Version classiqueVersion mobile

Géoarchéologie des îles de la Méditerranée

Matthieu Ghilardi

Partie 1. Anthropisation et mutations paysagères à la transition Paléolithique/Néolithique / Anthropization and landscape changes during the Late Paleolithic/Neolithic transition

Site Formation Processes at Akrotiri Aetokremnos, Cyprus: Why is the site so controversial?

Simmons Alan et Mandel Rolfe D


There have been numerous claims for pre-Neolithic occupations of many of the Mediterranean islands. Generally, these are not well supported by archaeological, chronological, and stratigraphic data. This changed with the excavation of Akrotiri Aetokremnos, a small collapsed rockshelter that has provided solid evidence of the earliest well- documented human presence on Cyprus. A large suite of radiocarbon ages indicates that the site was occupied around 12000 cal. BP, contemporary with the Late Epi-Paleolithic of the Near Eastern and Anatolian mainlands. Since Cyprus was not connected to the mainland by a landbridge during the Quaternary, Aetokremnos provides direct evidence for pre-Neolithic seafaring abilities. More controversial than the site’s age is the association of extinct endemic pygmy hippopotami with cultural materials, as this relates to the continuing discussion of human-caused Pleistocene extinctions. The claim of a direct association has been challenged, however, despite well-published archaeological and geoarchaeological data that support it. This paper summarizes the arguments for the association of cultural remains with pygmy hippopotami, specifically focusing on the site’s stratigraphic sequence, chronology, and formation processes. These data are consistent with the hypothesis that the remains of the pygmy hippopotami are in direct association with cultural features and artifacts that comprise the site’s two primary archaeological strata.

Texte intégral

Introduction and Research Background

1Nowhere has the critical application of geoarchaeology to archaeology been more apparent than at the controversial site of Akrotiri Aetokremnos in Cyprus. For years it was believed that the Mediterranean islands were not occupied until relatively late, during the later pre-pottery Neolithic, around 9 000 years ago. There had been a few claims for Paleolithic sites, but these were based on scant evidence (Simmons, 2014). That scenario, however, changed with Aetokremnos, which firmly dates to around 12000 cal. BP, during the Late Epi-Paleolithic. Nevertheless, the site and its implications remain controversial even after over 20 years since its excavation (Simmons, 1988, 1999, 2004, 2013a). This contribution addresses the controversy surrounding Aetokremnos, focusing on its geoarchaeological context, which has been addressed in great detail elsewhere (e.g. Mandel and Simmons, 1997; Mandel, 1999). To fully under- stand Aetokremnos, a circum-Mediterranean context is necessary. Conventional wisdom was that the Mediterranean islands were generally too impoverished to have supported hunter and gatherer economies, and could only have been substantially occupied by food producers, or Neolithic peoples (Simmons, 2007). And even then, only Cyprus, and perhaps Crete, had Aceramic (pre-pottery) Neolithic occupations, which were relatively late (ca. 9000 cal. BP) by mainland standards. The Neolithic on most of the other islands occurred only during the later Ceramic Neolithic, at around 6500 cal. BP or later. This scenario has now been discounted, with recent research showing that Cyprus at least has a Neolithic record that is approximately as old as that on the adjacent Anatolian and Levantine mainlands, including both Pre-Pottery Neolithic A (PPNA), starting around 11500 cal. BP, and Pre-Pottery Neolithic B (PPNB), starting around 10500 cal. BP, phases (Simmons, 2007). In the context of this contribution, it is necessary to expand to the pre-Neolithic. As noted above, claims for such occupations are poorly supported. The better claims relate mainly to the Epi-Paleolithic, and, until the excavation of Aetokremnos, most were restricted to islands relatively close to the mainland. Furthermore, the Epi-Paleolithic on many of these was late by Near Eastern Epi-Paleolithic standards (Simmons, 2014). Claims for even earlier usages of the islands, extending as far back as the Lower Paleolithic in some cases, have generally been dispelled. There are, however, now a few increasingly compelling arguments for at least limited early (Middle Paleolithic and likely Neanderthal, and perhaps even earlier) human use of some of the islands, such as Crete (Strasser et al., 2010) and the Ionian Islands (Ferentinos et al., 2012). This is particularly significant for the former, since it, like Cyprus, is “oceanic,” separated from the main- land by considerable distances of open water and with no evidence for a landbridge during the Quaternary (Lort, 1977; Stanley, 1977). This, of course, implies considerable seafaring abilities. But, this is a difficult type of archaeology to document for several reasons. These include the low archaeological visibility typical of pre-Neolithic sites. Also, sites occupied by hunters and gatherers usually leave few material remains. In addition, because of their antiquity, these sites often are either destroyed by erosion or buried by sediment. Furthermore, documenting such sites is hampered by lack of stratigraphic context (many are surface remains), difficulties in numerical dating, and, until recently, a general unfamiliarity of Mediterranean island archaeologists with these often ephemeral pre-Neolithic remains (Simmons, 2014). Despite these issues, Aetokremnos stands out as one of the first sites to indicate indisputable pre-Neolithic occupation of a Mediterranean island. Unlike the majority of other instances, Aetokremnos has excellent chronology and stratigraphic integrity.

Environment and Geology

2Aetokremnos is located on the southern portion of the Akrotiri Peninsula, the dominant feature of the southern coast of Cyprus (Figure 1). The peninsula can be broadly defined as a shallow syncline of Miocene, Pliocene, and Quaternary rocks with an east-west axis. The peninsula represents a former island that is now joined to the mainland by a tombolo. Thre is a shallow salt lake (the Akrotiri Salt Lake) between the two spits that formed the tombolo. When the island was formed is not well-es-tablished, although it is certain that the eastern end of the Akrotiri Salt Lake was open to the sea as recently as the Roman period, as shown on many ancient maps and by a Hellenistic or Roman shipwreck discovered near the lake, some 150 m. from the present shoreline. Whether or not the island existed when Aetokremnos was occupied is unresolved (Heywood, 1982; Morel, 1960; Simmons, 1999). The Akrotiri Peninsula is hilly to the north and gently inclines towards the sea. The hilly portion consists primarily of chalks and marls, with intrusions of siltstones and alluvium. Many plateaus and ridges have a thick deposit of havara (a Cypriot term that broadly translates as “lime- stone”). These sometimes contain pockets of terra rossa. The greater part of the cliffs where Aetokremnos is located exposes Pliocene-age rocks of the Athalassa and Nicosia Formations. There are no karst features, such as sinkholes, on the peninsula. Late Pleistocene and Holocene aeolian sand sheets and dunes are banked up against and over some of the cliffs around Aetokremnos. There also are modern dunes on the peninsula. No major streams cross the peninsula, although the Kouris River is directly to the north. The peninsula shares the semi-arid Mediterranean climate common to Cyprus. Unfortunately, for the island as a whole, there are limited paleoenvironmental data available (Morel, 1960; Simmons, 1999).

Figure 1: Location map of site and Akrotiri Peninsula.

A Brief Summary of Investigations at Akrotiri Aetokremnos

3Aetokremnos itself is a collapsed rockshelter precariously perched on a cliff overlooking the Mediterranean (Figure 2). Parts of the shelter have clearly fallen into the sea due to erosion, and surface remains were not impressive, consisting primarily of the bones of extinct, endemic pygmy hippopotamus (Phanourios minutus Cuvier [Faure et al., 1983]). This by itself was not surprising: there are many pygmy hippo and dwarf elephant paleontological sites on Cyprus and other Mediterranean islands, such as Crete and Malta (Reese, 1995; Sondaar, 1986). These endemic fauna were assumed to have gone extinct long before any human presence on the islands. At Aetokremnos, however, there were hints of cultural activity in the form of a small number of chipped stone artifacts associated with the fauna visible on the surface. This finding was intriguing because, if the association of artifacts with this extinct fauna could be stratigraphically and chronologically verified by excavation, Aetokremnos would have direct implications on the controversial issue of Pleistocene extinctions – were they caused by human activity or environmental change (e.g. Martin and Klein, 1964)? Three seasons of excavation in 1987, 1988, and 1990 (and a survey season in 1991), led to the conclusion that humans were, in fact, implicated in an extinction episode (Simmons, 1999, 2013a). This claim set off the storm of controversy that still surrounds the site. Archaeologists are generally conservative and proposing that humans might have played a role in the eradication of endemic hippos, likely already under stress by the cold and dry Younger Dryas, was simply pushing the envelope too far for some (e.g. Ammerman and Noller, 2005; Buminovitz and Barkai, 1999; Zazzo et al., 2015). But, by using multiple lines of evidence, the data from Aetokremnos suggested just that. Excavation was conducted on most of the remaining part of the shelter (just over 50 square meters), resulting in a huge and dense well-preserved assemblage of nearly 300 000 animal remains (Figure 3A and 3B), with over 98.3% belonging to pygmy hippos, representing a Minimum Number of Individuals (MNI) of over 500 animals. Much of this was burned (ca. 30%), and the remains were jumbled with virtually no articulation. We also recovered a minimum of three dwarf elephants, over 70 great bustards, and over 70 000 marine shells. The artifact assemblage consisted of over 1 000 chipped stone artifacts, quite distinct from anything then known on the island, but fitting comfortably within mainland materials represented by a general Natufian phase of the Late Epi-Paleolithic (ca. 15000-12000 cal. BP) (Simmons 1999, 2007, 2013a). Distinctive thumbnail scrapers dominated the tool assemblage (Figure 4C), although other types of tools were present as well. Personal ornamentation, primarily in the form of beads, also was recovered (Figure 4 A and 4B), and 11 features consisting of hearths and burnt bone were recorded.

Figure 2: Site overview photo. The insert shows rooffall overlying and protecting cultural remains.

Figure 3: Photo of faunal remains, showing density once excavation had begun (A) and well-preserved hippo cranium (B).

Figure 4: Artifacts from Akrotiri Aetokremnos-shell beads (A), picrolite bead (B), and thumbnail scrapers (C).


4It was immediately realized that a detailed geoarchaeological study focusing on stratigraphy and site formation processes was crucial to the proper interpretation of Aetokremnos. Thus, from the beginning of the investigations, geoarchaeology was a critical component, with all of the detailed study being done by Mandel. Field investigations at Aetokremnos went through several stages. Work was initially hindered by the presence of large roof-fall blocks mantled by a veneer of colluvium. During the 1987 field season, excavation units were concentrated along the fringes of the shelter in order to avoid these obstacles. Several large roof-fall blocks were removed in 1988 to allow excavation within the interior of the shelter. However, one massive roof-fall block remained in place over a large area of the site. Hence the full thickness and lateral extent of the underlying deposits could not be determined until the final field season. Complete removal of the collapsed roof exposed a package of unconsolidated sediment that is about 1 m thick. The thickness of this package was fairly consistent from the outer edge to the back of the shelter. The sediment directly overlies the bedrock floor of the shelter. A total of fifty-five square meters were excavated at Aetokremnos, including units placed outside of the rockshelter All but about 2.5 m2 were excavated to bedrock. Excavation units were 1 m2 or less, and all sediment was dry-screened. Because the excavation of Aetokremnos progressed from the surface downward over a period of several years, it did not follow geologic protocol and number stratigraphic units sequentially from the bottom of the shelter upward. Instead, we followed archaeological convention in numbering units from the surface downward; that is, the lowest numerical unit is at the top of the stratigraphic sequence and also is the youngest deposit. Bone and artifact provenience was keyed into this numbering system during the excavation.

5Four major stratigraphic units were identified at Aetokremnos on the basis of general physical properties observed in the field (Mandel, 1999). Arabic numerals (1-4) were used to designate these units beginning with the uppermost one, Stratum 1. The stratigraphic units were subdivided into sedimentary units that are designated by the addition of an upper-case letter after the stratum number, for example, Stratum 2A. The sedimentary units are rock terms defined by lithologic characteristics, such as color, texture, and bedding. After describing the physical characteristics of the sedimentary units at Aetokremnos, sediment samples were collected for laboratory analyses. Rockshelter sediments are usually sampled from a single type section that characterizes all the strata in the site. At Aetokremnos, this sampling strategy was not feasible because all strata were not present in a single section. In order to interpret the stratigraphy across the site, it was necessary to sample sediments from several profiles. Sediment samples weighing approximately 500 g each were collected from the middle of each sedimentary unit represented in the profiles. Several localities outside the rockshelter were sampled to provide comparative data for possible sources of sediments within the shelter. Colluvial deposits mantling the roof at the back of the rockshelter were sampled from two profiles. In addition, samples were taken from deposits of aeolian sand on slopes adjacent to and below the rockshelter, and one sample was collected from the littoral zone of a sandy beach along the shore of the Akrotiri Peninsula. The beach sample was used to characterize the mineralogy and granulometry of sediment in the primary source area for aeolian deposits on the slopes surrounding Aetokremnos. Mechanical and chemical analyses of sediment samples were conducted at Kansas State University’s Soil Characterization Laboratory. Particle size distribution was determined using a modification of the pipette method developed by Kilmer and Alexander (1949) and the Soil Survey Staff (1982). The results of the particle-size analyses were used to determine mean grain size and standard deviation. Total phosphorus (P) was determined by digesting samples with a salicylic–sulfuric acid mixture in a Technicon Model BD-20 Block Digester (Technicon Industrial Method 334-74W/B). Calcium carbonate (CaCO3) equivalent was determined by the rapid titration method (Piper, 1950). The mineralogy, sphericity, and roundness of sand grains were determined through optical analysis with a binocular petrographic microscope.

Site Stratigraphy and Chronology

6The site’s stratigraphic sequence is illustrated in Figure 5. Stratum 1 is the uppermost stratigraphic unit and includes colluvium that mantles intact and collapsed portions of the rockshelter’s roof and colluvium, slopewash, rooffall, and attrition sediment above Stratum 2. Stratum 1 does not contain any in situ cultural materials. Stratum 1 is subdivided into five sedimentary units: Stratum 1A through 1E. Stratum 1A is 25-50 cm thick and consists of colluvium derived from slopes above the rockshelter. This unit mantled the large roof-fall blocks above the archaeological deposits, and it overlies Stratum 1B on the bedrock overhang at the back of the rockshelter. Angular limestone cobbles and boulders are scattered through a calcareous, fine-grained matrix, and sedimentary features are absent. A very thin surface soil with A-C horizonation is developed in the upper part of Stratum 1A. Stratum 1B is restricted to the bedrock overhang at the back of the rockshelter. This sedimentary unit rests directly on bedrock and is mantled by Stratum 1A. It is 25-40 cm thick and, like Stratum 1A, consists of poorly sorted colluvium derived from the slopes above the site. The boundary between 1A and 1B is marked by a buried soil developed in the upper part of Stratum 1B. Hence colluviation was interrupted by an episode of landscape stability and soil formation before Stratum 1A aggraded. The buried soil is distinct but weakly developed (A-AC profile). Stratum 1C consists of laminated, sandy and silty slopewash that was deposited near the north wall of the rockshelter. The surface of this unit slopes south- westward away from the back of the shelter. Points of entry for the slopewash are several shallow fractures in the bedrock along the top of the north wall of the shelter. As sediment-laden water flowed through these fractures and out of the shelter, it cut a small channel through Strata 2, 3, and 4 and into the under- lying bedrock. This channel was subsequently filled with laminated sediment that composes Stratum 1C. Individual laminae are 1-4 mm thick and gently dip to the southwest. Stratum 1D consists of laminated, dark brown (10YR 4/3, dry) slopewash above Stratum IE. It is restricted to the extreme northeast corner of the shelter, and its surface, like that of Stratum C, slopes southwestward away from the back of the shelter. The gradient was established by deposition of slope wash spilling out of a fracture in the bedrock located high up on the wall in that corner. The dark sediment was traced up through the cavity to its source: the buried soil developed in Stratum 1B. Stratum 1D prograded only a short distance to the west-southwest. Individual laminae are 1-2 mm thick and have sandy loam texture. Stratum 1E is a culturally sterile body of sediment that was largely generated by roof fall and grain-by-grain disintegration (attrition) of the bioclastic limestone. Angular fragments of roof fall 5-75 cm in diameter are scattered through a homogeneous, very pale brown (10YR 7/4, dry) sandy-loam matrix. Stratum 1E is about 30 cm thick and has not been modified by soil development.

Figure 5: Representative composite stratigraphic drawing (prepared by R. Watters); insert is a photo of the stratigraphyand location A shows generalized location within stratigraphic section.

  • 1 All laboratory data are presented in Mandel and Simmons (1997).

7Stratum 2 is the uppermost intact cultural zone and contains the majority of the artifacts, features, and bird remains, although hippo bones also are present. It is 20 to 30 cm in thick and largely consists of attrition and aeolian sediment, though large and small fragments of rock fall are scattered through it. This unit mantles a culturally sterile zone (Stratum 3) within most of the shelter, but directly overlies Stratum 4 in places at the front of the shelter. The lower boundary of Stratum 2 is usually abrupt. Stratum 2 is subdivided into Stratum 2A and 2B. Stratum 2A is at the top of Stratum 2 and is 15-45 cm thick. The upper two-thirds of Stratum 2A consists of calcareous, dark grayish brown (10YR 4/2, dry) sandy loam. The matrix is loose and interspersed with cobble- and boulder-size clasts representing roof fall. There is a large volume of fine, disseminated charcoal in the upper part of 2A, giving the matrix an ashy appearance. The lower half of 2A also is calcareous sandy loam, but it is not as enriched with charcoal; hence, it has a lighter color (grayish brown [10YR 5/2, dry] to light brownish gray [10YR 6/2, dry]). There is no evidence of soil development in Stratum 2A, and its upper and lower boundaries are abrupt. Sand dominates the < 2 mm fraction in Stratum 2A, with proportions ranging between 64% and 72%.1 The mean particle size ranges from 0.33 to 0.56 mm, and the standard deviation (0.41-0.55) indicates that the < 2 mm fraction is well sorted to moderately well sorted. The grain-size distribution of the sand fraction in most samples from Stratum 2A is bimodal, with peaks in the coarse and fine fractions. These data suggest that sediment comprising Stratum 2A has two sources. The bulk of the coarse sand probably came from attrition of the bioclastic limestone. However, some of the sediment was blown in from the beach. This interpretation is supported by the mineralogical analysis. The proportion of igneous and metamorphic grains increases from less than 1% in the very coarse and coarse sand fractions to 2-5% in the fine and very fine sand fractions. Stratum 2B is a thin, discontinuous layer of calcareous, reddish brown (5YR 5/4, dry) sandy loam immediately below. It is 1-8 cm thick and limited to areas where there is evidence of burning within 2A. A good example of this spatial relationship is the reddish zone (Stratum 2B) beneath the pile of ash composing Feature 1 in Stratum 2A. The color of Stratum 2B clearly is a result of thermal alteration of the sediment. The mean particle size in Stratum 2B is 0.41 mm; hence, it is slightly coarser than 2A. The fine-grained matrix has a high proportion of sand (72%), with coarse grains dominating the sand fraction. Analysis with a binocular microscope revealed that the abundance of coarse sand is a result of fusing of smaller gains. Given that Stratum 2B only occurs beneath hearths, the fusing of grains is attributed to intense heat generated by fires in the overlying cultural features.

8Stratum 3 is a sterile zone that often separates Stratum 2 from Stratum 4. Stratum 3 is 15-30 cm thick across most of the site, but it is much thinner or absent in several areas, especially near the front and back of the shelter. It consists of loamy sand and sandy loam representing accumulations of sediment by attrition and aeolian processes. Also, large and small fragments of rock fall are scattered through the fine-grained matrix. Stratum 3 is subdivided into two sedimentary units: Stratum 3A and 3B. Stratum 3A is a homogeneous layer of loose, pale brown (10YR 6/3, dry) sandy loam immediately below Stratum 2. The proportion of sand in 3A is about 60%, with modes in the fine and coarse fractions. Stratum 3A also is characterized by relatively high proportions of silt (22.6%) and clay (16.5%). The mean particle size is 0.34 mm with a standard deviation of 0.44 (well sorted). The grain-size data suggest that 3A is composed of a fairly large volume of aeolian sediment in addition to material generated by attrition and roof fall. The relatively high proportion of igneous and metamorphic mineral grains within the sand fraction supports this interpretation. Most notably, non-carbonate grains comprise up to 20% of the very fine sand fraction. Stratum 3B is a homogeneous layer of loose, very pale brown (10YR 7/4, dry) sandy loam immediately below 3A, or below 2B where 3A is missing. Stratum 3B was distinguished from 3A in the field on the basis of texture. The proportion of sand in 3B is 78%, with modes in the coarse and very coarse fractions. This bimodal grain- size distribution resembles the one for the bedrock sample. Mean particle size for 3B ranges from 0.52 to 0.59 mm, and the standard deviation (0.48-0.52) indicates that the < 2 mm fraction is well sorted to moderately well sorted. Although some of the sand grains are rounded, the proportion of igneous and metamorphic grains is 1% or less in all of the sand fractions. Hence, 3B appears to be largely composed of attrition sediment, with only a small amount of wind-derived material At a few places, cultural features in Stratum 2 truncate strata 3A and 3B. However, virtually no cultural materials were found in situ within Stratum 3.

9Stratum 4 consists of an extremely dense accumulation of bones, primarily hippos, resting directly on the shelter’s bedrock floor. The bones are in a loose, sandy loam matrix that has been darkened by fine, powdery charcoal. Stratum 4 is 10-50 cm thick and is distributed throughout most of the interior of the collapsed rockshelter. The bulk of the sediment composing this unit accumulated through attrition and aeolian deposition. However, boulder size and smaller clasts representing roof fall were common. Stratum 4 is subdivided into four sedimentary units: Stratum 4A, 4B, 4A/B, and 4C (Mandel, 1999). Stratum 4A is a laterally discontinuous unit of loose, light brownish gray (10YR 6/2, dry) sandy sediment above the main bone bed within Stratum 4. The texture of the < 2 mm fraction in Stratum 4A is loamy sand with a sand content of 85%. The mean particle size is 1.1 mm, making 4A the coarsest stratum in the rockshelter. Stratum 4B is 10-45 mm thick and consists of a loose matrix of bones, wood ash, and sediment. Simmons (1991, 1999) referred to 4B as a “bone midden” because of the tremendous volume of burned and unburned bone concentrated in this stratum. It is also designated as a sedimentary unit because geologic processes have contributed mineral grains to this portion of Stratum 4. Th fine-grained sediment (< 2 mm) in 4B is grayish brown (2.5Y 5/2, dry) sandy loam. Roof fall is scattered through this stratum, and most clasts are cobble-sized or smaller. Fine sand strongly dominates the sand fraction, with proportions ranging from 50% to 57%. The mean particle size is relatively fine, ranging from 0.17 to 0.31 mm. Hence, there appears to be a significant aeolian component in 4B. The high proportion of igneous and metamorphic minerals (35-40%) and quartz grains (20-30%) in the fine and very fine sand fraction supports this interpretation. In some portions of the rockshelter there are small inclusions of crumb shaped aggregates typical of Stratum 4A within grayish brown (2.5Y 5/2, dry), bone-rich sediment typical of Stratum 4B. Where this occurs, the stratigraphic unit is designated as 4A/B. The grain-size distribution of the sand fraction in 4A/B is bimodal, with primary and secondary modes in the fine and coarse fractions, respectively. The proportion of igneous and metamorphic grains increases from 1% in the very coarse sand fraction to 35–40% in the very fine sand fraction. The proportion of quartz grains similarly increases from 1-2% in the very coarse sand fraction to 20-25% in the fine sand fraction. These findings clearly indicate that Stratum 4A/B is a mixture of aeolian and attrition sediment. Stratum 4C is common between the bone midden (4B or 4A/B) and the bedrock floor of the rockshelter. Although 4C contains faunal remains, the density of bones in this stratum is not as great as that in 4B. Also, 4C does not contain a large volume of ash. The grain-size distribution in 4C is similar to that in 4B and 4A/B. The mean particle size ranges from 0.20 to 0.28 mm, with a standard deviation of 0.31-0.40 (very well sorted to well sorted). Fine sand strongly dominates the sand fraction, and there is a second mode in the coarse fraction of some samples. The abundance of fine sand suggests that there is a significant aeolian component in Stratum 4C. This interpretation is supported by the mineralogy of the sand fraction. The proportion of non-carbonate mineral grains increases from 5-10% in the coarse fraction to 40-50% in the very fine fraction.

10Proper interpretation of Aetokremnos requires close attention to the stratigraphy, especially the two primary strata. Stratum 2 is clearly cultural, containing the majority of the chipped stone and features, along with most of the birds and shell. Also, nearly 4,000 hippo bones were recorded in Stratum 2. Stratum 4, separated by Stratum 3, contains the majority of the hippo remains but also 11.8% of the chipped stone and three features (Table 1). Strata 2 and 4 form the crux of the controversy at the site: it was argued that both strata were cultural and were separated by only a short period of time. We further argued that the limited amount of hippo remains in Stratum 2 was the result of the species becoming quickly depleted, and that the site’s occupants were replacing hippo with other economic resources, such as bird and marine invertebrates. Opponents begrudgingly accepted that Stratum 2 was cultural, representing the oldest occupation on Cyprus, but argued that Stratum 4 was a natural deposition of hippo remains. They contended that the cultural materials from Stratum 2 were mixed into this lower stratum. Likewise, they argued that the hippo remains present in Stratum 2 were also the result of mixing (e.g. Buminovitz and Barkai, 1996; Ammerman and Noller, 2005).





Chipped stone


4,387 (2.0)

24 (0.1)

216 (1.0)

42 (4.1)


17,145 (7.9)

478 (14.9)

3,870 (18.1)

181 (17.7)

Stratum 1

123 (0.1)

42 (1.3)

627 (2.9)

47 (4.6)

Stratum 2

3,966 (1.8)

2,074 (64.7)


628 (61.5)

Stratum 3

503 (0.2)

40 (1.2)

104 (0.5)

3 (>0.1)

Stratum 4

192,335 (88.2)

547 (17.1)

2,074 (9.7)

129 (11.8)

1 “mixed” is primarily from interface of Strata 2 and 4, where there was no Stratum 3
2 Refers to NISP
3 Refers to NISP
4 Refers to MNI only
Table 1: Stratigraphic association of primary data from Akrotiri
Aetokremnos. Percents in parentheses (from Simmons, 2013a).

11Along with stratigraphy, a numerical chronology also was critical to interpreting Aetokremnos. A total of 71 radiocarbon assays are presently avail- able for the site (Tables 2 and 3). Of these, 31 were processed during the excavation seasons (Table 2). These included 17 samples from Stratum 2, five from Stratum 2/4, and six from Stratum 4. Several laboratories determined the radiocarbon ages of charcoal, bone, and organic-rich sediment, and indicated a short Late Epi-Paleolithic occupation of a few hundred years centered around 12000 cal. BP (Simmons and Wigand, 1994; Wigand and Simmons, 1999). The weighted average for Stratum 2 is 12045 cal. BP, and for Stratum 4 it is 10590 cal. BP These were calibrated using programs available in the late 1990s (Stuiver and Braziunas, [1993] and Bard et al., [1990]). Taking the same data, and using the current OxCal 4.2 IntCal13 curve, the results are 12695-12561 cal. BP for Stratum 2 and 11594-11132 cal. BP for Stratum 4. Using either program, this presented an interesting case of reverse stratigraphy, but when questionable bone and sediment determinations were rejected (see detailed discussion in Simmons and Wigand, 1994; Wigand and Simmons, 1999), the weighted average for Stratum 2 is 12100 cal. BP, and for Stratum 4 it is 11960 cal. BP (using OxCal 4.2 these are 12706- 12575 cal. BP for Stratum 2 and 12711-12405 cal. BP for Stratum 4). While this still suggests reverse stratigraphy, it is a much smaller range and shows that both strata 2 and 4 were virtually the same in radiocarbon years, indicating a relatively short time span between them. After excavation, six additional samples (not included in the above averages) were processed and were in concordance with the original determinations. However, the ages on hippo remains were problematic, as bone dates often are, causing some critics to argue that the remains were much older. This issue, however, now appears to be resolved (see below).

Table 2: Radiocarbon determinations from Akrotiri Aetokremnos (calibration using OxCal 4.2, IntCal 13 curve).
Note that all bone determinations are on hippos (details to be found in Simmons and Wigand, 1994; Wigand and Simmons, 1999). Note that Stratum 2-4 refers to the limited portions of the site where Stratum 2 directly overlaid Stratum 4, with no Stratum 3 present.

Table 3: Post-excavation radiocarbon determinations for Akrotiri Aetokremnos (calibration using 0xCal 4.2, IntCal 13 curve).
Note that all determinations excepting 0xA-15989 are on bone (detail to be found in Simmons, 2013a; Vigne
et al., 2009; Zazzo et al., 2015).


12Despite detailed publications focusing on the site’s stratigraphy and geoarchaeology (Mandel, 1999; Mandel and Simmons, 1997; Simmons and Mandel, 2007), skeptics have ignored evidence that conclusively shows that there is no stratigraphic disturbance or re-working at Aetokremnos. They concede that Stratum 2 is cultural, but argue that Stratum 4 is a natural accumulation, and that any association of artifacts and hippos is due to later mixing by the humans responsible for Stratum 2. Buminovitz and Barkai (1996), for example, actually redrew one of the original published sections, and Ammerman and Noller (2005) concluded that critical evidence was missing, even though all aspects of the site were thoroughly addressed in the publications noted previously. Ammerman (2013) even suggests that actual data were ignored in favor of promoting a personal agenda arguing for human extinction. Many of these criticisms appear to come from either uncritical reading or totally ignoring relevant and well-published data. Hence, several points are reiterated here, focusing on geoarchaeological aspects of the site. Other indications of a cultural and hippo association have been thoroughly discussed elsewhere (Simmons, 1996, 1999, 2004 and 2013a).

13The first point is that, contrary to some suggestions (Ammerman and Noller 2005; Vigne et al., 2013), Aetokremnos was not formed by a sinkhole that acted as a natural trap for the hippos. The primary problem with this assertion is that there is no karst on the Akrotiri Peninsula, and the bedrock and hydrogeology are unsuitable for the development of sinkholes. Another important line of evidence indicating the absence of a sinkhole at Aetokremnos is the sequence of roof debris. With collapsed sinkholes, first the roof falls in, and then animal remains accumulate on top of the roof debris. At Aetokremnos, the collapsed roof, which we documented prior to excavations, mantled and sealed the intact cultural deposits comprising Stratum 2, and the Stratum 4 bones were directly on the shelter’s bedrock floor and not on roof debris from a collapsed sinkhole.

14A second point is that there are rockshelter analogies all around Aetokremnos. Indeed, an adjacent shelter served as a test for rapid sediment accumulation. An excavated test pit revealed rapid aggradation of attrition and aeolian sediment over a period of only a few years.

15The third point is the skeptics’ argument that Stratum 2 is cultural but Stratum 4 is natural, disturbed by later cultural activities. This issue is, of course, critical. The stratigraphic sequence at Aetokremnos, however, is nearly textbook perfect. With the exception of one Stratum 2 feature cutting into Stratum 4, there is absolutely no evidence of anthropogenic mixing of deposits. Also, there also are no biogenic features, such as krotovina (filled-in animal burrows), indicative of mixing. And yet, there are artifacts, features, and burned bone in Stratum 4 that could not have gotten there by mixing. Likewise, the limited amount of hippo remains in Stratum 2 was not moved upward by humans digging into Stratum 4; there simply is no evidence for this. Despite the strong evidence presented by Mandel and Simmons (1997), Mandel (1999), and Simmons and Mandel (2007), indicating clear separation between strata 2 and 4, some authors (e.g. Ammerman and Noller, 2005; Bunimovitz and Barkai, 1999; Zazzo et al., 2015) continue to state that there is mixing caused by human activity in Stratum 2 digging into Stratum 4.

16A fourth point is that while a detailed taphonomic study was not conducted at the time of excavation, this task was accomplished in Wopschall’s (2015) recent Ph.D dissertation. Her work approached the site from a variety of perspectives, including post-depositional taphonomic processes, assessment of skeletal representations and its relationship to human behavior, and demographic profiles. These lines of inquiry led her to conclude that human involvement was in fact an important variable in Stratum 4. Part of the study, for example, was a demographic analysis that compared the Aetokremnos hippos to four modern ecological studies – three culling and one natural. The site’s assemblage is very similar to the culling assemblages and quite distinct from the natural one.

17A fifth point is chronological. As previously noted, there is a large suite of radiocarbon ages for Aetokremnos, and skeptics have argued that the bone was likely older than the cultural materials. Since chronology is so crucial, we have provided all of the radiocarbon determinations from Aetokremnos here, including a new high precision analysis (Zazzo et al., 2015) that has provided 35 new ages determined on bone (Table 3). We present them in uncalibrated and calibrated years before present (BP) and do not analyze them further, as this has been done elsewhere (Manning, 2010; Wigand and Simmons, 1999). It also is not necessary to provide details on the new radiocarbon ages here, since these are thoroughly discussed by Zazzo et al. (2015). Providing all of the ages in one place, however, allows for a full assessment. It is important to note, as well, that the newer ages, taking advantage of more precise refinements in radiocarbon dating, do not substantially differ from the original ages, contra to Ammerman’s (2013) belief that the quality of most of the original Aetokremnos determinations should be considered suspect due to the level of radiocarbon technology in the 1990s. What is important to observe is that the majority of the new ages on the hippo bones are nearly identical to the charcoal ages. Indeed, Zazzo et al. (2015) argue that the “best” ages are the ones originally obtained on charcoal, which included six samples, five of which were from Stratum 2. These six ages averaged 10506 ±181 BP (12738-11776 cal. BP), while the five from Stratum 2 averaged 10495 ±200 BP (12760-11643 cal. BP). They further argue that of the new ages, those determined on seven calcined bone are perhaps the most accurate of their samples, and these average 10552 ±141 (12724-12051 cal. BP), which is very close to the charcoal determinations (Zazzo et al., 2015). Hence, these ages are statistically the same as the mean age of the charcoal. Furthermore, the new bone ages from Stratum 4 are only slightly older than the ages for Stratum 2, which is in stratigraphic accord.

18If the new ages were older than the ones determined on charcoal, the original interpretation of an association of living hippos with humans would have been an issue. However, the new ages confirm our original interpretation. Despite the close synchrony of the new hippo ages with the existing charcoal ones, though, Zazzo et al. (2015) still argue against a direct human/hippo connection. They come to this conclusion on their belief that the abundant burned bone at Aetokremnos is the result of humans using hippo bone from animals that had been dead for several centuries. This is primarily based on additional chemical studies that they conducted on a small sample (see below); such analyses, however, are not definitive.

19A sixth point is that critics often have cited the lack of cutmarks as indicating that the hippos were not butchered (Olsen, 1999; Ammerman and Noller, 2005, Zazzo et al., 2015; Binford, 2002). Furthermore, these authors argue that the site’s small scrapers would have been inadequate butchering tools. The entire issue of cutmarks is immensely complex, but previous studies show them to rare or absent in many faunal assemblages. A classic example is Frison’s (1989) elephant butchery experiments, in which biface reduction flakes were adequate for butchering but left no cut marks. Thus, much of the chipped stone from Aetokremnos could have functioned in butchering activities, particularly depending on how the animals were cooked and processed.

20Finally, there is the issue of burning. As noted above, in addition to the radiocarbon analysis, Zazzo et al. (2015) also conducted chemical studies on some of the burned bone samples, including some heavily burned bone that is “blue-green” in color. They conducted several physic-chemical analyses that included FT-IR (Fourier transform infrared) spectroscopy, stable carbon isotope analysis, micro- PIXE (proton-induced X-ray emission), and diffuse reflectance spectroscopy studies. These analyses led them to conclude that the bone at Aetokremnos was likely burned dry for fuel by humans some centuries after the animals’ death, which they conclude was perhaps some 600 years previous to what the radio-carbon determinations suggest. Zazzo et al. (2015) do not appear to consider that if this were indeed the case, humans were still active in Stratum 4, since most of the burned bone is from that stratum, and there is no evidence of mixing. It also must be realized that the amount of “blue-green” bone was very small in the total faunal assemblage. For the sake of argument, even if one accepts Zazzo et al.’s contention that this bone was burned centuries after the animals’ deaths, it is not representative of the entire assemblage. Indeed, since it is argued that the site was occupied for a few hundred years, previously burned bone from early in the occupation could have been accidentally re-burned, resulting in the blue-green bone.

21These chemical analyses seem inconclusive, based on many studies of bone combustibility (Théry-Parisot and Costamagne, 2005; Théry-Parisot, Chabal, and Costamagne, 2010). Indeed, Théry-Parisot et al. (2005) note that “...bone burns as long as organic material is present. When the organic material is consumed, the mineral part, which makes up most of the bone, does not burn.” Hence, it seems highly unlikely that bone that was several hundred years old would have been useful for fuel, or could even have been burned. Additionally, Wopschall’s (2015) study also concludes that the bone was burned while fresh.


22We stand by our original conclusion of a human and hippo connection (Simmons, 1999, 2004, 2013a, 2014). This is re-affirmed by Wopschall’s (2015) taphonomic study and by the new radiocarbon ages as well as the original analyses. Without evidence of a sinkhole or stratigraphic mixing, one has to explain how over 500 hippos, three dwarf elephants, over 500 bird bones, over 2 000 shell fragments, and around 12% of the chipped stone got into Stratum 4, unless people collected them elsewhere and carried them into the shelter, which seems highly unlikely. These facts should unravel the story of a natural deposition. Certainly one site alone will not prove or disprove a human-caused extinction event. There are, after all, many nuanced extinction scenarios, with some caused by humans and others driven by environmental change, or by combinations of both (e.g. Diamond, 1989; Grayson, 1991; Martin and Klein, 1984). Nevertheless, the claim made for a human role in extinction at Aetokremnos really should not be all that controversial, since it has been demonstrated time and time again that when humans invade an island, they are rapid agents of extinction (Simmons, 2013b). In conclusion, when one examines all of the arguments that have been made against a human and hippo association at Aetokremnos, none withstand careful scrutiny. Conclusions are supported by a rigorous geoarchaeological investigation, and combined with other data provide a compelling case. Certainly disagreement in archaeology is healthy, and we are pleased that after 20 years, this dialogue continues. Unfortunately, many of the site’s critics have ignored a corpus of well published and detailed site data, and have sometimes resorted to unnecessary ad hominem attacks that serve little purpose. We are sure that the only evidence that would convince some of the more hardcore skeptics would be human remains directly associated with the hippos. This is an unlikely scenario, but we believe that Aetokremnos provides more than the potential “smoking gun” that Cherry (1990) hinted at, and that Knapp (2010) elaborated upon, in implicating humans as one causal element in the extinction of Cyprus’ endemic fauna. These were animals who previously had no known predators and who probably were nowhere as fierce as their large modern counterparts. Accordingly, they would have been “naïve” and easy to hunt, particularly if they were already stressed due to the climatic harshness of the Younger Dryas. Their ultimate demise at the hands of humans is the most parsimonious explanation of this intriguing site. Our use of multiple lines of evidence, coupled with the new radiocarbon ages determined on hippo bones and the taphonomic study, should lay to rest the controversial aspect of Aetokremnos, with the realization that Cyprus, like other islands, provides a case study of humans meeting endemic fauna, with an unfortunate consequence for the latter. What Aetokremnos does is push this common scenario back into the late Pleistocene.


23First and foremost, we would like to thank the Department of Antiquities for their support throughout the Aetokremnos project. Officials at both the central headquarters in Nicosia and at the Kourion Museum in Kouklia were of the upmost assistance. The Western Sovereign Base Area Archaeological Society also was of considerable assistance, and we would like, in particular, to thank Frank Garrod for his continued help. The Cyprus American Archaeological Research Institute (CAARI) provided much logistical and local assistance as well. David Reese had the huge task of analyzing the bone from the excavations, and his tireless efforts are greatly appreciated. The late William Farrand also visited the site and offered valuable geoarchaeological insights. We also thank the various staff members, too numerous to individually mention here, for their dedicated support on the project. Primary financial support came from the National Science Foundation and the National Geographic Society, with additional support from a variety of agencies, including the National Endowment for the Humanities, the Institute for Aegean Archaeology, the Brennan Foundation, the Desert Research Institute in Reno, and the University of Nevada at Las Vegas.



AMMERMAN A., «Introduction», Eurasian Prehistory, 10, 2013, p.9-30.

AMMERMAN A., NOLLER J., «New Light on Aetokremnos», World Archaeology, 37, 2005, p.533-543.

BARD E., HAMELIN B., F AIRBANKS R., ZINDLER A., «Calibration of the 14C timescale over the past 30,000 years using mass spectrometric U-Th ages from Barbados corals», Nature, 345, 1990, p.405-410.

BINFORD L., «Review of Faunal extinctions in an island society: pygmy hippopotamus hunters of the Akrotiri Peninsula, Cyprus by A. Simmons», American Antiquity, 65, 2000, p.771.

BUNIMOVITZ S., BARKAI R., «Ancient bones and modern myths: ninth millennium BC hippopotamus hunters at Akrotiri Aetokremnos, Cyprus?», Journal of Mediterranean Archaeology, 9, 1996, p.85-96.

CHERRY J., «The First Colonization of the Mediterranean Islands: a Review of Recent Research», Journal of Mediterranean Archaeology, 3, 1990, p.145-221.

DIAMOND J., «Quaternary megafaunal extinctions: variations on a theme by Paganini», Journal of Archaeological Science, 16, 1989, p.167-175.

FAURE M., GUERIN C., SONDAAR P., «Hippopotamus Minutus Cuvier, mise au point», in Buffetaut E., Mazin J., Salmon E., Actes du Symposium Paleontologique Georges Cuvier, Montbeliard, Montbeliard, France, 1983.

FER ENTINOS G., GKIONI M., GER AGA M., PAPATHEODOROU G., «Early Seafaring Activity in the Southern Ionian Islands, Mediterranean Sea», Journal of Archaeological Science, 39, 2012, p.2167-2176.

FRISON G., «Experimental use of Clovis weaponry and tools on African elephants», American Antiquity, 54, p.766-784, 1989.

GRAYSON D., «Late Pleistocene Mammalian extinctions in North America: taxonomy, chronology, and explanations», Journal of World Prehistory, 5, 1991, p.193-231.

HEYWOOD H., «The archaeological remains of the Akrotiri Peninsula», in Swiny H., Ancient Kourion Area, Nicosia, Zavallis Press, 1982.

KNAPP A.B., «Cyprus’s earliest prehistory: seafarers, foragers, and settlers», Journal of World Prehistory, 23, 2010, p.79-120.

LORT J., «Geophysics of the Mediterranean sea basins», in Nairm A., Kanes W., Stehili F., The Oceanic Basins and Margins, Vol. 4A: The Eastern Mediterranean, New York, Plenum Press, 1977.

KILMER V., ALEXANDER L., «Methods of making mechanical analyses of soils», Soil Science, 68, 1989, p.15-24.

MANDEL R., «Stratigraphy and sedimentology», in Simmons A., Faunal Extinctions in an Island Society: Pygmy Hippopotamus Hunters of the Akrotiri Peninsula, Cyprus, New York, Kluwer Academic/Plenum Publishers, 1999.

MANDEL R., SIMMONS A., «Geoarchaeology of the Akrotiri Aetokremnos rockshelter, southern Cyprus», Geoarchaeology, 12, 1997, p.567-605.

MANNING S., «Appendix. A New Radiocarbon Chronology for Prehistoric and Protohistoric Cyprus, ca. 11,000-1050 Cal BC», in Knapp, B., The Archaeology of Cyprus: From Earliest Prehistory Through the Bronze Age, Cambridge, Cambridge University Press, 2013.

MARTIN P., KLEIN R. (eds.) Quaternary extinctions: a prehistoric revolution, Tucson, University of Arizona Press, 1984.

MOREL S., The geology and mineral resources of the Apsiou Akrotiri Area, Geological Survey Department, Cyprus, Memoir No. 7, Nicosia, Government Printing Office, 1960.

OLSEN S., «Investigation of the Phanourios bones for evidence of cultural modification», in Simmons A., Faunal Extinctions in an Island Society: Pygmy Hippopotamus Hunters of the Akrotiri Peninsula, Cyprus, New York, Kluwer Academic/Plenum Publishers, 1999.

PIPER C., «The Determination of Calcium Carbonate; Rapid Titration Method», in Piper C., Soil and Plant Analysis, New York, Wiley Interscience, 1950.

REESE D., The Pleistocene vertebrate sites and fauna of Cyprus, Nicosia, Geological Survey Department, Ministry of Agriculture, Natural Resources and Environment, 1995.

SIMMONS A., «Extinct Pygmy Hippopotamus and Early Man in Cyprus», Nature, 333, 1988, p.554-557.

SIMMONS A., «Humans, Island Colonization and Pleistocene Extinctions in the Mediterranean: the View from Akrotiri Aetokremnos, Cyprus», Antiquity, 65, 1991, p.857-869.

SIMMONS A., «Whose myth? Archaeological data, interpretations, and implications for the human association with extinct Pleistocene fauna at Akrotiri Aetokremnos, Cyprus», Journal of Mediterranean Archaeology, 9, 1996, p.95-103.

SIMMONS A., Faunal extinctions in an island society: pygmy hippopotamus hunters of Cyprus, New York, Kluwer Academic/Plenum Publishers, 1999.

SIMMONS A., «Bitter hippos of Cyprus: the island’s first occupants and last endemic animals—setting the stage for colonization», in Peltenburg E., Wasse A, Neolithic Revolution: New Perspectives on Southwest Asia in Light of Recent Discoveries on Cyprus, Oxford, Levant Supplementary Series, Vol. 1, 2004.

SIMMONS A., The Neolithic Revolution in the Near East: Transforming the Human Landscape, Tucson, University of Arizona Press.

SIMMONS A., «Akrotiri Aetokremnos (Cyprus) 20 years later: an assessment of its significance», Eurasian Prehistory, 10, 2013a, p.139-155.

SIMMONS A. ,»Island faunas», in McDade, Grzimek’s Animal Life Encyclopedia-Extinction, Farmington Hills, MI Gale-Centage Learning, 2013b.

SIMMONS A., Stone age sailors: paleolithic seafaring in the Mediterranean, Walnut Creek, CA, Left Coast Press, 2014.

SIMMONS A., MANDEL R., «Not such a new light: a response to Ammerman and Noller», World Archaeology, 39, 2007, p.475-482.

SIMMONS A., WIGAND P., «Assessing the radiocarbon determinations from Akrotiri Aetokremnos, Cyprus», in Bar-Yosef O. Kra R., Late Quaternary Chronology and Paleoclimates of the Eastern Mediterranean, Tucson and Cambridge, Radiocarbon and American School of Prehistoric Research, 1994.

SOIL SURVEY STAFF, «Procedures for collecting soil samples and methods of analysis for soil survey», Soil Survey Investigations Report, 1, 1982.

SONDAAR P., «The island sweepstakes», Natural History, 95, 1986, p.50-57.

STANLEY D., «Post Miocene depositional patterns and structural displacement in the Mediterranean», in Nairn A., Kanes W., Stehili F., The Oceanic Basins and Margins, vol. 4A: the Eastern Mediterranean, New York, Plenum Press.

STUIVER M., BRAZIUNAS T., «Modeling atmospheric 14C influences and 14C ages of marine samples back to 10,000 B.C.», Radiocarbon, 35, 1993, p.137-189.

STRASSER T., PANAGOPOULOU E., RUNNELS C., MURRAY P., THOMPSON N., KARKANAS P., MCCOY F., WEGMANN K., «Stone age seafaring in the Mediterranean: evidence from the Plakias region for Lower Palaeolithic and Mesolithic habitation of Crete», Hesperia, 79, 2010, p.145-190.

THÉRY-PARISOT I., CHABAL L., COSTAMAGNO S., The taphonomy of burned organic residues and combustion features in archaeological contexts, Valbonne, Palethnologie, 2, 2010.

THÉRY-PARISOT I., COSTAMAGNO S., BRUGAL J., FOSSE P., GUILBERT R., «The use of bone as fuel during the Paleolithic, experimental study of bone combustible properties», in Mulville J., Outram A., The Zooarchaeology of Fats, Oils, Milk, and Dairying, Oxford, Oxbow Books, 2005.

THÉRY-PARISOT I., COSTAMAGNO S., «Fuel properties of bones. Experimental data and archaeological reflections on their use in Paleolithic sites», Gallia Prehistoire, 47, 2005, p.235-254.

VIGNE J.-D., ZAZZO A., CARRÈRE I., BRIOIS F., GUILAINE J., «The transportation of mammals to Cyprus sheds light on early voyaging and boats in the Mediterranean Sea», Eurasian Prehistory, 10, 2013, p.157-176.

WIGAND P., SIMMONS, A., «The dating of Akrotiri Aetokremnos», in Simmons A., Faunal extinctions in an island society: pygmy hippopotamus hunters of the Akrotiri Peninsula, Cyprus, New York, Kluwer Academic/Plenum Publishers, 1999.

WOPSCHALL K., Akrotiri Aetokremnos and the Cypriot Pygmy Hippopotamus: An Interdisciplinary Look at a Late Pleistocene Large Mammal extinction, thèse de doctorat, université du Washington, 2015.

ZAZZO A., MATTIEU L., QUIES A., REICHE I., VIGNE J.-D., «Direct dating and physic-chemical analyses casts doubts on the coexistence of humans and dwarf hippos in Cyprus», PLOSOne 10.1371, 2015, p.1-13.


1 All laboratory data are presented in Mandel and Simmons (1997).

Table des illustrations

Légende Figure 1: Location map of site and Akrotiri Peninsula.
Fichier image/jpeg, 196k
Légende Figure 2: Site overview photo. The insert shows rooffall overlying and protecting cultural remains.
Fichier image/jpeg, 656k
Légende Figure 3: Photo of faunal remains, showing density once excavation had begun (A) and well-preserved hippo cranium (B).
Fichier image/jpeg, 388k
Légende Figure 4: Artifacts from Akrotiri Aetokremnos-shell beads (A), picrolite bead (B), and thumbnail scrapers (C).
Fichier image/jpeg, 244k
Légende Figure 5: Representative composite stratigraphic drawing (prepared by R. Watters); insert is a photo of the stratigraphyand location A shows generalized location within stratigraphic section.
Fichier image/jpeg, 316k
Légende Table 2: Radiocarbon determinations from Akrotiri Aetokremnos (calibration using OxCal 4.2, IntCal 13 curve).Note that all bone determinations are on hippos (details to be found in Simmons and Wigand, 1994; Wigand and Simmons, 1999). Note that Stratum 2-4 refers to the limited portions of the site where Stratum 2 directly overlaid Stratum 4, with no Stratum 3 present.
Fichier image/jpeg, 464k
Légende Table 3: Post-excavation radiocarbon determinations for Akrotiri Aetokremnos (calibration using 0xCal 4.2, IntCal 13 curve).Note that all determinations excepting 0xA-15989 are on bone (detail to be found in Simmons, 2013a; Vigne et al., 2009; Zazzo et al., 2015).
Fichier image/jpeg, 583k


University of Nevada, Las Vegas, USA,

University of Kansas, USA,

© CNRS Éditions, 2016

Licence OpenEdition Books

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search