Wang, H., 2012. Early MIS 3 occupation of Mochena Borago Rockshelter, southwest Ethiopian highlands: implications for late pleistocene archaeology, paleoenvironments and modern human dispersals. Quat. Int. 274, 38e54. Brandt, S.A., Hildebrand, E., Vogelsang, R., Wolfhagen, J., Wang, H., 2017. A new MIS 3 radiocarbon chronology for Mochena Borago Rockshelter, SW Ethiopia: implications for the interpretation of late pleistocene chronostratigraphy and human behavior. J. Archaeol. Sci. Rep. 11, 352e369. Chernet, T., 2011. Geology and hydrothermal resources in the northern lake Abaya area (Ethiopia). J. Afr. Earth Sci. 61 (2), 129e141. Clarke, D.L., 1977. Spatial information in archaeology. In: Clark, D.L. (Ed.), Spatial Archaeology. Academic Press, London, pp. 1e32. Claßen, E., Zimmermann, A., 2004. Tessellations and triangulations e understanding early neolithic social networks. In: Magistrat der Stadt Wien (Ed.), Enter the
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061
10
R. Vogelsang, K.P. Wendt / Quaternary International xxx (2017) 1e10
Past. The E-way into the Four Dimensions of Cultural Heritage. CAA 2003, Computer Applications and Quantitative Methods in Archaeology, Proceedings of the 31st Conference, Vienna, Austria, April 2003, 467e471. Corti, G., Sani, F., Philippon, M., Sokoutis, D., Willingshofer, E., Molin, P., 2013. Quaternary volcano-tectonic activity in the Soddo region, western margin of the southern main Ethiopian rift. Tectonics 32 (4), 861e879. Douze, K., Delagnes, A., 2016. The pattern of emergence of a Middle Stone Age tradition at Gademotta and Kulkuletti (Ethiopia) through convergent tool and point technologies. J. Hum. Evol. 91, 93e121. Fisher, E., 2010. Late Pleistocene Technological Change and Hunter Gatherer Behavior at Moche Borago Rockshelter, Sodo-wolayta, Ethiopia: Flaked Stone Artifacts from the Early OIS 3 (60e43 Ka) Deposits. Unpublished PhD thesis. University of Florida, Gainesville. Foerster, V., Junginger, A., Langkamp, O., Gebru, T., Asrat, A., Umer, M., Lamb, H.F., Wennrich, V., Rethemeyer, J., Nowaczyk, N., Trauth, M., Schaebitz, F., 2012. Climatic change recorded in the sediments of the Chew Bahir basin, southern Ethiopia, during the last 45,000 years. Quat. Int. 274, 25e37. Foerster, V., Vogelsang, R., Junginger, A., Asrat, A., Lamb, H.F., Schaebitz, F., Trauth, M.H., 2015. Environmental change and human occupation of southern Ethiopia and northern Kenya during the last 20,000 years. Quat. Sci. Rev. 129, 333e340. Gutherz, X., 2000. Sondages dans l’abri sous-roche de Moche Borago Gongolo dans le Wolayta (Ethiopie). Ann. d'Ethiopie XVI 35e38. Gutherz, X., Jallot, L., Lesur, J., Pouzolles, G., Sordoillet, D., 2002. Les fouilles de l’abrisous-roche de Moche Borago (Soddo, Wolyata), Premier bilan. Ann. d'Ethiopie XVIII 181e190. Hildebrand, E., Brandt, S., Lesur, J., 2010. The Holocene archaeology of southwest Ethiopia: new insights from the kafa archaeological Project. Afr. Archaeol. Rev. 27, 255e289. Huntingford, G.W.B., 1929. Some account of the kamelilo-kapchepkendi dorobo (Okiek) of Kenya colony. J. R. Anthropol. Inst. G. B. Irel. 59, 333e378. Hurni, H., 1998. Agroecological Belts of Ethiopia: Explanatory Notes on Three Maps at a Scale of 1:1,000,000. Research Report for the Soil Conservation Research Programme, Ethiopia. Centre for Development and Environment. University of Bern, Switzerland, in association with The Ministry of Agriculture, Ethiopia. Kienlin, T. L., 1998. Die britische Processual Archaeology und die Rolle David L. Clarkes und Colin Renfrews: Herausbildung, Struktur, Einfluß. In: Eggert, M.K.H., Veit, U. (Eds.), Theorie in der Arch€ aologie. Zur englischsprachigen Diskussion. Tübinger Arch€ aologische Taschenbücher 1. Waxmann, Münster,
67e113. Kratz, C.A., 1999. The okiek of Kenya. In: Lee, R.B., Daly, R. (Eds.), The Cambridge Encyclopedia of Hunters and Gatherers. Cambridge University Press, Cambridge, pp. 220e224. Lesur, J., Vigne, J.-D., Gutherz, X., 2007. Exploitation of wild animals in south-west Ethiopia during the Holocene (4000 BC-500 AD): the finds from Moche Borago shelter (Wolayta). Environ. Archaeol. 12 (2), 139e159. Liebling, ThM., Pournin, L., 2012. Voronoi diagrams and Delaunay triangulations: ubiquitous siamese twins. Doc. Math. Extra Vol. ISMP 419e431. McBrearty, S., 2012. Sharpening the mind. Nature 491, 531e532. €deker, O., Fisher, E.C., Kehl, M., Brandt, S.A., Goldberg, P., Vogelsang, R., Meyer, S., Bo Zinaye, B., Bubenzer, O., 2017. Deciphering rockshelter evolution and sediment deposition in volcanic, mountainous, tropical humid landscapes: the case of Mochena Borago (Ethiopian Highlands). Geoarchaeology (Submitted). Muchemi, J.G., Ehrensperger, A., 2011. Ogiek Peoples Ancestral Territories Atlas: Safeguarding Territories, Cultures and Natural Resources of Ogiek Indigenous People in the Eastern Mau Forest, Kenya. ERMIS Africa & CDE, Nairobi/Bern. €rtliche Theorie. ArchNakoinz, O., 2009. Zentralortforschung und zentralo €ologisches Korresp. 39, 361e380. a Pleurdeau, D., Hovers, E., Assefa, E., Asrat, A., Pearson, O., Bahain, J.-J., Lam, Y.M., 2014. Cultural change or continuity in the late MSA/Early LSA of southeastern Ethiopia? The site of Goda Buticha, Dire Dawa area. Quat. Int. 343, 117e135. Shafer, H.J., 2016. Research design and sampling techniques. In: Hester, T.R., Shafer, H.J., Feder, K.L. (Eds.) Field Methods in Archaeology. Routledge, Oxford. , J.S., Hopmans, E.C., Tiermey, J.E., Russel, J.M., Huang, Y., Sinnighe Damste Cohen, A.S., 2008. Northern hemisphere controls on tropical southeast African climate during the past 60,000 years. Science 322, 252e255. Villa, P., Soriano, S., 2010. Hunting weapons of neanderthals and early modern humans in South Africa: similarities and differences. J. Anthropol. Res. 66 (1), 5e38. Viste, E., Sorteberg, A., 2013. Moisture transport into the Ethiopian highlands. Int. J. Climatol. 33, 249e263. Woldegabriel, G., Aronson, J.L., Walter, R.C., 1990. Geology, geochronology, and rift basin development in the central sector of the main ethiopia rift. Geol. Soc. Am. Bull. 102 (4), 439e458. Wurz, S., 1999. The howiesons poort backed artefacts from Klasies river: an argument for symbolic behaviour. South Afr. Archaeol. Bull. 54 (No. 169), 38e50. Zimmermann, A., 1992. Tesselierung und Triangulationals Techniken zur Bes€ologischer Funddichten. Archa €ologische Inf. 15, 107e112. timmung archa
Please cite this article in press as: Vogelsang, R., Wendt, K.P., Reconstructing prehistoric settlement models and land use patterns on Mt. Damota/SW Ethiopia, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2017.06.061