The taphonomy and palaeoenvironmental implications of the small mammals from Karain Cave, Turkey

The taphonomy and palaeoenvironmental implications of the small mammals from Karain Cave, Turkey

Journal of Archaeological Science 38 (2011) 3048e3059 Contents lists available at ScienceDirect Journal of Archaeological Science journal homepage: ...

2MB Sizes 14 Downloads 76 Views

Journal of Archaeological Science 38 (2011) 3048e3059

Contents lists available at ScienceDirect

Journal of Archaeological Science journal homepage:

The taphonomy and palaeoenvironmental implications of the small mammals from Karain Cave, Turkey Arzu Demirel a, *, Peter Andrews b, Is¸ın Yalçınkaya c, Ayhan Ersoy c a

Faculty of Arts and Science, Department of Anthropology, Mehmet Akif Ersoy University, Istiklal Campus, Burdur, Turkey Natural History Museum, London, UK c Faculty of Letters, Ankara University, Ankara, Turkey b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 22 October 2010 Received in revised form 1 July 2011 Accepted 4 July 2011

The small mammal accumulations in the Pleistocene deposits of Karain Cave are investigated to identify the predators and possible biases in the fossil assemblages. Seven small mammal assemblages are studied in chronological order from two chambers of the cave, the main chamber E and the side chamber B. The lowermost level within the whole sequence is Proto-Charentien, which corresponds to an early stage of the Middle Palaeolithic. The main part of the material belongs to Middle Palaeolithic layers. The most important aspect of the fossil record in the cave is the human occupation without any interruption through the Pleistocene to Holocene. The small mammal fossil evidence in the cave denotes the presence of opportunistic predators throughout the sequence with one exception, and these produce balanced samples of small mammal faunas in the habitat. The lack of bias in the small mammal faunas allow the interpretation of local environments, showing that partial steppe and arid conditions existed during deposition of the lowermost levels of the Middle Palaeolithic in Karain Cave and that these shifted into more temperate and wooded habitat in the upper levels. Evidence from the side chamber indicates some differences, with a more open grassy environment. In the Mediterranean part of Anatolia the temperate and moist conditions in the Middle Palaeolithic were superseded by more arid conditions in the Upper Palaeolithic, followed by a decrease in steppe conditions during the Epi-Palaeolithic period. Ó 2011 Elsevier Ltd. All rights reserved.

Keywords: Karain Cave Small mammal taphonomy Scanning electron microscope Predators Pleistocene Palaeoecology

1. Introduction The significance of small mammal faunas in palaeoecology and archaeology has long been recognized. Brothwell and Jones’s (1978) study is one of the leading studies in the field. They clearly point out that; “in contrast to larger mammals, this group exploits a considerable range of food resources and take advantage of a variety of microclimates”. Redding (1978) also specifies the relevance of small mammals, especially rodents, to archaeological studies, as they are more sensitive to changes in the local environment; and he emphasizes that the resulting environmental interpretation will be more detailed and precise. This has been demonstrated in detail in an analysis of southern African faunas, where small mammals have much higher correlations with climatic and vegetation variables (Andrews and O’Brien, 2000). Despite being significant ecological indicators, small mammals are widely subject to taphonomic factors and their faunal composition can be substantially changed especially by predation. * Corresponding author. Tel.: þ90 248 213 30 46; fax: þ90 248 213 30 99. E-mail address: [email protected] (A. Demirel). 0305-4403/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.jas.2011.07.003

The study of the processes of preservation and destruction of faunal remains, the subject of taphonomy, directly addresses the processes by which fossils are preserved and the extent to which they represent past organisms, evolutionary patterns and important biological events in the history of the earth (Efremov, 1940). Much of the early work on vertebrate taphonomy has concentrated on large mammals (Lyman, 1994), and while small mammal faunas are subject to a different suite of taphonomic processes from large mammals, both are subject to taphonomic processes that may alter the original composition of the faunal community. Particular species may be selected for or against, with the result that while some species may be missing altogether from a fossil assemblage compared with the living community from which it was derived, other species may be present in numbers unrelated to their original abundance (Korth, 1979; Dodson and Wexlar, 1979; Andrews and Evans, 1983; Kowalski, 1990). Archaeological small mammal assemblages accumulated as predator assemblages must have their accumulating agents identified as far as possible in order to identify potential predator bias get a better picture of past environments. The first detailed study on the taphonomy of small mammal remains was presented by Andrews (1990), and this has been

A. Demirel et al. / Journal of Archaeological Science 38 (2011) 3048e3059

followed by an increasing number of studies (Fernandez-Jalvo and Andrews, 1992; Denys et al., 1996, 1997; Fernandez-Jalvo et al., 1998; Dauphin et al., 1994, 1999; Matthews, 1999; Pinto Llona and Andrews, 1999; Avery, 2002; Andrews and Jenkins, 2000). Small mammal accumulations in archaeological sites are often the prey assemblages of predators, mainly owls, especially when the rodent remains are in question, as rodents comprise the main part of their diet. Predator activity is clearly defined by the digestion patterns on the small mammal remains; and these patterns are best distinguished using the scanning electron microscope (SEM), it is because the SEM has superior resolution when compared to an optical microscope and enables to examine surfaces in great detail. The present paper on the taphonomy of the small mammal remains of Karain Cave, Turkey, follows Andrews (1990) methodology, and it stands as a first study on small mammalian taphonomy in Turkey. Karain Cave is unique for revealing the only known remains of Neanderthals in Anatolia. Abundant small mammal remains have been discovered during the systematic excavations in the Pleistocene sequences of the cave, and the purpose of this paper is to describe the taphonomy of the small mammal accumulation in two of the excavated chambers of the cave and to use this to reconstruct the palaeoecology of the surrounding area. 2. Karain Cave Karain Cave (Black Cave) is a complex of caves located 30 km. NNW of Antalya on the southern coast of Turkey (Figs. 1 and 2). It lies at the edge of the extensive travertine Antalya Plain, set in the Cretaceous limestone of the southern flanks of the Taurus Mountains. The cave complex is built up in a series of tiers reflecting their karst origins and ranging between 430 and 450 m above sea level (Albrecht, 1988), while the entrance to the cave affords an excellent view over the whole plain. 2.1. The sedimentological context and faunal background The main excavations at Karain Cave were conducted in chambers E and B (Yalçınkaya et al., 1993) (see Fig. 2). Chamber E is the


main chamber (Kökten’s (1964) Chamber) that contains Lower and Middle Palaeolithic deposits. The sequence is composed of interfingering colluvia, travertines (including speleothems), layers of clayey-silts, and calcitic concretions often associated with the development of paleosols (Otte et al., 1995) (Fig. 3). Thermoluminescence (TL) and Electron Spin Resonance (ESR) dating studies provide an age range for the sediments in Karain Cave 160,000 to 60,000 years for the middle Palaeolithic layers. Four consolidated travertine layers have been identified (labelled “1” to “4” by black triangles on Fig. 3) in Chamber E, and these are believed to correspond to palaeoclimatic changes that interrupted the processes of sedimentation within the cave (Otte et al., 1995). These layers have revealed remains of Homo neanderthalensis (Tas¸kıran, 2002). The faunal assemblages have a mixed character, and several taphonomic groups (sensu Gautier, 1987) could be identified in the main Chamber (Otte et al., 1998): (a) The principal group consists of consumption refuse resulting from animals which were hunted and transported to the cave by human agents; (b) Another group identified as workshop refuse includes elements transported by humans but which do not possess nutritional value; animals which could have used the cave for hibernation or habitation; (c) Penecontemporaneous intrusions, which include the remains of animals not intentionally transported to the cave by humans: freshwater gastropods, hares, small birds, voles, murids, sousliks, hedgehogs, shrews and other rodents and amphibians. These could have arrived as prey of different predators such as fox, small mustelids, and raptors: owls (Bubo bubo and Asio otus) and buzzards (Buteo rufinus). In his early publications, Kökten (1964), points out that Chamber B was preferentially occupied after the Middle Palaeolithic period. In the light of new discoveries at the site, Yalçınkaya points out that the uppermost Epi-Palaeolithic finds are derived from the other chambers of the cave and in situ cultural layers start with Karain type Mousterian in the main Chamber E at the uppermost levels

Fig. 1. Geographical location of the Karain Cave (Otte et al., 1995: 288).


A. Demirel et al. / Journal of Archaeological Science 38 (2011) 3048e3059

Fig. 2. The Karain Cave System (Yalçınkaya, 1991: 59).

(Yalçınkaya, pers.comm.). Chamber B includes three different geological layers belonging to the Pleistocene period. The lowermost Middle Palaeolithic layers are succeeded by the Upper Palaeolithic layers which were dated back to 28,000 years. The uppermost layers are Epi-Palaeolithic and these are contemporary with the dwelling floor of the first inhabitants of Öküzini Cave in the same area with C14 dating of 16,000 years (Tas¸kıran, 2002). These layers are also called first, second and third cultural phases

respectively (Özçelik, 2003). The contemporaneity of Middle Palaeolithic layers is also observed in the lithic tool production technology in both chambers (Yalçınkaya et al., 2001). 3. Materials and methods The small mammal remains that are the subject of this paper are limited to the rodent cranial remains unearthed during the

Fig. 3. Karain Cave, the Main excavation block, south profile (Otte et al., 1995: 289).

A. Demirel et al. / Journal of Archaeological Science 38 (2011) 3048e3059

systematic excavations. The small mammal remains from Chamber E were collected during the 1989e2000, and from Chamber B during the 1996, 1998 and 1999 field seasons. The excavated material was screened and sorted in the field. The screen size used in the field for sorting the small mammal material is 1 mm. The taxonomic identifications of rodents were made by Emmanuel Desclaux of Département de Préhistoire du Muséum National d’Histoire Naturelle, CNSR, France. The fossil collection available for this analysis is housed in the University of Ankara, Faculty of Letters, Prehistory Laboratory, Turkey. Table 1 displays the amount of analyzed material, MNI and NISP of both chambers. The cave sediments were excavated in 10 cm units, and this practice sometimes resulted in units which had incorporated two distinct sedimentary layers. This could potentially affect the interpretation of the faunal data, but in the opinion of the excavators, such “mixed” units are rare and the grouping of geological layers into archaeological units (Table 2) “substantially obviates the problem” (Otte et al., 1998). For this study, the material from the archaeological layers is grouped according to the archaeological units so as to correlate the taphonomic with the archaeological results (see Table 2). Different proportions are used to show the abundance of the small mammals consumed by the predators in the archaeological assemblages. MNI is calculated depending on the most abundant skeletal element in the assemblage and shows the number of the prey eaten by the predators in the each assemblage. On the other hand, number of small mammals for each assemblage is calculated separately depending on the most abundant skeletal element belonging to same genus or species and aims to show the abundance of the same genus or species in the assemblage. The taphonomic modifications for the small mammal remains (NISP 3257, MNI 324, and number of small mammals 507) have been examined under a SMZ 2T light microscope under 8e20 magnifications at the University of Ankara, Faculty of Letters, Prehistory Laboratory. Scanning electron microscope analyses were made with a JSM-6400 JEOL, Electron Microscope equipped with NORAN System 6 X-ray Microanalysis System & Semafore Digitizer, in secondary and backscattered electron mode which is held in the Middle East Technical University, Department of Metallurgical & Materials Engineering, Scanning Electron Microscopy (SEM) Laboratory. Specimens were mounted on stubs and sputtercoated with gold. The methodology of the taphonomic analysis had the aim of identifying the processes of accumulation of the small mammal faunas present in the Karain sequence. In most cases they appear to be prey accumulations of predators. It is assumed that the predators identified in the course of this study had similar hunting preferences in the past with the living ones, but there is no direct way of testing this assumption. It is also assumed that distributions and habitats of small mammalian species in the Karain sequence have not changed from living populations. These assumptions are implicit in any attempt to analyze past environments by whatever method, and all conclusions in this paper are based on this uniformitarian principle. Table 1 Number of identified specimens (NISP) and minimum numbers of individuals (MNI) for Chamber E and B.

Mandible Maxilla Isolated incisor Isolated molar NISP MNI

Chamber E

Chamber B


348 154 8 979 1489 175

121 68 591 988 1768 149

469 222 599 1967 3257 324


Table 2 Correlation between the geological and archaeological units and the oxygen isotope stages (Otte et al., 1998: 419). Geological layers



Archaeological units

Oxygen isotope stages

Human bones present

I.1 I.2e6 I.7 II.1e3 III.1eIII.2 III.2.1. III.3e5 IV.1 IV.2e4 IV.5 V.1 V? 2e4 VI

0 e 1 e 2e e 3 e e e 4 e 5

10.000 70e60.000 120e110.000 e 250e200.000 e 350e300.000 e e e 400e370.000 e e

U.P. I H G F F E D C B A A A

1,2,3 4 5 6 7 8 9 10 10 10 11 12 13

e X e e X e X e e e e e e

The main taphonomic modification examined was the effects of digestion on the teeth and jaws of all species of rodents. There are two reasons for this: firstly, it has been demonstrated that this provides the best diagnostic evidence of the mode of accumulation of predator assemblages of small mammals (Andrews, 1990); and secondly, post-depositional damage is so great at Karain Cave that initial breakage and skeletal element distributions have been obscured. Five categories of modification were distinguished for arvicolid (microtine) rodents, following Andrews (1990) methodology. The Karain small mammal faunas, however, are dominated by murid rodents e especially the Chamber E faunas e and it quickly became apparent that the digestion categories applied to murids differ in degree from those for arvicolids. For example, when prey assemblages of living predators are examined, it is observed that while arvicolid molars have light digestion, murid teeth in the same assemblages show no sign of digestion. Similarly, in modern prey assemblages where arvicolid teeth are moderately digested, murid teeth in the same assemblage and subjected to the same degree of digestion show only early stages of digestion. This disparity was recognized implicitly in Andrews (1990), but we have attempted to formalize it here. Table 3 relates degrees of digestion for two main categories of small mammals to the different predator species that produce them, distinguishing between arvicolids, with acute salient angles to their teeth, and murids and soricids, with lower crowned teeth. This shows, for example, that predators that produce digestion category 3 on arvicolid teeth produce only digestion category 1 on murids, glirids, gerbillids, and sciurids. It is likely that some differentiation will be found with further study between murids and soricids. 4. Results 4.1. Chamber E Microfauna from archaeological assemblage E: Archaeological assemblage E comes from geological levels III.3e5 (see Table 2) and is dated to 350,000 to 300,000 years (Charantien culture). Level III.6 is a calcite level and three specimens from this level are also included with this assemblage. This is the assemblage which has the lowest MNI (n ¼ 20) and number of species (n ¼ 8) within Chamber E, but it is the oldest faunal assemblage in the cave, and includes some human remains (two vertebrae and a femoral diaphysis fragment) (Otte et al., 1998). The number of incisors is small, and digestion is heavy (Fig. 4). Isolated molars are more digested compared to in situ molars, with 21% digestion of molars affected. Degree of digestion is light and


A. Demirel et al. / Journal of Archaeological Science 38 (2011) 3048e3059

Table 3 Digestion categories compared for arvicolids and murids/glirids. Digestion Stage

Digestion absent or minimal Molars 0e3% Incisors 8e13% Light/moderate digestion Molars 4e6% Incisors 20e30% (tips only) Moderate digestion Molars 18e22% Incisors 50e70% Heavy digestion Molars 50e70% Incisors 60e80% Extreme digestion Molars 50e100% Incisors 100% (dentine corroded)

Predator Category

Molar digestion

Incisor digestion


Barn owl, long-eared owl, short-eared owl, Verreaux eagle owl

Barn owl, short-eared owl, snowy owl



Snowy owl, spotted eagle owl, great grey owl

Long-eared owl, Verreaux eagle owl, great grey owl



European eagle owl, tawny owl

European eagle owl spotted eagle owls, tawny owl, little owl



Little owl, kestrel, peregrine

Kestrel, peregrine



Hen harrier, buzzard, red kite

Hen harrier, buzzard




Percentages of digestion are shown on the left for molars and incisors separately, and these are categorized into five categories numbered 1 (low levels of digestion) to 5 (high levels of digestion). The predators that produce these digestion categories are shown on the right (adapted from Andrews, 1990).

mostly seen on murid molars, which are more abundant at this level than other taxa and in this assemblage Apodemus species are the dominant prey (MNI ¼ 10) (Table 4), comprising 65% of the fauna. Light degree of digestion on murid molars is equivalent to digestion category 3 of arvicolid molars because of the differences in tooth structure as explained earlier (see Table 3). Physical abrasion is observed on some specimens, a feature that is common in cave environments (Fig. 5). Microfauna from archaeological assemblage F: Archaeological assemblage F belongs to geological levels III.1, III.2 and III.2.1. The first two levels have been correlated with isotope stage 7, dated between 250,000e200,000 years (Table 2) and represent the lowermost levels of Zagros or Karain type Mousterian in the cave. This assemblage also produced some human remains which are attributed to Neanderthals including one-third phalanx and three phalanges of the left hand, a left mandibular fragment and two bone fragments (radius and cuboid) (Otte et al., 1998). All the digested in situ molars and most of the isolated digested molars are murid molars. The MNI of Muridae is high (MNI ¼ 92, Table 4) compared to the other species in the assemblage, and murids make up 76% of the microfaunal assemblage. The total digestion percentage for molars is 15.6% (Table 5), most of the murid molars are lightly digested (Fig. 6aeb) and specific digestion pattern to the category 3 predators on the arvicolid molars observed (Fig. 6c).

Fig. 4. Heavy digestion on the incisor of a murid mandible.

Microfauna from archaeological assemblage G: The fauna from archaeological assemblage G also comes from the Middle Palaeolithic layers of the cave (level II 1e3) and is later than assemblage F (Table 2). The taphonomic and taxonomic features of this assemblage are similar to those of assemblage F. Isolated molars are considerably more digested than the in situ molars, both in terms of frequency and degree: 9% of in situ molars and 20.7% of isolated molars have been digested (Table 5). Muridae again is the dominant group in assemblage G, and most of the murid teeth have higher levels of digestion than in the previous fauna. They have lost enamel along the enameledentine border (Fig. 7a,b), a feature present also on the molars of Cricetulus and Sciurus molars, which have similar enamel structure with Muridae (Fig. 7c,d). Arvicolid teeth have been also more heavily digested (Fig. 7e,f). Microfauna from archaeological assemblage I: The fauna from archaeological assemblage I represents the uppermost levels of Chamber E. These levels are the uppermost levels of Zagros or

Table 4 Taxonomic composition and minimum number of individuals of small mammals recovered from four archaeological units, E, F, G and I in Chamber E. E Muridae Apodemus Arvicanthis ectos Mus abbotti Total Muridae (MNI) Arvicolidae Microtus guentheri Microtus arvalis Microtus (Chionomys) nivalis Microtus gud Arvicola sapidus Total Arvicolidae (MNI) Spalacidae Spalax leucodon Gliridae Myomimus roachi Gerbillidae Meriones tristinami Cricetidae Mesocricetus auratus cf. brandti Cricetulus migratorius Sciuridae Sciurus anomalus Total other rodentia (MNI) Total MNI Number of total species




10 2 1 13

75 13 4 92

34 8 3 45

17 1 1 19

2 1 0 2 0 5

10 2 2 6 1 21

6 1 1 3 1 12

3 2 1 3 0 9













0 0

1 2

0 1

0 0

0 2 20 8

0 8 121 13

1 7 64 13

0 5 33 10

A. Demirel et al. / Journal of Archaeological Science 38 (2011) 3048e3059


Fig. 5. Light digestion and physical abrasion on a microtine molar.

Karain type Mousterian with the date of 70,000 to 60,000 (isotope stage 4) (Table 2). It should be noted that this assemblage produced human remains, including two permanent and one deciduous molars attributed to Neanderthals (Yalçınkaya, 1988). Even though murids comprise most of the fauna, arvicolid molars are more digested compared with the murid molars (2.7% and 21.8 respectively). Of the total rodent molars 9.3% of digested and this is closer to category 2 predators for arvicolids (Table 3). Murids are the most abundant group as in the other chamber E microfaunas, but only 2.7% of the murid molars are lightly digested (Fig. 8). 4.2. Chamber B In Chamber B, some differences are observed compared with chamber E, both taxonomically and taphonomically, such as

Table 5 Taphonomic modifications of small mammal jaw and dental elements recovered from four archaeological units E, F, G, I in Chamber E. E n

F %


G %


I %



Anatomical element Mandible 26 100 199 99.5 85 98.9 38 100 Maxilla 7 26.9 92 46 38 44.2 17 44.8 Incisor 0 0 5 1.3 2 1.2 1 1.3 Molar 43 27.6 525 43.8 276 53.5 135 59.2 NISP 76 821 401 191 MNI 13 100 43 19 Number of small mammals 20 121 64 33 Dentition % digestion M in situ 10 17.5 58 12.1 18 9 3 3.3 % isolated M digestion 11 25.6 97 18.5 57 20.7 17 12.6 % digestion I in situ 2 16.7 8 15.9 2 10 2 14.3 % isolated I in situ 0 0 0 0 1 50 0 0 Total M 100 1005 476 226 Light 18 85.7 135 87.1 59 77.6 18 85.7 Moderate 3 14.3 16 10.3 14 18.4 2 9.5 Heavy 0 0 6 3.9 3 4 1 4.8 Extreme 0 0 0 0 0 0 0 0 Total digestion 21 21 155 15.6 75 16 20 9.3 Total I 12 56 22 15 Light 1 50 7 87.5 2 66.7 1 50 Moderate 0 0 1 12.5 0 0 1 50 Heavy 1 50 0 0 1 33.3 0 0 Extreme 0 0 0 0 0 0 0 0 Total digestion 2 16.7 8 14.3 3 13.6 2 13.3

Fig. 6. SEM micrographs of rodent remains from F microfaunal assemblage. (a) Light digestion shown on a Mus molar; (b) Light digestion shown on an Arvicanthis molar; (c) Category 3 digestion as shown on an arvicolid molar (note the rounded corners of the occlusal surface).

increase in the number of arvicolids especially in P.II (Table 6) and the increase in the isolated incisors (Table 7). This is connected with the greater breakage of the jaws in this chamber. The bias against molars is very high, so the digestion percentages of the incisors provide the best evidence for evaluating the predator activity in the


A. Demirel et al. / Journal of Archaeological Science 38 (2011) 3048e3059

Fig. 7. SEM micrographs of rodent remains from G microfaunal assemblage. (a) Light digestion shown on an Apodemus molar; (b) Light digestion shown on an Arvicanthis molar; (c) Light digestion shown on a Cricetulus molar; (d) Light digestion shown on a Sciurus molar; (e) Moderate digestion shown on a Arvicola sapidus molar; (f) Heavy digestion shown on a Microtus guentheri molar.

Fig. 8. A SEM micrograph from I microfaunal assemblage. Heavy digestion shown on an Apodemus molar.

Chamber B microfaunal assemblages. The collecting bias against molars during excavation renders them suspect for taphonomic analysis, so that greater reliance is placed on digestion percentages of incisors, which are not under-represented (Table 7). Microfauna from archaeological assemblage P.III: The lowermost faunal assemblage in Chamber B is level P.III, which is contemporary with the Middle Palaeolithic sequences of Chamber E. The digestion percentage for the incisors from this fauna is 43% (Table 7), this is close to the level produced by European eagle owl which produces 48% digested incisors in prey assemblages (Andrews, 1990) and most of the incisors are lightly digested (Fig. 9). Even the molars are under-represented, 13.1% molar digestion in total is also most close to the produced by European eagle owl. Microfauna from archaeological assemblage P.II: The fauna from level P.II is dated to the Upper Palaeolithic levels of Chamber B and arvicolids have higher numbers only in P.II along with P.I compared to Muridae in the whole sequence (Table 6). This might show the predator’s prey preferences, or it might indicate changing environmental conditions at this time. Digestion percentages are similar to those for P.III (Table 7) and are close to those seen for European eagle owl, tawny owl and little

A. Demirel et al. / Journal of Archaeological Science 38 (2011) 3048e3059


Table 6 Taxonomic composition and minimum number of individuals of small mammals recovered from three archaeological units PIII, PII, and PI in Chamber B. P.III Muridae Apodemus Mus abbotti Total Muridae (MNI) Arvicolidae Microtus guentheri Microtus arvalis Microtus nivalis Microtus gud Arvicola sapidus Total Arvicolidae (MNI) Spalacidae Spalax leucodon Gliridae Myomimus roachi Gerbillidae Meriones tristinami Cricetidae Mesocricetus auratus cf. brandti Cricetulus migratorius Cricetus cricetus Sciuridae Sciurus anomalus Dipodidae Allagtaga Rodentia (only incisors) Total other rodentia Total MNI Number of total species



2 12 1 15 8 2 1 0 2 1 14

4 24 2 30 38 5 2 0 4 2 51

1 1 4 6 9 2 1 0 1 1 14










1 0 1

2 1 2

2 1 0

Fig. 9. A SEM micrograph from P.III microfaunal assemblage. Light digestion shown on a rodent incisor.




0 26 33 62 11

1 55 70 151 14

0 28 36 56 10

Microfauna from archaeological assemblage P.I: The P.I microfaunal assemblage is dated to the Epi-palaeolithic period. There is marked decrease in both the number of species and numbers of individuals in this fauna. Incisors are 50.4% digested which is closest to the produced by European eagle owl (48%, Andrews, 1990) compared to the tawny owl and the little owl in the same group (Table 3). Only 3.4% of incisors are heavily digested and therefore the possible presence of the little owl is unlikely in these levels. Light to moderate digestion on the murid molars (Fig. 11a), light digestion on the arvicolid molars (Fig. 11b) and also light digestion on Spalax leucodon molars (Fig. 11c) is observed on the small mammal remains.

owl (Table 3). These predators produce digestion both on the enamel and the dentine of the incisors, ranging from light to heavy, sometimes extreme digestion produces wavy surfaces on the dentine with enamel only remaining as small islands on the teeth (Fig. 10aed). Isolated molars are more abundant than in situ molars, which is a pattern only occurs in the prey assemblages of European eagle owl and tawny owl (Andrews, 1990). Table 7 Taphonomic modifications of small mammal jaw and dental elements recovered from three archaeological units PIII, PII, and PI in Chamber B. P.III

Anatomical element Mandible Maxilla Incisor Molar NISP MNI Number of small mammals Dentition % digestion M in situ % isolated M digestion % digestion I in situ % isolated I in situ Total M Light Moderate Heavy Extreme Total digestion Total I Light Moderate Heavy Extreme Total digestion









32 11 102 152 297 26 62

61,6 21,2 98,1 48,7

68 45 379 685 1177 95 151

35,8 23,7 99,7 60,1

21 12 110 151 294 28 56

37,5 41,5 98,2 44,9

5 23 2 47 214 26 1 1 0 28 114 35 8 6 0 49

8,1 15,1 16,7 46,1

11 110 5 173 842 102 16 4 0 122 404 115 46 15 2 178

7 16,1 20 45,6

15 35 1 57 205 39 7 4 0 50 115 38 16 2 2 58

27,8 23,1 20 51,8

92,9 3,6 3,6 0 13,1 71,4 16,3 12,2 0 43

83,6 13,1 3,3 0 14,5 64,6 25,8 8,4 1,1 44,1

78 14 8 0 24,4 65,5 27,6 3,4 3,4 50,4

5. Discussion There is strong evidence from all levels in the Karain sequence for the action of predators in the accumulation of the small mammal faunas. All these faunal assemblages show evidence of digestion of teeth, both molars and incisors, and because of the high level of post-depositional breakage, assessing the degrees and frequencies of digestion provides the best method of identifying the types of predators involved. From this we will attempt to assess the degrees of bias present in the small mammal faunas so as to arrive at a balanced reconstruction of the environment indicated by the faunas. Microfauna from archaeological assemblage E: The digestion patterns and frequencies indicate category 3 predators such as European eagle owl (Bubo bubo) and tawny owl (Strix aluco) which are both opportunistic predators. The percentage of digested molars is similar to the levels of digestion produced by tawny owls today in its prey assemblages. The prey assemblages of this owl include many species of rodents, birds and insectivores, with an upper size limit for prey of 150 g, and they are characterized by high levels of species equitability (Andrews, 1990). Microfauna from archaeological assemblage F: Murids are the major prey in this fauna, which could indicate favourable environmental conditions for murids, but it could also be due to their accumulation by a predator which prefers murids. Total digestion percentage is close to the produced by European eagle owl and tawny owl predators and the digestion pattern is also consistent with the same group (light for the murids and heavy for the arvicolids (Table 3). In this case we do not feel able to distinguish, but


A. Demirel et al. / Journal of Archaeological Science 38 (2011) 3048e3059

Fig. 10. SEM micrographs of rodent remains from P.II microfaunal assemblage. (a) Light digestion shown on a rodent incisor; (b) Moderate digestion shown on a rodent incisor; (c) Heavy digestion shown on a rodent incisor; (d) Extreme digestion shown on a rodent incisor.

note that both owls are woodland dwellers that hunt in more open areas, particularly near water (Mikkola, 1983), and the abundance of murids at this level would not, therefore, appear to be taphonomically induced. Their abundance at this level is likely to be a direct reflection of their abundance in the local environment in the Pleistocene, and to indicate well wooded conditions. Microfauna from archaeological assemblage G: This assemblage is similar to assemblage F in terms of taphonomic and taxonomic features. Having the isolated molars more digested than the in situ molars both in terms of frequency and degree is a characteristic of European eagle owl and tawny owl assemblages (Andrews, 1990). Of those predators, tawny owl produces heavier digestion on the molars of its prey and was the probable accumulator of the fauna at this level. Microfauna from archaeological assemblage I: The assemblage is made up of smaller sized animals, and larger species such as Sciurus anomalus, Cricetulus migratorius and Arvicola sapidus are no longer present in this fauna. Presence of A. otus (long-eared owl) is recorded for these levels (Otte et al., 1998: 426), and it is a selective hunter of microtines with an upper size limit for its prey similar to that of Tyto alba (barn owl) up to 100e200 g (Andrews, 1990). The greater abundance of smaller sized animals in this assemblage may be the result of selective hunting behaviour of a predator, and similarly, the absence of C. migratorius might be the result of the selective behaviour of this owl in favour of arvicolids.

Microfauna from archaeological assemblage P.III: The degrees of digestion and the size range of the prey assemblage strongly suggest the presence of the European eagle owl in this level. This owl is an opportunistic predator and takes whatever available in the environment, thus, almost equivalent MNI’s of murids and arvicolids might suggest differences in the environment at the time of accumulation. Lesser numbers of Apodemus compared to P.II and the relative increase of the arvicolid species compared to Mousterian levels of Chamber E is also another indication of the environment at this time period. Microfauna from archaeological assemblage P.II: Digestion patterns and percentages are moderately high and indicate that the European eagle owl and tawny owl are again the most likely predators of this assemblage. The upper limit for prey size of the tawny owl is about 150 g in weight (Andrews, 1990) and presence of digestion on the molars of the relatively large mammals such as S. anomalus (280e480 g, Demirsoy, 2003), and A. sapidus (70e250 g., Nowak, 1999) suggesting the European eagle owl is a more likely predator for the accumulation of small mammal remains at this level. Microfauna from archaeological assemblage P.I: Digestion patterns are again moderately high, indicating the European eagle owl or the tawny owl, and the presence of the larger sized animals in the assemblage (e.g. S. anomalus, S. leucodon and A. sapidus) supports the action of the European eagle owl rather than tawny

A. Demirel et al. / Journal of Archaeological Science 38 (2011) 3048e3059


and geological layers. According to the results of this study, the most likely predators of these prey assemblages are as follows: Chamber E (the main chamber),  Archaeological owl,  Archaeological owl,  Archaeological  Archaeological

assemblage E  European eagle owl or tawny assemblage F e European eagle owl or tawny assemblage G e tawny owl, assemblage I e long-eared owl

Chamber B (the side chamber),  Archaeological assemblage P.III e European eagle owl,  Archaeological assemblage P.II e European eagle owl,  Archaeological assemblage P.I e European eagle owl or tawny owl

Fig. 11. SEM micrographs of rodent remains from P.I microfaunal assemblage. (a) Apodemus mandible fragment with light to moderate digestion on the molars; (b) Arvicolid mandible fragment with light digestion on the molars; (c) Light digestion shown on a Spalax molar.

owl. These are both opportunistic predators, so that the species represented in this fauna should be in proportion to their actual presence. The results of this study have shown that the small mammal assemblages found in the Karain Cave sequences were prey assemblages of predators that are associated with different cultural

In Chamber E, archaeological level E is Charentien and F and G are Middle Palaeolithic Mousterian levels (Otte et al., 1998). The F and G small mammal assemblages are taxonomically and taphonomically similar. The taphonomy of small mammal assemblages F and G has shown that the main predators responsible for the bone accumulations were the European eagle owl and tawny owl. European eagle owl today is an opportunistic predator, taking a wide variety of prey up to the size of small deer, but there are two potential biases arising out of its activity patterns: nocturnal animals are taken more commonly than diurnal, because the owl hunts mainly at night; and wetter and more open parts of the owl’s hunting range may be over-represented, because it prefers hunting in these areas (Mikkola, 1983). Both of these biases are observed in eagle owls living today, and it is presumed that they were present in the Late Pleistocene as well. Tawny owl is also characterized by its diversity in prey assemblages and an opportunistic feeder, adjusting its hunting behaviour to the environment and its diet whatever is available in its hunting territory (Andrews, 1990). The same goes for the presumption that eagle owls in the past were opportunistic predators, as they are today, and if this is the case it may be concluded that these two fossil assemblages, along with assemblage E from Chamber E, are accurate representations of the animals living within their hunting ranges. The most important feature of the F and G faunas is the high number of species and almost identical species composition. The only difference between them is the disappearance of Mesocricetus in assemblage G (Table 5) and the appearance of S. anomalus which lives in wooded conditions (Demirsoy, 2003) in this assemblage suggesting transition from steppe and arid conditions to more temperate and wooded conditions during the Middle Palaeolithic. This continues the trend from levels of assemblage E, which lacks glirids and sciurids, and the presence of Arvicanthis also provides some evidence of grassland (Fernandez-Jalvo et al., 1998: 160). The other rodents represented with fewer numbers in this assemblage, such as Spalax and Meriones favour steppe conditions (Demirsoy, 2003). The arvicolids in these levels today are mostly found in meadow and grassland habitats (Demirsoy, 2003; Nowak, 1999). Assemblage F records the first appearances of Chionomys nivalis, A. sapidus, Myomimus roachi, Mesocricetus sp. and C. migratorius. Among these species Mesocricetus and Myomimus prefer steppe and arid environments today and C. migratorius prefers meadows, steppes and open woodland conditions (Demirsoy, 2003), increase in the number of Mus abbotti is also another indication of open and steppic conditions (Storch, 1988). The Apodemus species present in Karain microfaunal assemblages are A. sylvaticus, A. flavicollis and A. mystacinus, of which A. flavicollis is the most dependent on trees and is the only species which lives in dense woodland (Demirsoy,


A. Demirel et al. / Journal of Archaeological Science 38 (2011) 3048e3059

1996, 2003; Van Den Brink, 1967). The presence of A. sapidus may be evidence of water in the surrounding area (Mitchell-Jones et al., 1999), and this is also indicated by the evidence of sedimentological data and large mammals such as hippopotamus indicating presence of a lake (Yalçınkaya and Otte, 1999). It is harder to assess the environment of assemblage I at the top of the section in chamber E because it is interpreted as having been accumulated by the long-eared owl. The long-eared owl is a specialist predator primarily of mammals and of arvicolids in particular, but they are known to take murids where voles are absent (Andrews, 1990). It is likely that the small mammal assemblage is biased by this predator, which makes it speculative to reconstruct its environment but it is possible that it continues the trend towards more open steppic conditions with presence of woodlands around the cave. Taking taphonomic bias into account, there is a strong indication of environmental change throughout the chamber E sequence. The evidence from chamber E shows a trend from open steppe to more wooded conditions according to the small mammal evidence. Neanderthal remains are associated with assemblage F suggest temperate and partly steppe and arid conditions, with the presence of a lake as evidenced by the sedimentological and large mammalian data. As far as it can be determined at present, the microfauna from the uppermost unit, assemblage I, with further Neanderthal remains, also indicates open, perhaps steppe environments, but the manner in which this fauna was accumulated, and the nature of the taphonomic bias, makes this conclusion uncertain. The lowermost Chamber B fauna, assemblage P.III, is the only faunal assemblage equivalent to the Mousterian levels of Chamber E. The predator inferred to have accumulated the P.III small mammal assemblage is the European eagle owl. Lower representation of Apodemus suggests less wooded areas at that time and the presence of M. abbotti, which known for preferring open and steppe environments, is also another evidence of reduced woodland at this level. P.III records the first appearance of Cricetus which lives in arid steppe conditions (Demirsoy, 1992). Other rodent species here such as Spalax, Myomimus and Mesocricetus favour steppes and Sciurus favours woodland habitats. The relative increase in the arvicolid species, which prefer meadows, moorland, grassland and to a lesser degree woodland habitats, would seem to indicate conditions favourable by this owl, but probably not as wooded as the later Chamber E levels. The most distinctive feature of Upper Palaeolithic P.II microfaunal assemblage is the greater abundance of arvicolid species (Fig. 12) and diverse species composition. The level of digestion and the size range of this assemblage indicate that it was accumulated by a European eagle owl. The relative abundance of arvicolids, compared with murids, together with the high species richness

suggests a mixed habitat woodland habitat at this level. Similarly, along with Meriones tristinami, the sole presence of Allactaga in this assemblage is strong indication of steppe or dry conditions (Nowak, 1999) in the vicinity of Karain Cave in the Upper Palaeolithic. This conclusion is supported by the presence of the Cricetid species Mesocricetus, C. migratorius and Cricetus cricetus (MNI of 5 out of total 151) which live in steppe conditions today. It is concluded that mixed habitats were present, including both patches of woodland and more open arid environments. The post-glacial Epi-Palaeolithic P.I microfaunal assemblage was again shown to be accumulated by the European eagle owl or the tawny owl. Arvicolid species shows a further increase in relative abundance compared with murids (Fig. 12), but the sample size is the lowest of the Chamber B faunas. The higher MNI of arvicolids compared to murids in this fauna could be evidence of increasing grassland, but it has been shown that the improvement of climatic conditions around this time led to the spread of woodlands in the an, 2002). The presence of Apodemus and Anatolian Plateau (Özdog Sciurus in the P.I fauna indicates woodland, moorland and bushy environments with thick vegetation. On the other hand, Cricetulus and Mesocricetus prefer steppes. This species composition shows that the surrounding of Karain Cave was mostly thickly vegetated, but some steppe was still present during the Epi-Palaeolithic period. 6. Summary and conclusions Small mammal assemblages have great potential to understand the past environments in the archaeological sites as they are very sensitive to changes and selective regarding their habitats, but also vulnerable to any taphonomic factors which greatly effect and change their taxonomic composition. We have been able to identify the predator responsible for most of the seven units in the cave sequences, and from this we have been able to assess the degree of bias present in most of the faunas. The small mammal faunas at Karain Cave were accumulated by at least three species of owl: European eagle owl or tawny owl in most cases and possibly long-eared owl in assemblage I. The microfaunal assemblages in Chambers E and B have similar species composition, but the proportions of arvicolids and murids, is greater in Chamber B. Although the two chambers were probably interconnected during the Pleistocene (Yalçınkaya, 1987), the differences in species proportions show that the small mammal faunas accumulated independently in the two chambers during a slightly different time frame. The environments indicated in the two cave sequences are similar, consisting of partial steppe and arid conditions, along with grassland and woodland vegetation, but the higher proportions of arvicolids in Chamber B suggests that grassland was locally more abundant than woodland. The change in faunal composition up the Chamber E sequence indicates a shift from open steppic conditions during the middle Palaeolithic to more wooded conditions during the upper Palaeolithic. Acknowledgements

Fig. 12. Graphic illustration of murids/arvicolids by stratigraphic level.

The authors are grateful to Professor Harun TASKIRAN, Associate Professor Metin KARTAL, Dr. Kadriye OZCELIK and Dr. Beray KOCAKUNDAKCI, the members of the Karain Cave Excavation Project for their invaluable help in various ways and to Mr. Cengiz TAN, staff of the Middle East Technical University, SEM Facilities for the micrographs of the material which are of great value for this study. We also wish to thank to Mr. Emmanuel DESCLAEUX of Département de Préhistoire du Muséum National d’Histoire Naturelle, CNSR who carried out the taxonomic identification of the

A. Demirel et al. / Journal of Archaeological Science 38 (2011) 3048e3059

studied material. The authors are also grateful to the anonymous referees for their suggestions to improve the text. Arzu Demirel’s research was funded by ARIT (American Research Institute in Turkey), Peter Andrews by the L.S.B. Leakey Foundation for Anthropological Research. References Albrecht, G., 1988. An Upper Paleolithic sequence from Antalya in Southern Turkey, results of the 1985 cave excavations in Karain B. In: L’Homme de Néandertal, vol. 8. La Mutation, Etudes et Recherches Archéologiques de L’Universite de Liege, Liege, No: 35, pp. 23e35. Andrews, P., 1990. Owls, Caves and Fossils. Natural History Museum Publications, London. Andrews, P., Evans, E.M.N., 1983. Small mammal bone accumulations produced by mammalian carnivore. Paleobiology 9, 289e307. Andrews, P., Jenkins, E., 2000. The taphonomy of the small mammal faunas. In: Barham, L. (Ed.), The Middle Stone Age of Zambia, South Central Africa. Western Academic&Specialist Press, Bristol, pp. 57e61. Andrews, P., O’Brien, E.M., 2000. Climate, vegetation and predictable gradients in mammal species richness in southern Africa. Journal of Zoology 251, 205e231. Avery, D.M., 2002. Taphonomy of micromammals from cave deposits at Kabwe (Broken Hill) and Twin Rivers in Central Zambia. Journal of Archaeological Science 29, 537e544. Van Den Brink, F.H., 1967. A Field Guide to the Mammals of Britain and Europe. Collins, London. Brothwell, D., Jones, R., 1978. The relevance of small mammal studies to archaeology. In: Brothwell, D., Thomas, K.D., Clutton-Brock, J. (Eds.), Research Problems in Zooarchaeology. Institute of Archaeology, London, pp. 47e57. Dauphin, Y., Kowalski, C., Denys, C., 1994. Assemblage data and bone and teeth modification as an aid to paleoenvironmental interpretations of the open-air Pleistocene site of Tighenif (Algeria). Quaternary Research 42, 340e349. Dauphin, Y., Williams, C.T., Andrews, P., Denys, C., Fernandez-Jalvo, Y., 1999. Diagenetic alterations of micromammal fosil bones from Olduvai Bed-I of the Lower Pleistocene sequence at Olduvai Gorge, Tanzania. Journal of Sedimentary Research 69, 612e621. Demirsoy, A., 1992. Yas¸amın Temel Kuralları. Omurgalılar/Amniyota (Sürüngenler, Kus¸lar ve Memeliler), Cilt III/Kısım III. Meteksan Yayınları, Ankara. rafyası: Hayvan Cog rafyası. Meteksan Demirsoy, A., 1996. Genel ve Türkiye Zoocog Yayınları, Ankara. Demirsoy, A., 2003. Türkiye Omurgalıları: Türkiye Omurgalı Faunasının Sistematik ve Biyolojik Özelliklerinin Aras¸tırılması ve Koruma Önlemlerinin Saptanması, Memeliler. Meteksan Yayınları, Ankara. Denys, C., Dauphin, Y., Rzebik-Kowalska, B., Kowalski, K., 1996. Taphonomic study of Algerian owl pellet assemblages and differential preservation of some rodents: palaeontological implications. Acta Zoologica Cracoviensia 39, 103e116. Denys, C., Andrews, P., Dauphin, Y., Willams, T., Fernandez-Jalvo, Y., 1997. Towards a site classification: comparison of stratigraphic, taphonomic and diagenetic patterns and processes. Bulletin de la Societe Geologique de France 168, 751e757. Dodson, D., Wexlar, P., 1979. Taphonomic investigation of owl pellets. Paleobiology 5, 275e284. Efremov, I., 1940. Taphonomy: a new branch of paleontology. Pan-American Geologist 74, 81e93. Fernandez-Jalvo, Y., Andrews, P., 1992. Small mammal taphonomy of Gran Dolina, Atapuerco (Burgos), Spain. Journal of Archaeological Science 19, 407e428.


Fernandez-Jalvo, Y., Denys, C., Andrews, P., Williams, T., Dauphin, Y., Humphrey, L., 1998. Taphonomy and palaeoecology of Olduvai Bed-I (Pleistocene, Tanzania). Journal of Human Evolution 34, 137e172. rafya Dergisi Kökten, I., 1964. Karain’in Türk Prehistoryasındaki yeri. Türk Cog 18e19, 22e23. Korth, W., 1979. Taphonomy of microvertebrate fossil assemblages. Annals of Carnegie Museum 48, 235e285. Kowalski, K., 1990. Some problems of the taphonomy of small mammals. International symposium on the evolution and biostratigraphy of the arvicolids, pp. 285e296. Pinto Llona, A.C., Andrews, P., 1999. Amphibian taphonomy and its application to the fossil record of Dolina (Middle Pleistocene, Atapuerco, Spain). Palaeogeography, Palaeoclimatology, Palaeoecology 149, 411e429. Lyman, L.R., 1994. Vertebrate Taphonomy. Cambridge Manuals in Archaeology. Cambridge University Press. Matthews, T., 1999. Taphonomy and the micromammals from Elands Bay Cave. South African Archaeological Bulletin 54, 133e140. Mikkola, H., 1983. Owls of Europe. Poyser Natural History. Academic Press, London. Mitchell-Jones, A.J., Amori, G., Bogdanowicz, W., Krystufek, B., Reijnders, P.H.J., Spitzenberger, F., Stubbe, M., Thissen, J.B.M., Vohralik, V., Zima, J., 1999. The Atlas of European Mammals. Poyser Natural History. Academic Press, London. Nowak, R.M., 1999. Walker’s Mammals of the World, 6th edn., vol. II. The Johns Hopkins University Press, Baltimore and London. Otte, M., Yalçınkaya, I., Tas¸kıran, H., Kozlowsky, J., Bar-Yosef, O., Noiret, P., 1995. The Anatolian Middle Paleolithic: new research at Karain Cave. Journal of Anthropological Research 51, 287e299. Otte, M., Yalçınkaya, I., Kozlowsky, J., Bar-Yosef, O., Lopez, Bayon I., Tas¸kıran, H., 1998. Long-term technical evolution and human remains in the Anatolian Palaeolithic. Journal of Human Evolution 34, 413e431. _ Adımlar: Paleolitik Çag . Arkeo Atlas 1, 46e57. DBR Yayıncılık, an, M., 2002. Ilk Özdog _ Istanbul. arası B Gözü’nde Üst Paleolitik Sorunu. Anadolu/ Özçelik, K., 2003. Karain Mag Anatolia 25, 83e95. Redding, R.W., 1978. Rodents and the archaeological palaeoenvironment: considerations, problems and the future. In: Meadow, R.H., Zeder, M.A. (Eds.), Approaches to Faunal Analysis in the Middle East. Peabody Museum Bulletin 2, Cambridge, pp. 64e68. Storch, G. Von, 1988. Eine jungpleistozäne/altholozäne Nager-Abfolge Von Anatolya, SW-Anatolien (Mammalia-Rodentia). Z. Säugertierkunde 53, 76e82. _ arası. Arkeo Atlas 1, 50. DBR Yayıncılık, Istanbul. Tas¸kıran, H., 2002. Karain Mag Yalçınkaya, I., 1987. 1985 yılı Karain Kazıları. 8. Kazı Sonuçları Toplantısı I, T.C. Kültür ı. Eski Eserler ve Müzeler Genel Müdürlüg ü, Ankara, pp. ve Turizm Bakanlıg 21e37. Yalçınkaya, I., 1988. 1986 yılı Karain Kazıları. 9. Kazı Sonuçları Toplantısı I, T.C. Kültür ü, Ankara, pp. ı. Eski Eserler ve Müzeler Genel Müdürlüg ve Turizm Bakanlıg 15e37. Yalçınkaya, I., Otte, M., 1999. 1997 Yılı Karain Kazıları. 20. Kazı Sonuçları Toplantısı I, ı. Anıtlar ve Müzeler Genel Müdürlüg ü, Ankara, pp. 23e37. T.C. Kültür Bakanlıg Yalçınkaya, I., Otte, M., Bar-Yosef, O., Kozlowski, J., Léotard, J.M., Tas¸kıran, H., 1993. The excavations at Karain Cave, Southwestern Turkey: an interim report. In: Olszewski, D.I., Dibble, H.L. (Eds.), The Paleolithic Prehistory of the ZagrosTaurus. University Museum Monograph 83, University Museum Symposium Series Volume V. University Museum of Pennsylvania, pp. 101e117. Yalçınkaya, I., Tas¸kıran, H., Atıcı, A.L., Kösem, M.B., Özçelik, K., Kartal, M., Erek, C.M., 2001. 1999 yılı Karain Kazıları. 22. Kazı Sonuçları Toplantısı I, T.C. Kültür ı. Anıtlar ve Müzeler Genel Müdürlüg ü, Ankara, pp. 9e20. Bakanlıg