Holocene environment and subsistence patterns from Capsian and Neolithic sites in Tunisia

Quaternary International xxx (2013) 1e12

Contents lists available at ScienceDirect

Quaternary International journal homepage: www.elsevier.com/locate/quaint

Holocene environment and subsistence patterns from Capsian and Neolithic sites in Tunisia Nabiha Aouadi a, *, Yosra Dridi b, Wafa Ben Dhia c a

Centre National de Recherches Préhistoriques, Kairouan 3100, Tunisia LAMPEA-UMR 6636, Aix-en-Provence, France c Institut des Langues Appliquées, Moknine 5050, Tunisia b

a r t i c l e i n f o

a b s t r a c t

Article history: Available online xxx

Faunal analysis was carried out on four Holocene Tunisian prehistoric sites from both Capsian and Neolithic cultures. Capsian hunting strategies from the examination of prey selection and carcass exploitation were investigated from two sites: Bir Hmairiya and SHM-1. Capsian populations showed great adaptability to their physical environment. At SHM-1 and Bir Hmairiya, prehistoric populations developed subsistence strategies focused on hunting of a wide range of local wild prey with more exploitation of medium and small mammals, especially hartebeest and gazelle with additional exploitation of food resources from other terrestrial mammals and land snails. Cultural and economic patterns indicate that these hunters-gatherers populations were using long-term occupation sites, a mode of sedentary lifestyle during the Capsian. The Neolithic economy is documented through archaeozoological studies of faunal remains from two sites: Kef el Agab and Doukanet el Khoutifa. Neolithic populations kept the same lifestyle as the Capsian tradition but with gradual investment in meat, milk, and wool production. Comparing Kef el Agab and Doukanet el Khoutifa, patterns of mammal exploitation do not display great differences. The occupants practiced hunting and gathered land snails. The main change lies in the increase in Ovicaprid remains at Doukanet el Khoutifa, which would suggest a greater control of breeding activity. Pastoralism was already established at Doukanet el Khoutifa by the first half of the 7th millennium cal BP. This pastoralism involved more sheep and goat than bovines, but without dog or pig. Ó 2013 Elsevier Ltd and INQUA. All rights reserved.

1. Introduction

humans groups was gathering land snails, hunting wild animals, and picking fruits. Taphonomic and archaeozoological research are required to recognize the post-depositional processes that affected faunal remains. We first present taphonomic and archaeozoological aspects of the four sites. We then consider subsistence patterns of Capsian and Neolithic groups and discuss the palaeoenvironmental data. Bone and tooth remains were determined to the most specific level possible. We used ungulate size classes (modified from Brain, 1981) for unidentified faunal remains. We studied the faunal remains from different kinds of cultural sites in order to gain further insight into the economy of these societies. The earliest stages of animal domestication in North Africa and especially in Tunisia are of great interest in particular when we compare them to human behavioral patterns during the Capsian, just prior to domestication. During the last two decades, archaeozoological investigations in the Maghreb have tried to understand the economic strategies of Capsian populations; summaries of research into the subsistence strategies of Capsian and Neolithic populations is provided elsewhere (Merzoug, 2012; Ouchaou, 2012).

In this paper, we present the results of detailed archaeozoological analyses of all excavated faunal remains from four prehistoric rammadiyet in Tunisia. A rammadiya is a specific name given to an enormous accumulation of ash, land snail shells, burned stones, lithic artifacts and faunal remains (Gobert, 1937). Two sites belong to the Capsian (Camps, 1974; Rahmani and Lubell, 2012), Bir Hmairiya, attributed to Typical Capsian and the Upper Capsian site SHM-1, while Kef el Agab and Doukanet el Khoutifa belong to the Neolithic. In eastern Maghreb, Capsian culture is one of the most studied Epipalaeolithic culture and is inherent to Tunisia and eastern Algeria. Capsian is subdivided in two phases: Typical Capsian (ca 10,000e8000 cal BP) and Upper Capsian (ca 9000e 7500 cal BP). The Capsian groups are considered as the last huntersgatherers in the Maghreb. The mode of subsistence of those

* Corresponding author. INP, Prehistory, sidi abid el ghiriani, Kairouan, Tunisia. E-mail address: [email protected] (N. Aouadi). 1040-6182/$ e see front matter Ó 2013 Elsevier Ltd and INQUA. All rights reserved. http://dx.doi.org/10.1016/j.quaint.2013.07.028

Please cite this article in press as: Aouadi, N., et al., Holocene environment and subsistence patterns from Capsian and Neolithic sites in Tunisia, Quaternary International (2013), http://dx.doi.org/10.1016/j.quaint.2013.07.028

2

N. Aouadi et al. / Quaternary International xxx (2013) 1e12

In the present study, detailed analysis of the faunal remains from SHM-1 permits new insights into the way the Capsian inhabitants of the site exploited animals. The analyses from all four samples of Holocene ungulate assemblages presented here is mostly aimed at the reconstruction of animal exploitation and site function along with the deduction of the paleoclimatic patterns that affected Capsian and Neolithic societies. 2. Site locations and cultural context Bir Hmairiya is located in southern Tunisia in the mineral basin of Gafsa, next to the Metlaoui area which is now quite dry (Fig. 1). This rammadiya is composed of two open-air loci: one Typical Capsian and the other Upper Capsian (Marty, 1966; Ben-Dhia, 2002). Teste excavated these in 1936 and 1949 (Marty,1966), but unfortunately there is no documentation of his excavation procedure. Two arbitrary loci were excavated, and deposits at both were composed of ash, burned stones, land snails and very scarce mammal remains. Today, the site is in a very badly preserved state, perturbed and attacked by weathering, trampling and erosion. For this study, we analyzed the Typical Capsian faunal assemblage recovered from sector SE2. The open-air site of SHM-1 is situated on the eastern Tunisian coast, north of Sousse between two physiographic units: the northwestern highlands, and the eastern and southern Mediterranean coast. The rammadiya was first explored by Harbi-Riahi and Zoughlami (1971). Between 2002 and 2007, the site was

excavated by an ItalianeTunisian research program (Mulazzani, 2013). Seven main Upper Capsian occupational layers have been detected; all radiocarbon dated between the 9th (7880
30 BP, 8587e8968 cal BP) and the second half of the 8th millennium (6840
30 BP, 7136e7478 cal BP; Saliège et al., 2013). The Kef el Agab site consists of a huge cave and a sub-circular slope in front, both of which compose the rammadiyat. The site is located in northern Tunisia near Jendouba and was first excavated by Bardin in 1947e1948 (Bardin, 1953). In November 2010, a new trench was opened on the slope. A radiocarbon date, obtained on a human bone fragment from Bardin’s excavation, is 5984
24 BP (6747e6887 cal BP) (M. Mannino 2010, personal communication). The stratigraphic provenance of the sample is unknown, as the excavation was not conducted following the stratigraphy of the sediment. Finally, the site of Doukanet el Khoutifa in the Siliana region, is composed of a shelter with a vast terrace of 2200 m2 and was excavated by Zoughlami between 1973 and 1976 in Sectors 1 and 2 (Zoughlami, 2009). Two main levels are distinguished from bottom to top: a Yellow level (thickness 0.4 m) and a Dark one (thickness 1.6 m), both belonging to the Neolithic and dating between the 8th (6750
200 BP, 7266e7970 cal BP) and the first half of the 7th millennium cal BP (6000
100 BP, 6638e 7158 cal BP) (Zoughlami, 2009). We analyzed faunal remains from both levels. Radiocarbon dates for the four sites studied are given in Table 1.

Table 1 Selected radiocarbon dates of Holocene sites from Tunisia (OES: Ostrich egg-shell; CG: Cerastoderma glaucum; Ch: Charcoal; H: Human bone). All calibrations use CALIB 6.1.0 and Intcal09.14c for terrestrial samples or Marine09.14c for marine samples taking in to account a DR correction factor for Mediterranean waters of 58
85 (Reimer and McCormac, 2002; Reimer et al., 2009). Site

Culture

Layer

Sample

Lab. code

Age BP

Cal BP 2s

Ref.

SHM-1

Upper Capsian

Kef el Agab

Neolithic

1 7 Unknown

OES CG H

SacA 23655 ENEA-866 Hd-28982

7880
35 6840
30 5984
24

8587e8968 7136e7478 6747e6887

Doukanet el Khoutifa

Neolithic

Yellow Dark

Ch Ch

MC 828 MC 835

6750
200 6000
100

7266e7970 6638e7158

Saliège et al., 2013 Saliège et al., 2013 M. A. Mannino, pers. comm. Zoughlami 2009 Zoughlami 2009

3. Methods of analyses

Fig. 1. Map of the studied Capsian and Neolithic sites. 1: Bir Hmairiya, 2: SHM-1, 3: Doukanet el Khoutifa, 4: Kef el Agab.

This study included detailed taphonomic and archaeozoological analyses of faunal remains from all four sites. We first cleaned and marked all specimens and listed them in an Excel database. Sorting and identification followed the methodology of Klein and CruzUribe (1984) and Barone (1986). Taxonomic identifications of faunal remains were done first by direct observations and comparisons with fossil and modern specimens from Tunisia (stored at the Archaeozoological Laboratory in Kairouan) and more generally from Africa by consulting the literature (Gabler, 1985; Peters, 1986; Peters et al., 1997; Jousse, 2003). In rare cases, we could determine sex for some remains (on pelvis, size and shape of horn, differences in bone size). To age remains, we used species-specific eruption sequences and wear patterns on teeth. We used mortality profiles of ungulates (wild and domestic) to understand the origin of vertebrate remains and thus human behavioral patterns. Bones were assigned to one of three age classes: young, prime, and old. Stiner (1990) indicates that recent humans typically focus on prime-age adults when they select prey. Faunal analysis was carried out following standard taphonomic and archeozoological methods (Reitz and Wing, 1999). All bones were subjected to analyses of weathering features (macroscopic

Please cite this article in press as: Aouadi, N., et al., Holocene environment and subsistence patterns from Capsian and Neolithic sites in Tunisia, Quaternary International (2013), http://dx.doi.org/10.1016/j.quaint.2013.07.028

N. Aouadi et al. / Quaternary International xxx (2013) 1e12

under strong direct light and microscopic examination of the surface traces), breakage patterns (with study of percussion notches and pits), tooth marks of carnivores and of burning (Blumenschine et al., 1996). Taphonomic analysis included study of the effect hydrological transport may have had on the bone samples. Skeletal elements were arranged in hydrological transportability groups (Voorhies, 1969). We also analyzed the external physical aspects of recovered remains (biochemical and abrasion marks). Several indices were calculated in order to evaluate bone fragmentation (Table 4). We choose to use those proposed by Brugal and Patou-Mathis (1993) for cranial and post-cranial remains. The general skeletal preservation index or IGCS is the ratio between the cranial NISP and the post-cranial NISP. For a complete bovid skeleton, the assumed ratio is 0.27. The bone breakage intensity index, or IFO, is the ratio between the number of complete bones and the number of broken bones for each taxon. This index shows the amount of the destruction for different levels. The dental preservation index, or ICD, is the ratio between isolated teeth in NISP and post-cranial NISP. The ICD evaluates dental preservation for each species relative to a complete skeleton. Besides, we used the Completeness Index (CI) of Marean (1991) which evaluates the degree to which postdepositional processes have affected our assemblages by calculating damaged tarsals and carpals after removing those affected by human or carnivore attacks. In order to pinpoint more details about bone collectors, we analyzed bone surface modification mainly by carnivore tooth marks (mechanical and chemical attacks, Fosse et al., 1998), or butchering traces (marrow extraction, evisceration, skinning, use of bones for tools). We followed Villa and Mahieu (1991) in examining patterns of breakage and their interpretation (green vs. dry bone breakage) by describing the bone outline (transverse, V-shaped, longitudinal) and the fracture angle (oblique, right, mixed). We followed Brain (1980) in studying depth and width of gnaw marks on bones. Cut mark incisions have a V-shaped section and internal micro-striation (Potts and Shipman, 1981). In addition, we recorded the position of cut marks along the identifiable bone. Shaft and epiphyseal fragments were carefully studied. Then, depending on the preserved original circumference of the bone, we classified and measured (length, width, thickness) on the unidentifiable remains and entered the data into another database (subdivisions are long bone fragments, metaphyseal fragments and epiphyseal fragments; see Bunn, 1983 for details). We also recorded side and state of fusion when recognizable. Current archaeozoological models illustrate a wide range of faunal interpretations. For example, studying mammalian skeletal element abundance allows elucidation of butchery and transport practices. We followed the framework of Behrensmeyer (1978, 1991) and Blumenschine and Selvaggio (1988) to reconstitute the history of vertebrate accumulation (sedimentary context, breakage patterns, shafts of flakes, taxonomic spectra, skeletal representation). We first analyzed skeletal part representation for each species using parameters from Grayson and Frey (2004) such as NISP (number of identified specimens), MNI (minimum of number of individuals). Each body part (skull, axial, appendicular) are represented in NISP. In analyzing butchery activities (cut marks), we preferred to employ minimal animal units (MAU). Ungulate exploitation patterns were considered within the global established context. Variation in the mineral composition of each skeletal element influences its survivorship (Lyman, 1994; Lam et al., 1999). Skeletal element abundance in assemblages is monitored either by carcass treatment or by bone preservation. In order to elucidate the effect of pre-depositional and postdepositional processes on bone preservation, we analyzed the

3

relationship between bone density and bone survivorship. Here, we used bone mineral density values for wildebeest (Lam et al., 1999). If bones were not affected by selective destruction (e.g. density), we then studied the possibility of selective transport of body parts. In fact, ethnographic research (Binford, 1978; O’Connell et al., 1988a, 1988b, 1990, 1992) showed that the nutritive value of body parts established their transport and abundance on archaeological sites. Binford (1978) was the first to derive utility indices for anatomical units of animals (sheep and caribou). He used meat, grease, and marrow value of each element and calculated the modified general utility index (MGUI). We preferred in some cases to employ the unstandardized food utility index of Metcalfe and Jones (1988) because they include the whole economic bone utility in their calculations. We tested the relationship between bone survivorship (based on the minimum number of animal units expressed in MAU) and the food utility index (FUI) (dealing with the weight of useable tissue from current animal body parts). Bone utility index and skeletal element abundance showed a positive relationship in hunting context. In a scavenging case, a reverse relationship is demonstrated between both variables. Therefore, we investigated any relationships between the recovered skeletal representation in the assemblage (expressed in %MAU or %NNISP) and different parameters (density, FUI, hydrological transportability). We used Spearman’s rank eorder correlation coefficient (a ¼ 0.05) to test all hypothesis. All statistical tests were performed using PAST (Hammer et al., 2001). In summary, all methods used in this study were aimed at providing and analyzing information about site formation, faunal accumulators, and human behavior. 4. Material Bir Hamiriya and Doukanet el Khoutifa were only excavated during the last century but SHM-1 and Kef el Agab were recently reexcavated and produced important samples of artifacts and faunal remains, the latter of which is the subject of the research reported here. 4.1. Bir Hmairiya Bone material comes from the earlier excavations of Teste, and unfortunately no stratigraphic information is available about provenance. The faunal remains were probably selected during the excavation, as we found only teeth and complete bones, and all bone fragments are missing. Only the bone collection from SE2 has been found, and it is Typical Capsian according to Marty (1966). Lacking the description of the stratigraphic sequences, we treated the whole faunal assemblage as a single aggregate unit. 4.2. SHM-1 The first excavation at SHM-1 was conducted between 1969 and 1971 by Riahi and Zoughlami. A field season of exhaustive and methodical excavation was conducted by Mulazzani and Boussofara between 2002 and 2007, covering an area estimated to be 94 m2 using a grid system of 1 m2 and different levels with each object mapped by theodolite with reference to a datum point. All excavated deposits were sifted (Mulazzani, 2010; Belhouchet et al., 2013). Seven levels were grouped into two phases: phase 1 (from level 1 until level 4) and phase 2 (from level 5 until level 7). The vertical distribution of the faunal remains appears continuous, although there are breaks between different levels. We treated the whole assemblage as two units. We analyzed both series: from the earlier excavations of Riahi and Zoughlami and the recent stratigraphic and planigraphic

Please cite this article in press as: Aouadi, N., et al., Holocene environment and subsistence patterns from Capsian and Neolithic sites in Tunisia, Quaternary International (2013), http://dx.doi.org/10.1016/j.quaint.2013.07.028

4

N. Aouadi et al. / Quaternary International xxx (2013) 1e12

trenches opened by the TunisianeItalian team. This second collection has been analyzed following the diachronic sequence (Aouadi et al., 2013). The faunal assemblage showed highly fragmented bone. We focused our taxonomic identifications mainly on isolated teeth. 4.3. Kef el Agab In 2010, Aouadi and Belhouchet carried out a 2 m2 excavation to reconstruct the stratigraphic sequence and apply new chronometric dating methods to the site. This produced a large number of samples. We also analyzed the Kef el Agab bone series from Bardin’s (1953) excavations and from our 2010 tests. 4.4. Doukanet el Khoutifa From Doukanet el Khoutifa, we studied the faunal assemblage from the excavations of Zoughlami (1973e1976) which recovered important material including human remains, stone and bone tools and a large sample of vertebrate faunal remains that were not well described. 5. Results 5.1. Faunal assemblages Wild mammal species composed the faunal spectra from Bir Hmairiya. This sample yielded an NSP of 123 (we counted identifiable and unidentifiable elements, Table 2), of which 84.5% could be attributed to taxon and part of skeleton. There are mainly remains of bovids (Aurochs, Barbary sheep, Alcelaphini and Antilopini), with some rhinocerotidae and equids. Carnivores, leporids and birds are very scarce. The faunal spectra are dominated by Barbary sheep (Ammotragus lervia) and hartebeest (Alcelaphus buselaphus) (Table 3, Fig. 2A).

Table 2 Percentages of determined bones (NISP: number of identified specimens, NUSP: number of unidentified specimens, NSP: number of specimens). Parameters

Bir Hmairiya

SHM-1

Kef el Agab

Doukanet el Khoutifa

Total NISP Total NUSP Total NSP % determined

104 19 123 84.5

261 2914 3175 8.2

378 331 709 53.3

1196 4506 5702 21.0

From SHM-1, we counted an NSP of 3175 (Table 2), of which only 8.2% were attributable to taxon. Bovids are the most frequently represented taxa. Hartebeest and gazelle are the main species, followed by aurochs and leporids. Table 3 includes a diverse range of mammals mainly from medium (e.g. hartebeest), large (e.g. aurochs) and small (e.g. gazelle) bovids (Table 3, Fig. 2B). Our analyses of mammalian skeletal elements document very poor bone survival. Skeletal part representation reveals a high frequency of cranial elements and low representation of vertebrae, ribs, and complete limbs. The faunal assemblage from SHM-1 is richer in number and species in phase 2 (more recent, formed by levels 5, 6, 7) than phase 1 (levels 1, 2, 3, 4). However, small mammals and reptiles are more common in phase 1, which may reflect palaeoclimatic changes. From Kef el Agab, we counted an NSP of 709, of which more than half (53.5%) were identifiable (Table 2). Overall, remains essentially come from bovids (aurochs, buffalo, gazelle, hartebeest and Barbary sheep) with some carnivores (Table 3). Remains of barbel from nearby Wadi Medjerda also indicate fishing activities. Caprini is the largest group of domestic fauna (Table 3, Fig. 2C). In test S1, we found faunal remains belonging mainly to wild bovids (gazelle and aurochs). Small quantities of domestic species are present, especially Capra/Ovis taxa. In terms of individuals, there are eight aurochs, three water buffalo, many hartebeest (n ¼ 11) and gazelle (n ¼ 15), three Bohor reedbucks, three Barbary sheep and five wild boars. Carnivores are represented by two golden jackals, one hyena, and four felids. Domestic Caprini are abundant with 23 individuals (Table 3). The faunal assemblage from Doukanet el Khoutifa is composed of mixed taxa from wild and domestic species. We analyzed 5702 elements, of which only 21% were identified (Table 2). Remains of domestic Caprini (Ovis and Capra) and of wild taxa (Bovini, Alcelaphini and Antilopini) are the main taxa in the assemblage (Table 3, Fig. 2D). Sheep and goat remains dominate the sample. Nevertheless, faunal remains from the Yellow level are mainly composed of aurochs and Dorcas gazelle. Bovids, suids, equids, leporids and birds are present in both levels but are more frequent in the Dark one. However, carnivores (canids, felids) and hedgehog are present only in the Dark level. The skeletal representation of the ovicaprids is characterized by a high frequency of tarsals and phalanges, followed by loose teeth and then by forelimb elements (carpals and metacarpals). There is more intense bone fragmentation and tooth dispersion in the Dark level, whereas in the Yellow level, cranial elements are more often complete and isolated teeth are less abundant. Vertebrae, ribs, pelvis, femora, and tarsals are more abundant in the Dark level. The bone assemblages from both levels have low frequencies of carnivore tooth marks. Element distribution of Barbary sheep (Fig. 3) shows that all skeletal

Table 3 List and abundance (in NISP and MNI) of faunal remains from Holocene sites in Tunisia (þ: present). Taxa

Bir Hmairiya NISP

Bos primigenius, aurochs aff. Bos primigenius Bos sp. Syncerus caffer, African buffalo Bos taurus, domestic cattle Alcelaphus buselaphus, red hartebeest aff. Alcelaphus buselaphus Damaliscus sp., topi and their allies Gazella dorcas, dorcas gazelle Gazella cuvieri, mountains gazelle

SHM-1 MNI

NISP

Kef el Agab MNI

NISP

12 2 2 2

6 1 1 1

4

1

5

1

29

3

18 13

9 8

9 3

1 2

4 2

3 2

Doukanet el Khoutifa MNI

NISP

MNI

40

8

43

7

5 4

3 3

43 2 5 11 2

9 2 2 3 2

26 33 5

3 5

8

3

Please cite this article in press as: Aouadi, N., et al., Holocene environment and subsistence patterns from Capsian and Neolithic sites in Tunisia, Quaternary International (2013), http://dx.doi.org/10.1016/j.quaint.2013.07.028

N. Aouadi et al. / Quaternary International xxx (2013) 1e12

5

Table 3 (continued ) Taxa

Bir Hmairiya NISP

Gazella ruffifrons, red-fronted gazelle Gazella sp., gazelle Kobus kob, waterbuck Redunca redunca, bohor reedbuck Ammotragus lervia, barbary sheep Capra/Ovis, goat/sheep Capra hircus, goat Ovis aries, sheep Hippotragini, hippotragine Sus scrofa, wild boar aff. Sus scrofa aff. Giraffa camelopardalis, giraffe Ceratotherium sp., rhinoceros Equus sp., equids Panthera leo, lion Crocuta crocuta, spotted hyaena Canis aureus, common jackal Caracal caracal, caracal Felis libyca, African wildcat Felis sp. Vulpes vulpes, fox Lepus sp., hare Oryctolagus cuniculus, European rabbit Atelerix algirus, North African hedgehog Rodent Hystrix cristata, North African crested porcupine Birds Fishes Tortoise Snake Amphibian Shellfish Land snail Sea shell

SHM-1 MNI

16

2

26

4

1

1

5 3

NISP

1 2

1 1

1 1

1

2 5 12 8 0 0

10 24 19 18 2 1

0.2 0.2 0.63 0.44 * *

0.2 0.3 0.24 0.18 * *

0.2 0.27 0.38 0.18 * *

44 33 24 3

6 9 3 2

38 24 21 1

0.15 0.37 0.14 2

0.9 0.45 0.41 *

0.86 0.45 0.23 *

MNI 1 5 1 1 4 4 5 11

49

10

3

3

6 17 42 9 21

3 3 10 4 9

12

5

10

4

1 1 3

1 1 3 11 2

4 1

1

1

1 1 9 3

1 1 4 1

21 1

7 1

5

5

14 5

8 5

9

1

Table 4 Fragmentation indices for mammal remains of the four sites. IFO: bone breakage intensity index. IGCS: general skeletal preservation index. ICD: dental preservation index. P NISP Complete Fragment IFO IGCS ICD NISP NISP 12 29 31 26 2 1

NISP

7

elements are present in the sample, although the abundance of each element is not representative of anatomical completeness: there is a much lower frequency of axial and forefoot in comparison to forequarter and hindfoot elements. The global skeletal conservation index (IGCS) of mammals at Kef el Agab showed more conservation of cranial remains than post-cranial but also that a selection of cranial remains was made by the Bardin team (Table 4). From the Dark level, the IGCS of Caprini is 0.33 and for the Yellow level it is 0.6, showing more Caprini cranial remains in the Yellow than in the Dark. Both levels have more cranial remains than postcranial ones.

Bir Hmairiya Bovini Alcelaphini Antilopini Caprini Carnivores Leporids SHM-1 Bovini Alcelaphini Antilopini Reduncini

Doukanet el Khoutifa MNI

18

13 205 52 2 6 4 þ þ

þ

NISP

1 30 2 2 19 123 50 99

9

1

Kef el Agab MNI

2 2

1 2

2 3

2 2

2

2

1

1

9 9

4

3 3

3 3

2

2

22

þ

Table 4 (continued) P NISP

þ

Complete NISP

Carnivores 32 16 Leporids 49 19 Kef el Agab Bovini 54 37 Alcelaphini 60 37 Antilopini 62 33 Reduncini 6 5 Caprini 89 66 Carnivores 9 5 Doukanet el Khoutifa (Yellow level) Bovini 4 1 Alcelaphini 2 1 Antilopini 15 8 Caprini 12 7 Leporids 4 2 Doukanet el Khoutifa (Dark level) Bovini 97 45 Alcelaphini 36 8 Antilopini 46 26 Reduncini 4 1 Caprini 421 186 Carnivores 9 6 Leporids 20 5

Fragment NISP

IFO

IGCS

ICD

16 30

1 0.63

0.55 0.02

0.5 0.02

17 23 29 1 23 4

2.17 1.6 1.13 5 2.86 1.25

0.58 1.7 0.47 1 1.6 1.75

0.58 1.6 0.16 0.6 1.4 0.25

3 1 7 5 2

0.33 1 1.14 1.4 1

0 0.3 0.23 0.6 0.25

0 0.3 0.23 0.6 0.25

52 28 20 3 235 3 15

0.86 0.28 1.3 0.33 0.79 2 0.33

0.8 0.16 1 0 0.33 0.5 0

0.78 0.16 0.95 0 0.31 0.5 0

The IFO index of bovids is different in the four sites: Bir Hmairiya is 0.2e0.63, SHM-1 is 0.15e2, Kef el Agab is between 1.13 and 5, the Yellow level of Doukanet el Khoutifa is between 0.33 and 1.4, differing in the Dark level where it is between 0.28 and 1.3. Kef el

Please cite this article in press as: Aouadi, N., et al., Holocene environment and subsistence patterns from Capsian and Neolithic sites in Tunisia, Quaternary International (2013), http://dx.doi.org/10.1016/j.quaint.2013.07.028

6

N. Aouadi et al. / Quaternary International xxx (2013) 1e12

Agab has the most complete bones while SHM-1 and the Dark level of Doukanet el Khoutifa have the lowest frequency of complete bones, and thus the highest frequencies of broken bone. The ICD for bovids at Bir Hmairiya is between 0.18 and 0.38, for SHM-1 it is between 0.23 and 0.86, at Kef el Agab it is between 0.16 and 1.6, the Yellow level at Doukanet el Khoutifa is between 0.23 and 0.6 and for the Dark level between 0.16 and 0.95. At Kef el Agab there is a greater frequency of isolated teeth than in the other sites, with a high degree of fragmentation of mandibles and dispersion of isolated teeth and especially good preservation of Alcelaphini teeth. Bovini teeth from SHM-1 are the best preserved of all sites. On the other hand, teeth of Antilopini from the Dark level of Doukanet el Khoutifa are the best conserved. In sum, differential abundance of parts of bones is likely due to in situ destruction and biological agents. Taphonomic analysis from SHM-1 shows a deficiency of complete bones caused by destruction during the occupation of the site. The CI index of bovid carpals and tarsals at SHM-1 is between 75 and 100% indicating that it is very unlikely post-depositional processes are the only factors responsible for the missing elements. In conclusion, the indices of bone fragmentation (Table 4) indicate that differential preservation, human strategies, and carnivore gnawing altered all faunal assemblages. Nevertheless, carnivore activity seems to have been very limited, especially at SHM-1. 5.2. Palaeoenvironment

Fig. 2. Faunal abundance (in %NISP). A: Bir Hmairiya, B: SHM-1, C: Kef el Agab, D: Doukanet el Khoutifa.

The distribution of faunal group association into four environmental types (Fig. 4) shows the preponderance of an open, but not desert, environment (Type C) in different proportions at all four sites. From the environment detected by the faunal spectra, SHM-1 was in the most open one, while Doukanet el Khoutifa was the least. This fits with the topography of the surrounding area. We introduced cenogram construction in this study in order to infer palaeohabitat from analysis of the structure of extinct mammalian groups. The correct use of the cenogram method permits inference of habitat type for fossil communities and effects caused by diverse pressures on the mammalian community (Travouillon and Legendre, 2009). Cenogram plots (Fig. 5) were used to deduce insights into palaeoclimatic patterns using ungulate body mass distribution and study the correlation between habitat and mammalian structure (Travouillon and Legendre, 2009). Valverde (1964) was the first to use the term cenogram for a graph displaying the relationships between the size of predators and the size of their prey species in a mammalian community. His cenograms were constructed by plotting rank ordered taxa versus head-body length. Legendre (1986, 1989) adapted this method for palaeoecological studies. Body sizes were expressed in log of body mass (g). Legendre (1989) established visual models compiling faunal community and environment. Current mammalian communities provide models for establishing the general environmental characteristics of fossil mammalian faunas. We constructed cenogram graphs for the mammalian communities from our sites (Fig. 5). They show an arid environment at Bir Hmairiya but an open and more humid one at Kef el Agab, reflecting the different geographical locations of the two sites. Climatic patterns were better with wet conditions in phase 2 of SHM-1 with more diversified faunal spectra than in phase 1. The cenogram method is used in conjunction with other methods as an index of biodiversity (Fig. 6). The Simpson diversity Index D (1949) tends towards a value of 1 when diversity is minimal so that one species is present in the assemblage. The Shannon Index H0 (1948) of biodiversity has a value of 0 when only one taxon is present in the site, while it is maximal when each species of the spectra has the same

Please cite this article in press as: Aouadi, N., et al., Holocene environment and subsistence patterns from Capsian and Neolithic sites in Tunisia, Quaternary International (2013), http://dx.doi.org/10.1016/j.quaint.2013.07.028

N. Aouadi et al. / Quaternary International xxx (2013) 1e12

7

Fig. 3. Ratio diagram of Barbary sheep skeletal portions using NISP from Doukanet el Khoutifa. Differences are in logarithmic scale. The reference line (y ¼ 0) represents the number of bones in a complete Ovis skeleton.

influence in the assemblage. The H0 Index is maximal in all sites and in particular at Kef el Agab (H0 ¼ 2) indicating that there is no one species dominant. The faunal spectrum at Kef el Agab shows equal repartition within species.

The Neolithic sites (Kef el Agab and Doukanet el Khoutifa) are the most diversified. This is confirmed by the Simpson Index D, which is far from the value of 1 (all sites are between 0.2 and 0.3). The Hill Index (1973) combined both the Shannon and Simpson

Fig. 4. Distribution of faunal group association from several Holocene sites. A: Escarped environment little woody, B: Woody environment, C: Open environment not desert, D: Open desert environment. I: Distribution of faunal group association by % of MNIc from Bir Hmairiya. II: Distribution of faunal group association by % of MNIc from SHM-1. III: Distribution of faunal group association by % of MNIc from Kef el Agab. IV: Distribution of faunal group association by % of MNIc from Doukanet el Khoutifa.

Please cite this article in press as: Aouadi, N., et al., Holocene environment and subsistence patterns from Capsian and Neolithic sites in Tunisia, Quaternary International (2013), http://dx.doi.org/10.1016/j.quaint.2013.07.028

8

N. Aouadi et al. / Quaternary International xxx (2013) 1e12

Fig. 5. Cenograms plots from several Holocene sites. I: Cenogram plot from Bir Hmairiya. II: Cenogram plot from SHM-1. III: Cenogram plot from Kef el Agab. IV: Cenogram plot from the Dark level of Doukanet el Khoutifa.

indices; when it is close to 1, the faunal diversity is low. The Hill index from phase 1 of SHM-1 and from the Yellow level of Doukanet el Khoutifa shows low diversity in faunal spectra. The Taxonomic Richness Index of Margalef (1958) explains how many species are

present in the assemblage. The faunal assemblage from the Dark level of Doukanet el Khoutifa is the most varied, whereas phase 1 of SHM-1 and the Yellow level of Doukanet el Khoutifa contain lesser numbers of taxa.

Fig. 6. Biodiversity indices from Holocene sites in Tunisia. H0 : Shannon Index, D: Simpson Index, Hill: Hill Index, R: Margalef Index.

Please cite this article in press as: Aouadi, N., et al., Holocene environment and subsistence patterns from Capsian and Neolithic sites in Tunisia, Quaternary International (2013), http://dx.doi.org/10.1016/j.quaint.2013.07.028

N. Aouadi et al. / Quaternary International xxx (2013) 1e12

9

5.3. Economic strategies We tried to test some hypotheses about the economic strategies of the human groups from the four sites. Bone survival of skeletal elements depends on the structural density of the bones. To understand the effects of post-depositional processes on the faunal assemblage and to detect the origin of bone accumulation, we tested whether the presence of skeletal elements is linked to their density. For bone density, we used the data for wildebeest bones provided by Lam et al. (1999). Then, we investigated the relationship between the standardized food utility index (S)FUI (Metcalfe and Jones, 1988, in some cases we used the MGUI index) and skeletal element abundance (%MAU) in order to illustrate the possible selective transport of bones into or out of the residential base. We focused on the exploitation patterns of bovids by Capsian groups. There is no statistically significant relationship between bone density (data from Lam et al., 1999) and the % survival of Antilopini at SHM-1. Thus, the loss of gazelle remains is due either to pre- or post-depositional taphonomic processes. Nevertheless, the high degree of bone fragmentation may have influenced investigation into the relationship between bone density and faunal representation at SHM-1. In order to illustrate the economic utility model in accordance to skeletal part transport strategy, the MGUI index was compared against bone survival of skeletal elements (in %MAU) of Gazella at SHM-1. As expected, the results of both indices showed no correlation and bone frequency of gazelle shows no relation to MGUI index. Therefore, neither selective destruction (expressed by bone density) nor selective transport of skeletal elements can be cited as responsible for the missing bones. The retrieval of bone marrow, the making of bone tools and mainly taphonomic processes are the likely explanations for the high numbers of bone splinters on the site. Most of the bone remains (identified and unidentified elements) from all sites show clear signs of green breakage (smoother surfaces and oblique breakage planes). Taphonomic processes (mainly the dominance of fracturing green bones) indicate that most bone breakage occurred during the time of occupation of the site, producing large numbers of fragments. The mortality profile of gazelle bones from SHM-1 shows dominance of bones belonging to young individuals (Fig. 7) (data references are those of Davis, 1980). Skeletal part representation of gazelle at Kef el Agab shows the supremacy of autopodial elements (Fig. 8). Small bones such as tarsals, carpals, and phalanx are also well-represented in the assemblage. The mortality patterns of gazelle post-cranial remains from Kef el Agab show the dominance of a 5e10 month interval. Mature animals are rare and no very old individuals are present.

Fig. 7. Mortality profile of Gazella bones from SHM-1.

Fig. 8. Skeletal part representation for Gazella remains from Kef el Agab (S1 and S1external edge) (skeletal representation of goat from Barone, 1986).

Furthermore, for Alcelaphini remains from Kef el Agab, the regression line for bone survivorship based on %MAU and bone density indicates a non significant relationship between the two variables (R2 ¼ 0.14). Finally, there is no significant relationship between (S)FUI and %NNISP of Alcelaphini bone from Kef el Agab indicating that bones were unaffected by selective transport. In general, the pattern of skeletal element representation in wild mammals indicates primary access to carcasses by the occupants of Kef el Agab. Through examination of skeletal representation of small mammals (gazelles, Bohor reedbuck) by plotting the (S)FUI against %MAU, their carcasses may have been brought back whole to the cave. From Doukanet el Khoutifa, Caprini bone scatter plots show no relationship between density and %MAU nor between (S)FUI and % NNISP within Ammotragus lervia remains. Thus, the representation of bones was not affected by their density or by their caloric value. 6. Discussion The results from faunal remains to intercalate the four studied sites into the pattern of Holocene palaeoclimate variation for the eastern Mediterranean and adjacent regions indicate an arid landscape at Bir Hmairiya. SHM-1 was an open grassland landscape supporting grazing animals (hartebeest and gazelle) with some browsers (such as Giraffa). In addition, the presence of taxa heavily dependent on water, such as reedbuck (Redunca) and African buffalo (Syncerus), document the presence of water sources nearby. Kef el Agab was in an open savanna with some mountainous surroundings. Nevertheless, the presence of Syncerus caffer and reduncines indicates proximity to a perennial source of water, possibly the Wadi Medjerda, which has been active since the Iberomaurusian (Zielhofer et al., 2004, 2009). Doukanet el Khoutifa was also in an open savanna with some mountainous surroundings. The faunal analysis indicates several aspects, including skeletal element representation with absence of selective transport, surface modification (weathering, root marks, trampling), breakage patterns and cut marks from all stages of activities suggest that domestic animals were butchered at the site. Hunting was an important economic activity as indicated by the high percentage of wild mammal remains (aurochs, gazelle, hartebeest). Table 5 summarizes the most important results for each site. A high frequency of determination (84.5%) is found at Bir Hmairiya and the lowest frequency is at SHM-1 (8.2%). The taxonomic

Please cite this article in press as: Aouadi, N., et al., Holocene environment and subsistence patterns from Capsian and Neolithic sites in Tunisia, Quaternary International (2013), http://dx.doi.org/10.1016/j.quaint.2013.07.028

10

N. Aouadi et al. / Quaternary International xxx (2013) 1e12

richness indices in the Dark level of Doukanet el Khoutifa, phase 2 of SHM-1 and at Kef el Agab show that they are the richest in species. The diversity index indicates that the Dark level of Doukanet el Khoutifa is the most diversified assemblage. The assemblage from each site produced ungulate mortality patterns heavily dominated by prime age individuals (even the samples are very small), illustrating the hunting strategy e carnivores hunt juvenile or senile animals (see Steele, 2002, 2003 for a detailed discussion of ageing methods).

indicates dry conditions. All mammal species enumerated here no longer inhabit the area; natural and human factors caused their extinction. Gazelle, hartebeest, and giraffe shared the open woodland in eastern Tunisia during parts of the Holocene. Likewise, for SHM-1, the presence of African buffalo indicates proximity to a perennial source of water. In levels 5, 6 and 7, the degree of anthropogenic activity seems to be higher. As expected, most data indicate an open habitat pattern for SHM-1 and Bir Hmairiya and a mosaic habitat at Kef el Agab and Doukanet el Khoutifa.

Table 5 Summary of the main archaeological features of each site. Site

Bir Hmairiya

SHM-1

Kef el Agab

Doukanet el Khoutifa

Dates

*

6747e6887 cal BP (A.M. Mannino, personal communication)

Culture Type of faunal remains

Typical Capsian Wild species

Layer 1: 8587e8968 cal BP Layer 7: 7136e7478 cal BP (Saliège et al., 2013) Upper Capsian Wild species

% determined bones Taxinomic richness (Margalef Index) Correlation between % survival and density Mortality profile Cut marks Bone tools Human remains Ceramic Environment

84.5% *

8.2% 2.2 (stage 1) et 3.8 (stage 2)

Mediterranean Neolithic Wild species þ some domestic species 53.3% 3.8

Absent

Absent

Absent

Dark level: 6638e7158 cal BP Yellow level: 7266e7970 cal BP (Zoughlami, 2009) Tellian Neolithic Domestic species þ some wild species 21% 2.8 (yellow level) and 4.8 (dark level) Absent

Prime dominated Present Absent? Present? Absent Arid

Prime dominated Present Present Present Few ceramic Open landscape

Prime dominated Present Present Present Few ceramic Open Savanna þ Mountain surroundings

Prime dominated Present Present Present Number of ceramic Open Savanna þ Mountain surroundings

The presence of nearly all skeletal elements of the wild mammals at Kef el Agab and Doukanet el Khoutifa indicates good bone survival and primary access to their body parts by prehistoric populations and the practice of active hunting. For domestic species at both sites, the missing elements are due to the high degree of bone fragmentation. Remains from Caprini (Ovis and Capra) are more abundant than those from aurochs. In both sites, there are no remains of dog or pig. 7. Conclusions Recent archaeozoological studies outline the evolution of subsistence behaviors from Capsian to Neolithic populations during the Holocene in Tunisia. The faunal assemblages from SHM-1 and Bir Hmairiya provide information concerning Capsian subsistence during the lower Holocene. Large game hunting strategies were developed by hunter-gatherer populations in order to optimize carcass exploitation of various animals (hartebeest, gazelle). Capsian populations showed great adaptability to their physical environmental their hunting strategies could be reconstructed by examination of prey selection and carcass exploitation. At SHM-1, prehistoric populations developed subsistence strategies focused on hunting of a wide range of local wild prey with more exploitation of medium and small sized mammals, especially hartebeest and gazelle with additional heavy exploitation of food resources from other terrestrial mammals and land snails and from marine fish and molluscs. Mammalian faunal spectra from SHM-1 indicate a mixed steppe landscape for this area that changed slightly during the occupation of the site. Most palaeoclimatic proxies indicate environmental deterioration and onset of drier conditions in the midHolocene (Faust et al., 2004; Zielhofer and Faust, 2008). The first levels (1e4) of SHM-1 most likely correspond in faunal composition and dates to the fossil assemblage from Bir Hmairiya which also

The presence of reedbuck, waterbuck, and topi at Kef el Agab could indicate an area with more abundant green forage than is found today in northern Tunisia. African buffalo (at SHM-1 and Kef el Agab) inhabit valleys (as at Kef el Agab located 4 km north of Wadi Medjerda, the biggest valley in northern Tunisia) and grassland. Cultural and economic patterns indicate that these huntersgatherers populations were using long-term occupation sites, e.g. a mode of sedentary lifestyle during the Capsian. Hartebeest was the major hunted animal in the diet of Capsian populations throughout the lower and mid-Holocene in Tunisia. Taphonomic and archaeozoological data reported here from Holocene sites in Tunisia show that humans were the primary agent of modification and consumption of carcasses. The Neolithic economy is illustrated by archaeozoological analysis of two sites: Kef el Agab and Doukanet el Khoutifa. Changes are observed in species composition during the Holocene: wild species decreased and domestic ones increased. At Kef el Agab, wild fauna prevail, but Doukanet el Khoutifa showed increased consumption of domestic taxa. Kef el Agab and Doukanet el Khoutifa are located high on steep slopes that provided prehistoric groups the ability to track the movement of wild mammals in the surrounding area. The disproportionate paucity of complete long-bones in the four sites is explained by: the high degree of bone fragmentation at SHM-1; the previously selective choice of complete elements by the previous excavation teams at Bir Hmairiya and Kef el Agab; and the loss of cranio-dental remains at Doukanet el Khoutifa (given by J. Zoughlami in 1976 to Oxford University for determination). In accordance with the economic utility model (Binford, 1978), the composition of skeletal elements of wild mammals at Kef el Agab can be understood by selective transport of the choicest elements by the Neolithic occupants of the site. The age categories of Ovicaprid remains from Kef el Agab show a low percentage of old

Please cite this article in press as: Aouadi, N., et al., Holocene environment and subsistence patterns from Capsian and Neolithic sites in Tunisia, Quaternary International (2013), http://dx.doi.org/10.1016/j.quaint.2013.07.028

N. Aouadi et al. / Quaternary International xxx (2013) 1e12

animals, suggesting that wool production may not have been a major reason for their breeding. While it seems that their exploitation was mainly related to meat consumption, milk and wool may have also been exploited. At Kef el Agab, the presence of carnivores such as spotted hyena and jackal indicates that these animals may have been responsible for accumulating a small amount of bones. However, the presence of rather high numbers of wild species implies that hunting was a much-practiced activity at the settlement. From Doukanet el Khoutifa, the wild component of the bone sample is limited to a small number of species, in particular gazelle, Barbary sheep and most likely aurochs and Alcelaphini. The frequency of Caprini is rather high compared to that of cattle, and so the percentage of the sheep/goat in the assemblage points to the relatively great importance of these animals in the economy of the Neolithic population. Despite the varying numbers of skeletal elements from domestic animals at Doukanet el Khoutifa, the presence of cut marks from all stages of activities, suggest that those animals were killed and butchered at the camp. Comparing Kef el Agab and Doukanet el Khoutifa indicates that patterns of mammal exploitation in both sites do not display great differences. The occupants practiced hunting and gathered land snails. The main change lies in the increase in Ovicaprid remains at Doukanet el Khoutifa, which would suggest a greater control of breeding activity. Pastoralism was already established at Doukanet el Khoutifa by the first half of the 7th millennium cal BP. This pastoralism involved more sheep and goat than bovines, but without dog or pig. In northern Morocco, the Neolithic layer at Kaf Taht el-Ghar, dated to 6050
120 BP, contains more remains of Caprini than bovines, but includes pig (Ouchaou, 2012). From Abu Tamsa in Cyrenaica (Libya), dated between 7695
60 BP and 7275
40 BP (Bed IV) and 6605
40 BP (Bed III) there are some bone remains from Caprini without any remains of bovines or dog (de Faucamberge, 2012). The most likely explanation for the lack of bovines and dog and the presence of Caprini in the first layers of Neolithic sites is that domestication of Caprini in the Maghreb was allochthonous. The breeding of Caprini fits very well with the topography of the first types of camps of the Neolithic groups (caves and shelters). Some questions remain unanswered: if there were attempts at domestication in the Maghreb, what species were domesticated? Why did land snail gathering and hunting of wild mammals persist in Neolithic contexts? Acknowledgements We are grateful to David Lubell for inviting us to present this study at the 11th Biennial Conference of the Society for Africanist Archaeologists (SaFa), for his useful comments, and for his English revision of this text. The authors wish to thank Simone Mulazzani (UMR 7041 ArScAn) for the constructive remarks on an earlier version of this article and for helpful discussions. Dating of Kef el Agab has been obtained at the Heidelberger laboratory, thanks to collaboration with Marcello A. Mannino, of the Max Planck Institute for Evolutionary Anthropology of Leipzig, and with Sahra Talamo. References Aouadi, N., Dridi, Y., Maini, E., Curci, A., Brugal, J.-Ph., Mannai-Tayech, B., 2013. La faune de la rammadiya capsienne de SHM-1. In: Mulazzani, S. (Ed.), Le capsien de Hergla (Tunisie). Culture, environnement et économie, Reports in African Archaeology, vol. 4. Africa Magna Verlag, Frankfurt, pp. 318e330. Bardin, P., 1953. La grotte du Kef-el-Agab (Tunisie). Gisement néolithique. Libyca 1, 271e305. Barone, R., 1986. Anatomie comparée des mammifères domestiques. In: Tome premier: Ostéologie. Ecole Nationale Vétérinaire - Laboratoire Anatomie, Lyon.

11

Behrensmeyer, A.K., 1978. Taphonomic and ecologic information from bone weathering. Paleobiology 4, 150e162. Behrensmeyer, A.K., 1991. Terrestrial vertebrate accumulations. In: Allison, P.A., Briggs, D.E.G. (Eds.), Taphonomy: Releasing the Data Locked in the Fossil Record. Plenum, New York, pp. 291e327. Belhouchet, L., Mulazzani, S., Pelegrin, J., 2013. Evolution of a 9the8th mill. cal BP Upper Capsian site: the techno-typological study of bladelet production at SHM-1 (Hergla, Tunisia). Quaternary International. http://dx.doi.org/10.1016/ j.quaint.2013.05.006. Ben-Dhia, W., 2002. Etude du gisement de Bir Hamairia d’après la collection conservée au Musée National de Bardo (Tunis). In: Mémoire de Magistère en Archéologie et Patrimoine. Université de Tunis 1, Faculté des Sciences humaines et sociales de Tunis. Binford, L.R., 1978. Dimensional analysis of behavior and site structure: learning from an Eskimo hunting stand. American Antiquity 43 (3), 330e361. Blumenschine, R.J., Selvaggio, M.M., 1988. Percussion marks on bone surfaces as a new diagnostic of hominid behaviour. Nature 333, 763e765. Blumenschine, R.J., Marean, C.W., Capaldo, S.D., 1996. Blind tests of inter-analyst correspondence and accuracy in the identification of cut marks, percussion marks, and carnivore tooth marks on bone surfaces. Journal of Archaeological Science 23, 493e507. Brain, C.K.,1980. Some criteria for the recognition of bone-collecting agencies in African caves. In: Behrensmeyer, A.K., Hill, A.H. (Eds.), Fossils in the Making: Vertebrate Taphonomy and Paleoecology. University of Chicago Press, Chicago, pp. 107e130. Brain, C.K., 1981. In: The Hunters or the Hunted? An Introduction to African Cave Taphonomy. University of Chicago Press, Chicago, p. 376. Brugal, J.-P., Patou-Mathis, M., 1993. L’assemblage osseux de l’abri des Canalettes: présentation générale. In: Meignen, L. (Ed.), L’abri des Canalettes, un habitat moustérien sur les grands Causses (Nant, Aveyron): fouilles 1980e1986, C.N.R.S (ed.), pp. 77e87. Bunn, H.T., 1983. Comparative analysis of modern bone assemblages from a San hunter-gatherer camp in the Kalahari Desert, Botswana, and from a spotted hyena den near Nairobi, Kenya. In: Clutton-Brock, J., Grigson, C. (Eds.), Animals and Archaeology, British Archaeological Reports International Series, Oxford, pp. 143e148. Camps, G., 1974. Les civilisations préhistoriques de l’Afrique du Nord et du Sahara. Doin, Paris. Davis, S.J.M., 1980. A note on the dental and skeletal ontogeny of Gazella. Israel Journal of Zoology 29, 129e134. de Faucamberge, E., 2012. Néolithisation et Néolithique en Cyrénaïque (Libye). Encyclopédie Berbère 34, 5474e5481. Faust, D., Zielhofer, C., Baena Escudero, R., Diaz del Olmo, F., 2004. High-resolution fluvial record of late Holocene geomorphic change in northern Tunisia: climatic or human impact? Quaternary International 23, 1757e1775. Fosse, P., Brugal, J.P., Guadelli, J.L., Michel, P., Tournepiche, J.F., 1998. Les repaires d’hyènes des cavernes en Europe occidentale: présentation et comparaison de quelques assemblages osseux. In: Economie préhistorique: les comportements de subsistance au Paléolithique. XVIIIèmes rencontres internationales d’Antibes, 1997, APDCA (ed), pp. 43e61. Gabler, K.O., 1985. Osteologische Unterscheidungsmerkmale am postcranial Skelett zwischen Mahnenspringer (Ammotragus lervia) Hausschaf (Ovis aries) und Hausziege (Capra hircus). Gobert, E.G., 1937. Les escargotières, le mot et la chose. Revue Africaine 81, 638e 645. Grayson, D.K., Frey, C.J., 2004. Measuring skeletal part representation in archaeological faunas. Journal of Taphonomy 2 (1), 27e42. Hammer, Y., Harper, D.A.T., Ryan, P.D., 2001. Paleontological statistics software package for education and data analysis. Palaeontologia Electronica 4 (1), 9. Harbi-Riahi, M., Zoughlami, J., 1971. La Rammadiya de Sebkha Halk el Mennzel, pp. 181e184. Africa III et IV. Hill, M.O., 1973. Diversity and evenness: a unifying notation and its consequences. Ecology 54 (2), 427e432. Jousse, H., 2003. Impact des variations environnementales sur la structure des communautés mammaliennes et l’anthropisation des milieux: exemples des faunes holocènes du Sahara Occidental. L’université Claude Bernard e Lyon 1. Klein, R.G., Cruz-Uribe, K., 1984. The Analysis of Animal Bones from Archaeological Sites. University of Chicago Press, Chicago. Lam, Y.M., Chen, X., Pearson, O.M., 1999. Intertaxonomic variability in pattern of bone density and the differential representation of Bovid, Cervid, and Equid elements in the archaeological record. American Antiquity 64, 343e362. Legendre, S.., 1986. Analysis of mammalian communities from the late Eocene and Oligocene of Southern France. Palaeovertebrata 16 (4), 191e212. Legendre, S., 1989. Les communautés de Mammifères du Paléogène (Eocène supérieur et Oligocène) d’Europe occidentale: structures, milieux et évolution. Münchner Geowissenschaftliche Abhandlungen, München 16, p. 110. Lyman, R.L., 1994. Vertebrate Taphonomy. Cambridge University Press, Cambridge, p. 524. Marean, C.W., 1991. Measuring the post-depositional destruction of bone in archaeological assemblages. Journal of Archaeological Science 18, 677e694. Margalef, R., 1958. Temporal succession and spatial heterogeneity in natural phytoplankton. In: Buzzati-Traverso, A.A. (Ed.), Perspectives in Marine Biology. University of California Press, Berkeley, pp. 323e349. Marty, M., 1966. Le Capsien de Bir Hamaïria. Libyca XIV, 115e160. Merzoug, S., 2012. Nourriture (Préhistoire): alimentation carnée. Encyclopédie Berbère 34, 5622e5625.

Please cite this article in press as: Aouadi, N., et al., Holocene environment and subsistence patterns from Capsian and Neolithic sites in Tunisia, Quaternary International (2013), http://dx.doi.org/10.1016/j.quaint.2013.07.028

12

N. Aouadi et al. / Quaternary International xxx (2013) 1e12

Metcalfe, D., Jones, T., 1988. A reconsideration of animal body-part utility indices. American Antiquity 53, 486e554. Mulazzani, S., 2010. L’habitat Epipaléolithique de SHM-1 et les sites environnants au bord de la Sebkha lagune de Halk el Menjel (Hergla - Tunisie) entre le VIIe et le VIe millénaire cal bc (Thèse de doctorat). Université Paris 1 e PanthéonSorbonne, université degli studi di Bologna. Mulazzani, S. (Ed.), 2013. Le capsien de Hergla (Tunisie). Culture, environnement et économie. Reports in African Archaeology, vol. 4. Africa Magna Verlag, Frankfurt. O’Connell, J.F., Hawkes, K., Blurton Jones, N., 1988a. Hadza scavenging: implications for Plio/Pleistocene hominid subsistence. Current Anthropology 29 (2), 356e363. O’Connell, J.F., Hawkes, K., Blurton Jones, N., 1988b. Hadza hunting, butchering, and bone transport and their archaeological implications. Journal of Anthropological Research 44 (2), 113e161. O’Connell, J.F., Hawkes, K., Blurton Jones, N.G., 1990. Reanalysis of large mammals body part transport among the Hadza. Journal of Archaeological Science 17, 301e316. O’Connell, J.F., Hawkes, K., Blurton Jones, N., 1992. Patterns in the distribution, site structure and assemblage, composition of Hadza Kill-butchering sites. Journal of Archaeological Science 19, 319e345. Ouchaou, B., 2012. Néolithisation du Maroc: comportements de subsistance et domestication animale. Encyclopédie Berbère 34, 5512e5520. Peters, J., 1986. Osteomorphology and Osteometry of the Appendicular Skeleton of African Buffalo, Syncerus caffer and Cattle, Bos primigenius F. taurus Bojanus. Occasional Papers. Laboratorium Voor Paleontologie. Rijksuniversit Gent, p. 9. Peters, J., Van Neer, W., Plug, I., 1997. Comparative postcranial osteology of hartebeest (Alcelaphus buselaphus), scimitar oryx (Oryx dammah) and addax (Addax nasomaculatus), with notes on the osteometry of gemsbok (Oryx gazella) and Arabian oryx (Oryx leucoryx). Annales des Sciences zoologiques du Musée royal de l’Afrique centrale 280. Tervuren. Potts, R., Shipman, P., 1981. Cutmarks made by stone tools on bone from Olduvai Gorge, Tanzania. Nature 291, 577e580. Rahmani, N., Lubell, D., 2012. Early Holocene climate change and the adoption of pressure technique in the Maghreb: the Capsian sequence at Kef Zoura D (Eastern Algeria). In: Desrosiers, P. (Ed.), The Emergence of Pressure Blade Making. From Origin to Modern Experimentation. Springer, pp. 139e156. Reimer, P.J., McCormac, F.G., 2002. Marine radiocarbon reservoir corrections for the Mediterranean and Aegean seas. Radiocarbon 44 (1), 159e166. Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey, C., Buck, C.E., Burr, G.S., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P., Hajdas, I., Heaton, T.J., Hogg, A.G., Hughen, K.A., Kaiser, K.F.,

Kromer, B., McCormac, F.G., Manning, S.W., Reimer, R.W., Richards, D.A., Southon, J.R., Talamo, S., Turney, C.S.M., van der Plicht, J., Weyhenmeyer, C.E., 2009. IntCal09 and Marine09 radiocarbon age calibration curves, 0e50,000 years cal BP. Radiocarbon 51 (4), 1111e1150. Reitz, E., Wing, E.S., 1999. Zooarchaeology. Cambridge University Press. Saliège, J.-F., Magnani, G., Mulazzani, S., 2013. Datations 14C de la Rammadiya SHM1. In: Mulazzani, S. (Ed.), Le capsien de Hergla (Tunisie). Culture, environnement et économie, Reports in African Archaeology, vol. 4. Africa Magna Verlag, Frankfurt, pp. 145e150. Shannon, C.E., 1948. A mathematical theory of communication. The Bell System Technical Journal 27, 379e423, 623e656. Simpson, E.H., 1949. Measurement of diversity. Nature 163 (4148), 688. Steele, T.E., 2002. Accuracy of age determinations from tooth crown heights: a test using an expanded sample of known age red deer (Cervus elaphus). In: Recent Advances in Ageing and Sexing Animal Bones, Durham, pp. 119e128. Steele, T.E., 2003. Using mortality profiles to infer behavior in the fossil record. Journal of Mammalogy 84 (2), 418e430. Stiner, M.C., 1990. The use of mortality patterns in archaeological studies of hominid predatory adaptations. Journal of Anthropological Archaeology 9, 305e351. Travouillon, K.J., Legendre, S., 2009. Using cenograms to investigate gaps in mammalian body mass distributions in Australian mammals. Palaeogeography, Palaeoclimatology, Palaeoecology 272, 69e84. Valverde, J.A., 1964. Remarques sur la structure et l’évolution des communautés de Vertébrés terrestres, I: Structure d’une communauté, II: Rapports entre prédateurs et proies. In: La terre et la vie, Paris, vol. 111, pp. 121e153. Villa, P., Mahieu, E., 1991. Breakage patterns of human long bones. Journal of Human Evolution 21, 27e48. Voorhies, M.R., 1969. Taphonomy and Population Dynamics of an Early Pliocene Vertebrate Fauna, Knox County, Nebraska e Contributions to Geology. University of Wyoming. Special Paper, 1. Zielhofer, C., Faust, D., 2008. Mid- and Late Holocene fluvial chronology of Tunisia. Quaternary Science Reviews 27, 580e588. Zielhofer, C., Faust, D., Baena Escudero, R., Diaz del Olmo, F., Kadereit, A., Moldenhauer, K.M., Porras, A., 2004. Centennial-scale late-Pleistocene to midHolocene synthetic profile of the Medjerda Valley, northern Tunisia. The Holocene 14 (6), 851e861. Zielhofer, C., Recio Espejo, J.M., Nunez Granados, M.A., Faust, D., 2009. Durations of soil formation and soil development indices in a Holocene Mediterranean floodplain. Quaternary International 209, 44e65. Zoughlami, J., 2009. Le Néolithique dans la Dorsale tunisienne. Kef el Guéria et sa région. Centre de Publication Universitaire, Tunis, p. 216.

Please cite this article in press as: Aouadi, N., et al., Holocene environment and subsistence patterns from Capsian and Neolithic sites in Tunisia, Quaternary International (2013), http://dx.doi.org/10.1016/j.quaint.2013.07.028