2006

UDC 572

VOLUME 30

3/2006 VOLUME 30 Num. Curr. 74

CODEN: COANDS

ISSN 0350-6134

Collegium Antropologicum Zagreb – CROATIA, September 2006

No. 3

CONTENTS ORIGINAL SCIENTIFIC PAPERS I. Jankovi}, I. Karavani}, J. C. M. Ahern, D. Brajkovi}, J. Mauch Lenardi} and F. H. Smith

Vindija Cave and The Modern Human Peopling of Europe . . . . . . . . 457

S. Forenbaher and P. Raji} [ikanji}

The Prehistoric Hillfort at Grad (Pelje{ac, Dalmatia) – Preliminary Results of Intensive Surface Survey . . . . . . . . . . . . . 467

M.-H. Cazes

An Example of Demographic Anthropology, the Study of Matrimonial Exchanges – Endogamy, Choice of Spouse and Preferential Marriage . . 475

R. Manfredi, S. Sabbatani and D. Agostini

Trend of Mortality Observed in a Cohort of Drug Addicts of the Metropolitan Area of Bologna, North-Eastern Italy, During a 25-Year-Period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 479

A. Dvornik-Radica, V. Rudan, V. Jure{a, D. Ivankovi}, M. Rumboldt, E. Smoje, D. Vrdoljak and N. Mrdulja{-\uji}

Do We Need the »Adolescent Crisis« Diagnosis? . . . . . . . . . . . . . 489

M. Vrca Botica, I. Zeli}, I. Pavli} Renar, B. Bergman Markovi}, S. Stojadinovi} Grgurevi} and I. Botica

Structure of Visits Persons with Diabetes in Croatian Family Practice – Analysis of Reasons for Encounter and Treatment Procedures using the ICPC-2 . . . . . . . . . . . . . . . . . . . . . . . . 495

I. Vasilj, S. âvaljuga, P. Petrovi}, Lj. Ostoji}, Z. Ostoji}, A. Kvesi} and V. Martinovi}

Cerebrovascular Insult Hospital Cases in the Clinical Hospital Mostar (Bosnia and Herzegovina) From 1999 to 2003 – An Example of an Institutional Register . . . . . . . . . . . . . . . . . . . . . . . . . 501

Lj. Ostoji}, A. Bradari}, K. Mi{e, Z. Ostoji}, J. Lovri}, P. Petrovi}, A. Ujevi}, M. Erceg, S. Jankovi} and J. Tocilj

Pulmonary Function in Persons Who are Professionally Exposed to Formaldehyde Fumes . . . . . . . . . . . . . . . . . . . . . . . . . . 507

M. Alilovi}, T. Pero{-Golubi~i}, J. Tekavec-Trkanjec, S. Smojver-Jeèk and R. Li{~i}

Epidemiological Characteristics of Sarcoidosis Patients Hospitalized in the University Hospital for Lung Diseases »Jordanovac« (Zagreb, Croatia) in the 1997–2002 Period . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513

A. V~ev, I. Begi}, R. Ostoji}, D. Jur~i}, D. Boì}, I. Soldo, R. Gmajni}, G. Kondà, E. Khaznadar and N. Mi}unovi}

Esomeprazole Versus Pantoprazole for Healing Erosive Oesophagitis . . 519

D. Raheli}, M. Kujundì}, @. Romi}, K. Brki} and M. Petrove~ki

Serum Concentration of Zinc, Copper, Manganese and Magnesium in Patients with Liver Cirrhosis. . . . . . . . . . . . . . . . . . . . . . . . 523

M. Maru{i}, V. Prese~ki, M. Kati~i}, M. Dominis and S. Kaleni}

The Place and Role of Serologic Methods in Detecting Helicobacter Pylori Infection . . . . . . . . . . . . . . . . . . . . . . . . 529

Coll. Antropol. 30 (2006) 3, Zagreb, Croatia

T. Mi{evi}, B. Brklja~i}, L. Zibar, M. Jaki}, S. Kurbel, R. Radi} and S. Mi{evi}

Duplex Sonography of Arteriovenous Fistula in Chronic Hemodialysis Patients . . . . . . . . . . . . . . . . . . . . . . . . . . . 535

D. Mijatovi}, K. Buli}, I. Dèpina and J. Unu{i}

The Supply of Blood in the Skin Territory Above the Lower Part of the Serratus Anterior Muscle . . . . . . . . . . . . . . . . . . . . . . 543

D. Pa{ali}, G. Feren~ak, B. Gr{kovi}, M. [esto and A. Stavljeni}-Rukavina

Association of Two Genetic Variations of Lipoprotein Lipase, S447X and Hind III, with Coronary Artery Disease and Hypertriglyceridemia. . . 549

R. Terzi}, A. [ehi}, N. Teran, I. Terzi} and B. Peterlin

Frequency of HFE Gene Mutations C282Y and H63D in Bosnia and Herzegovina . . . . . . . . . . . . . . . . . . . . . . . . . 555

S. Polovina, M. Cvjeti~anin, J. Mili~i} and T. Polovina Prolo{~i}

Dermatoglyphs and Brachial Plexus Palsy . . . . . . . . . . . . . . . . 559

G. Pavli{a, J. Papa, L. Pavi} and G. Pavli{a

Bilateral MR Volumetry of the Amygdala in Chronic PTSD Patients . . 565

R. Poljak-Guberina, A. êlebi}, O. @ivkovi}, M. Guberina and A. Mulja~i}

Denture Repairs in Different Regions of Croatia in Relation to Prosthodontic Teams . . . . . . . . . . . . . . . . . . . . . . . . . . 569

N. Petri~evi}, M. Katunari}, K. Mehuli}, P. Simeon, K. Rener-Sitar and A. êlebi}

Selection of Appropriate Artificial Frontal Teeth Size Using Dimensions of Hard Palate . . . . . . . . . . . . . . . . . . . . . . . . . 573

T. Baruah, S. Mondal, A. Kumar Gharami and D. Kumar Adak

The Tai-Phake of Assam, India – A Morphometric Study and Population Comparison with Neighbouring Groups . . . . . . . . . . . 579

H. Lali}, N. Kalebota and M. Kabalin

Measures for Achieving Recruits' Enhanced Fitness – A Transversal Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585

D. Mijatovi}, K. Buli} and V. Nikoli}

Quantification Model for Muscular Forces and Momentums in Human Lower Extremities . . . . . . . . . . . . . . . . . . . . . . . 593

V. Srhoj, N. Rogulj, N. Zagorac and R. Kati}

A New Model of Selection in Women’s Handball . . . . . . . . . . . . . 601

G. Markovi} and D. Sekuli}

Modeling the Influence of Body Size on Weightlifting and Powerlifting Performance . . . . . . . . . . . . . . . . . . . . . . . 607

I. Goi}-Bari{i}, A. Bradari}, M. Erceg, I. Bari{i}, N. Foreti}, N. Pavlov and J. Tocilj

Influence of Passive Smoking on Basic Anthropometric Characteristics and Respiratory Function in Young Athletes . . . . . . 615

J. Pau{i}, M. âvala and R. Kati}

Relations of the Morphological Characteristic Latent Structure and Body Posture Indicators in Children Aged Seven to Nine Years . . . . . 621

A. Perini}

The »Harsh Inhabitants of Hvar« in the Speech of Vinko Pribojevi} (A. D. 1525) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 629

H. Greil

Patterns of Sexual Dimorphism from Birth to Senescence . . . . . . . . 637

M. Balla, R. Angelopoulou, G. Lavranos and P. Manolakou

Gonadal Cell Proliferation Dimorphism . . . . . . . . . . . . . . . . . . 643

REVIEWS P. Manolakou, R. Angelopoulou and G. Lavranos

Sex Determinants in the Genome – Lessons from the Animal Kingdom . . 649

R. Angelopoulou, G. Lavranos and P. Manolakou

Establishing Sexual Dimorphism in Humans . . . . . . . . . . . . . . . 653


G. Lavranos, R. Angelopoulou, P. Manolakou and M. Balla

Hormonal and Meta-Hormonal Determinants of Sexual Dimorphism . . . . . . . . . . . . . . . . . . . . . . . . . . . 659

A. [erman, M. Vlahovi}, Lj. [erman and F. Buli}-Jaku{

DNA Methylation as a Regulatory Mechanism for Gene Expression in Mammals . . . . . . . . . . . . . . . . . . . . . 665

S. Missoni

Nutritional Studies in Croatia – A Century of Research . . . . . . . . . 673 BOOK REVIEW

P. Raji} [ikanji}

Prehistoric Herders of Northern Istria (Croatia): The Archaeology of Pupi}ina Cave, volume I/ Pretpovijesni sto~ari sjeverne Istre: Arheologija Pupi}ine pe}i, 1. svezak (Eds. Preston Miracle and Sta{o Forenbaher. Monografije i katalozi 14, Arheolo{ki muzej Istre, Pula, 2006) . . . . . . . . . . . . . . . . . . 697 IN MEMORIAM Milan F. Pospí{il Obituary . . . . . . . . . . . . . . . . . . . . . . . . . 699


COLLEGIUM ANTROPOLOGICUM Indexed in: Current Contents Social Sciences Citation Index Index Medicus/MEDLINE International Current Awareness Service: Anthropology Abstracts in Anthropology Anthropological Literature Linguistics and Language Behavior Abstracts INIST/CNRS Sociological Abstracts Science Culture SARL UnCover CSA Sociological Abstracts International Center for Scientific Research (CIRS) Ulrich’s International Periodical Directory

Coll. Antropol. 30 (2006) 3: 457–466 Original scientific paper

Vindija Cave and The Modern Human Peopling of Europe Ivor Jankovi}1, Ivor Karavani}2, James C. M. Ahern3, Dejana Brajkovi}4, Jadranka Mauch Lenardi}4 and Fred H. Smith5 1 2 3 4 5

Institute for Anthropological Research, Zagreb, Croatia Department of Archaeology, Faculty of Humanities and Social Sciences, University of Zagreb, Zagreb, Croatia Department of Anthropology (3431), University of Wyoming, Larami, USA Institute for Quaternary Paleontology and Geology, Croatian Academy of Sciences and Arts, Zagreb, Croatia Department of Anthropology, Loyola University, Chicago, USA

ABSTRACT Vindija cave in Croatia has yielded the youngest securely dated Neandertal skeletal remains in Central/Eastern Europe. In addition, these remains have been found in association with archaeological material exhibiting Upper Paleolithic elements. Due to its geographic location and date, the Vindija remains are particularly crucial for the understanding of initial modern human peopling of Europe and the nature of the Neandertal demise. The significance of archaeological and paleontological finds and hominin fossils from this site is discussed in the light of new finds at Vindija and recent developments in the fields of paleoanthropology and prehistoric archaeology. Furthermore, the impact of revised chronology for several crucial specimens and sites throughout Europe, including Vindija, is discussed. Key words: Vindija cave, modern human origins, Neandertals, human evolution, Upper Paleolithic

Introduction and Brief Site History The site of Vindija is a large cave, about 50 m in length, 28 m in width, and almost 20 m in height (Figure 1). It is located in the Hrvatsko Zagorje region of Croatia, 9 km northwest of Ivanec and about 20 km west from the center of Vara`din1 It was first mentioned as a potentially interesting archaeological site by D. Hirc2. Initial archaeological excavations were conducted by S. Vukovi}3–5 starting in 1928, but it was not until the mid-1970s that large-scale excavations started under the direction of M. Malez1,6. It was under his direction that the majority of the paleontological, archaeological, as well as the entire hominin sample was unearthed between 1974 and 19867–9. Since then, several additional hominin fossils have been identified10–12, and the archaeological and faunal assemblage has been a subject of detailed analyses12–18. The stratigraphic sequence of the site is complex, consisting of over 12 m of deposits, divided into 13 basic stratigraphic units (A–M). Complexes F, G and K are further subdivided into Fg, Fs, Fd, Fd/d, G1 to G5, and K1 to K3 layers12,19,20. Units A to D are Holocene, while units D to M yielded material dated to the Pleistocene (Figure 2).

Faunal and sedimentological analysis suggests that the climate during the formation of complex G (OIS 3) was variable but at times similar to the recent one, while the younger complex E/F (OIS 2) was deposited under somewhat cooler climatic conditions. Of major interest for the modern human origins debate in Europe is the material from complex G. This stratigraphic unit yielded most of the Neandertal bones from the site. The archaeological assemblage is quite complex. While the tools from G3 are attributed to the Mousterian with some Upper Paleolithic elements present, the G1 assemblage provides a more complicated picture12,15. It is in this layer that a Neandertal mandible (Vi-207) was found in association with Aurignacian or Aurignacian like split base bone point (Vi-3437) (Figure 3). Additionally, three massive-base bone points (so-called Mlade~ type) were found in the same layer. Such bone points are distinctly Upper Paleolithic tools. The stone tool assemblage from G1 exhibits a mixture of Mousterian and Upper Paleolithic types15 (Figure 4). One well-made bifacial stone point made from non-local raw material shows similarities to material from Hungary usually attributed to the Szele-

Received for publication July 5, 2006

457

I. Jankovi} et al.: Vindija Cave and The Modern Human Peopling of Europe, Coll. Antropol. 30 (2006) 3: 457–466

Fig. 1: Vindija cave (photo: I. Karavani}).

tian industry (Figure 4; 4). Whether the archaeological material from G1 represents the Aurignacian or some other variant of the initial Upper Paleolithic, as a »transitional« industry (e.g. Szeletian), or the late Mousterian with Upper Paleolithic components remains uncertain15,21–23. Complex F has yielded archaeological material attributable to the Aurignacian sensu lato (layer Fd/d) and Epigravettian (layers Fd/s, Fs, and Fg), while

Fig. 2: Stratigraphic profile of the Vindija Cave (modified after Rukavina 1983).

458

Fig. 3: Split-base bone point Vi 3437 and hominin mandible Vi 207.


Fig. 5: Accumulated fossil remains of ungulata in taphodermes of layers/complexes D, E/F, G, and H.

Fig. 4: Selected artefacts from Vindija level G1: 1. probable a pseudo- tool (previously published as denticulated piece), 2. burin. 3. sidescraper. 4. leaf-shaped bifacial piece, 5. flake with marginal retouch on distal end (previously published as an endscraper on flake), 6. massive base bone point, split base bone points. (Modified after Karavani} 1995: Fig. 3; Drawing by Marta Perki}.

the E layer is Epigravettian12–14. In layer D, modern human (Homo sapiens sapiens) skeletal material has been found alongside material attributed to the Epigravettian. The majority of the anatomically modern human sample comes from this layer, although the inscriptions on several fragments suggests that they were found near the border with the E sequence, and a few fragments might belong to the Holocene layer B. In this paper, we will concentrate on the finds from complex G, as those are crucial to the »Neandertal question« and the modern human peopling of Europe.

Vindija Faunal Sample During the Upper Pleistocene, Vindija cave was situated on the southern edge of the Alpine ice sheet, which at the times of the glacial maximum covered the Alps. However, Vindija also lies near the edge of the Pannonian Plain, which explains the steppe elements in the classical forest faunal community during the OIS 2 and 3. As majority of the Vindija finds are faunal, the zooarchaeological sample from this site has been studied at numerous

times19,20,24–26. With a better understanding of taphonomy of the site and more detailed studies of specific taxa, new patterns emerged. A recent revision of the ungulate fauna removes Coelodonta antiquitatis (woolly rhino), Saiga tatarica (saiga antelope), and Equus cf. germanicus from D, E/F, and G complexes at Vindija18. The presence of the first two taxa was considered to be evidence of extremely cold paleoclimatic conditions, while the equid was considered indicative of open, steppe environments during the period in which these complexes were deposited19. Rangifer tarandus (reindeer) is representative of the tundra zones and more open parts of the taiga. Fossil remains of reindeer were reported to be present in complexes D, E, F, and G19, but now we know that only a few skeletal remains (MNE 5) were accumulated in complexes E/F and G. On the other hand, our recent revision has added Capreolus caprolus (roe deer) to the faunal lists of the E/F and G complexes. Results have shown that the abundance of ungulate remains is highest in complex G (Figure 5). The revised faunal associations better accord with the palaeoclimatic reconstructions based on sedimentological characteristics27 and paleovegetation28. Results of this new revision of the Vindija faunal assemblage call into question the previous reconstruction of alternating »cold« and »warm« faunal communities during the deposition of the E, F and G complexes. The Vindija ungulate assemblage has undergone a complex taphonomic history. Traces of animal modification (e.g. gnaw marks) point to the activities of small-sized carnivores (e.g. fox and marten) and rodents. Of the larger-sized carnivores, Ursus spelaeus (cave bear) is ubiquitous throughout the Vindija sequence, and it is the only large carnivore present in the lowest strata of complex G19. Cave bears probably occupied the cave for hibernation17. The other larger-sized carnivores are Panthera spelaea (cave lion) and Canis lupus (wolf); both are present in the assemblages of complexes D, E, F and G. There are only rare gnaw marks, most probably from wolf, on the ungulate remains, and these appear to have been made on bone refuse left by the hominins. In contrast, our new taphonomic analyses have produced widespread evidence of hominin selection and modification (e.g. body part selection, breakage patterns, butchery marks). This shows that hominins were the most important accumulators of the ungulate assemblage. 459


Chronometric Dating and the Early Modern Human Sequence in Europe The significance of establishing a reliable chronological framework in human evolutionary studies cannot be overemphasized, and improvement in dating techniques and redating of a number of finds has led to major reevaluations of both data and interpretations concerning modern human origins29. Recently, several key fossils have been redated by more precise methods. This led to exclusion of several specimens previously held to be among the earliest modern humans in Europe from the debate (e.g., Vogelherd and Velika Pe}ina, now dated to the Neolithic30,31). As the Vindija Neandertal remains are crucial to the debate, dating of various stratigraphic layers of this site has been attempted several times, but not without problems32. Neandertal remains from later G1 were directly dated by AMS and yielded a date of 28–29 kya, thus making them the youngest Neandertals in the region30. Recently, the new technique of ultrafiltration of collagen samples has been applied and the same G1 fossils have been redated to 32–33 000 14C years ago33. Until the same methods are applied to other crucial specimens of approximately same time period (both late Neandertals and early anatomically modern humans in the region) it is impossible to create the much needed timeframe of overlap of these two populations in Europe. The main problem with radiocarbon dating is a high error margin for material older than about 30 kya. Newer techniques, such as AMS, ultrafiltration, etc., add to the accuracy of dating and make these methods less destructive34. However when the time of overlap is expected to be several thousand years at best, the error margin is still unacceptably high. Further, many specimens from this crucial time period (e.g. Mlade~, Kostenki, etc.) are likely to be older than reported35,36. Therefore, the redating of the Vindija specimens does not necessarily widen the temporal gap between indigenous European Neandertals and anatomically modern newcomers. At present, and based on the radiocarbon dates of the finds, candidates for the oldest anatomically modern human remains from Europe are those from Kent's Cavern, England, Brassempouy and La Quina in France, Kostenki in Russia, Oase, Cioclovina and Baia de Fier in Romania, and Mlade~ in Czech Republic. However, there are problems with all of these sites. Kent's Cavern 4 is a human maxillary fragment found in 1927 in a large cave system near Torquay, England37. It was found below the layer containing what was described as »Aurignacoid» industry22,38–40, making the association of archeological industry and human fossil questionable. The fossil was directly dated to around 31 kya40,41, but it may be as old as 35–37 kya42. Although this specimen was described as modern in morphology, the fragmentary state makes this assesment uncertain36 and new analyses are still in progress. The exact nature of the »Aurignacoid» industry also needs to be subjected to careful re-analysis. Several isolated teeth and phalanges found at Brassempouy in France yielded dates between 30000 and 33500 years ago43. As is the case with Kent's Cavern finds, the archae460

ological industry of this site needs serious reexamination before it can be confirmed as Aurignacian sensu stricto. An additional problem is that the metric values of the human fossils fall both within modern human and Neandertal ranges43 contra 44. La Quina 25 is stratigraphically associated with the radiocarbon date of around 32 kya45, and not directly dated. Further, the specimen is juvenile which always presents an additional problem in taxonomic assesments. The Kostenki 1 specimen has recently been directly dated to around 32 kya46, but a detailed morphological analysis is still unpublished. Recently, human fossils from several Romanian sites have been directly dated47. A skull, tibia and scapula from the Woman's cave (Baia de Fier) were found in 1952 and the postcranial remains have been dated to around 30 kya48. The archaeological finds from the site have been described as Mousterian, while the upper layers contain some type of Upper Paleolithic industry. As the layers in the cave are mixed, the association of archaeological industries, as well as various human fossil elements are unclear. The skull from Cioclovina cave, most likely male49 is now dated to around 29 kya50 has been described by Rainer and Simionescu51 as »Homo sapiens fosillis…with Neanderthalian characters«, and although it is morphologically modern in overall gestalt, its supraorbital region is very robust and there is bunning on the occipital bone22,36,49. Cranial and postcranial remains from Pestera Muierii48 are approximately 30 000 years old50, but not associated with archaeological industry. The most recent finds come from Pestera cu Oase in Romania and are dated to arround 35 kya52,53. These were also not found in association with archaeological material. Trinkaus and colleagues52,53 note the presence of several archaic features on these otherwise anatomically modern specimens (e.g. pronounced juxtamastoid eminence on Oase 3, robust and laterally oriented zygomatic bones and large molars in Oase 2). At least one feature (lingual bridging of the mandibular foramen present on the left ramus of Oase 1 mandible) is unknown in modern humans predating Oase remains but is common in Neandertals and some of the later modern humans in Europe52–53. No archaeological industry was found at this important site, limiting our knowledge of these earliest anatomically modern humans in Europe to their anatomical features. New direct dating of the human remains from Mlade~ (Lautsch) in Moravia, Czech Republic54 suggests an age of around 31 kya for these anatomically modern humans55. Although the association with Aurignacian lithics was previously suggested, the exact nature of the deposition at the site is uncertain22 and while Mlade~ type bone points were found, the lithic material is scarce, and the split base bone points that are common in other Aurignacian-like industries of the earliest Central/Eastern European Upper Paleolithic are absent22. Therefore, the question of whether these tools represent an early Aurignacian-like (transitional) industry, or later Aurignacian sensu stricto, remains open. In addition, as in Oase sample, several archaic features are seen in some of


the Mlade~ specimens. These include occipital bunning in Mlade~ 3, 5 and 6, and robust supraorbital regions in Mlade~ 5 and 6, as well as large palatal and dental dimensions and some other anatomical details in the sample22,49,55–64, all features that are common in earlier Neandertal populations.

Industries of the Earliest Upper Paleolithic of Europe If we use the traditional approach based primarily on typology and technology in order to define Middle (Mousterian and its variants) vs. Upper Paleolithic industries in Europe, we face the problem of several so-called 'transitional’ industries. These include the Châtelperronian of France and northern Spain, Szeletian and Jankovichian of central and parts of eastern Europe, Uluzzian of Italy (Tuscany, Calabria, southern Adriatic part, Uluzzo Bay, etc.), Streletskian of eastern Europe, Jerzmanowician of eastern Germany and Poland, Althmülian of southern Germany, Bohunician of Czech Republic, Brynzeny and Kostenki Szeletian of Russia and several other unnamed or site-specific assemblages from Poland, Slovakia, Czech Republic, Romania, etc. in which various elements of Mousterian appear alongside the Upper Paleolithic types or types produced using technology commonly associated with the Upper Paleolithic. All these industries seem to have their origin in local Mousterian variants and no abrupt change can be seen22,65–97. Except for documented associations of Neandertal remains and Châtelperronian artifacts from La Roche à Pierrot at St. Cesaire and Grotte du Renne at Arcy-sur-Cure98–101 there are no diagnostic hominin fossils associated with any of these earliest Upper Paleolithic finds22,102–103. Thus, even if we accept the earliest Aurignacian as a single industrial complex that has its origins outside this area104 (both of these premises being far from proven) and attribute it to anatomically modern newcomers (for which there are no known hominin/industrial associations) we are left with the problem of who is responsible for these pre-Aurignacian assemblages. Typological thinking is responsible for the acceptance of the Aurignacian as a single widespread complex commonly associated with the spread of morphologically modern humans into Europe21,22. We believe that, in light of the currently available evidence (or the lack thereof) this view should be carefully reexamined. Simplification of this model can be summarized as follows: As more and more studies66–70,73,75,76,78,81,82,105 show that the earliest Upper Paleolithic (»transitional«) industries in Europe develop within the local framework from (and including various elements of) the Mousterian complex, the earliest distinctly Upper Paleolithic industry associated with anatomically modern humans should be Aurignacian, brought here as they move into the region106. Here authors vary in opinion on whether and how much influence modern newcomers and their culture had on the technological/behavioral change of late Neandertals. Thus, in this model, the Aurignacian is re-

garded as a single imported complex that can be recognized in the archaeological record by the appearance of certain tool types and automatically assigned to anatomically modern populations. While this sounds simple enough, it is not. First, detailed archaeological studies show that several tool types (especially bone tools) used as indicative of Aurignacian are in fact commonly found in various aforementioned »transitional« industries21,67,89,95,107,108. Further, the Early Aurignacian differs from the Late Aurignacian21. Finally, there are great differences between assemblages of typical Aurignacian from Western Europe, and that of Central/Eastern Europe15,21,90,108. All this makes it clear that there may be a different pattern of behavioral, and most likely, populational change in Western vs. Central/Eastern Europe. This is in agreement with several anatomical studies49,57,109. While this transition (whatever the mode of it) was more abrupt in Western Europe, evidence suggests a more gradual pattern for Central and Eastern areas of this region. Therefore, we believe it is quite likely that some Neandertal populations had a significant role in the formation of early modern European gene pool (via assimilation into anatomically more modern populations), while other Neandertal groups had none. As in the case of the initial Upper Paleolithic (aka »transitional«) industries, except for the Châtelperronian, makers of the earliest Aurignacian sensu lato are unknown as there is no clear association of diagnostic hominin and archaeological material. Although a new study and dating of an interstratified sequence of Châtelperronian and supposed Aurignacian suggests coexistence of these industries at least in some sites110, determination of this industry as Aurignacian should be reexamined. As mentioned, tool types indicative of Aurignacian commonly appear in other transitional industries of Central Europe. Again, no association of human bones was found in these layers, therefore all we can say is that there are two contemporaneous yet somewhat different cultural traditions present at the site. One of these is known to be associated with late Neandertals. One more point concerning the appearance of the Aurignacian should be mentioned. Although its origins were commonly seen in the Middle Eastern assemblages of anatomically modern humans, some authors trace its initial rise in several independent centers in Europe111. This explanation makes more sense if the Aurignacian is not a single widespread complex but actually represents different Early Upper Paleolithic assemblages that share several tool types (previously considered to be indicative of a single industrial complex). In this light there is no need to see these industries as a product of a single population. This also raises possibilities of different explanations for shared similarities (trade, influence, population mixing, etc). However, we should bear in mind that population contacts differ in their pattern. Interbreeding and peaceful coexistence, trade, etc., might dominate some of these interactions, while in others patterns of contact might differ. Therefore, models based on data from Western Europe should not be used for Central/Eastern Europe. 461


The Middle to Upper Paleolithic Transition at Vindija and its Significance for the Modern Human-Peopling of Europe Vindija Cave has an important place in the understanding of the initial anatomically modern human peopling of Europe. The significance of the association of Neandertal remains with an Upper Paleolithic industry has been a subject of a considerable debate15,80,112–114, as it has been argued that the association is in fact artificial and the result of the cryoturbation that has been noted in some parts of the cave. A partial Neandertal mandible (Vi-207) found in direct association with the characteristic Upper Paleolithic tool type (a split-based bone point, Vi-3437) adds to the complexity of the picture. We contend that the arguments presented in favor of artificial mixing of these are weakened by careful consideration of data. It is true that the excavations at Vindija, in many ways, followed techniques that had already been abandoned in Paleolithic archaeology at that time in most of Europe (especially in France where the past mistakes of numerous excavations during the early part of the 20th century led to deeper understanding of the importance of careful and detailed collecting and documenting of finds and features). However, such arguments can only go so far. Practice of selective collecting of »more important« or bigger and diagnostic finds of recognizable importance does not automatically cast a shadow on all of the data. While important data was lost, resulting from non-collecting or selective collecting of items (such as debitage or smaller non-diagnostic fragmentary bones etc), the majority of recognizable tools, bones, bigger pieces of debitage etc. were collected and recorded according to stratigraphic units. Cryoturbation, while present at the site115, has not been noted for the part of the cave where the associated mandible and bone point have been found8,12,15,112. Further, G1 consists of characteristic reddish clay, easily recognizable and distinct from both upper and lower parts of the sequence. This reddish clay was embedded in both

Vi-207 mandible and the Vi-3437 bone point and can still be observed on another massive bone point from this stratigraphic layer. In a recent paper, Ahern and colleagues12 reported additional Neandertal remains, one of which (a proximal radial shaft Vi 13.8) has embedded reddish clay sediment that is characteristic of layer G1. Neandertal attribution of this specimen12 is suggested by the strong curvature of the shaft and the medial orientation of radial tuberosity116,117. The presence of further Neandertal specimens from layer G1 additionally disproves the claim for artificial mixing of layers and arguments against the Neandertal association with the G1 Upper Paleolithic industry. There is an interesting pattern when we compare archaeological assemblages of various Vindija layers. In older layers (unit K) typical Mousterian tools predominate and there is a clear evidence of the use of Levallois technology that is common in most European Mousterian assemblages. The most abundant raw material in unit K is local quartz16,118, and flake technology predominates in tool production. Level G3 presents a mixture of typical Mousterian tools, such as sidescrapers, but there are also Upper Paleolithic types of stone tools (such as endscrapers), and alongside flake technology, bifacial and blade technology was used in production of tools from this layer. It is important to note that no evidence of Levallois technology is seen in layer G3 of Vindija15. There is also evidence of more selective use of raw material, as there are more tools on chert in this layer12,16,118. The level G1 assemblage shows an even more pronounced shift towards the use of higher quality raw material (i.e., chert) compared to the older layers of the site, and there are no tools made on quartz12,16,118. Upper Paleolithic elements in stone tools are more abundant than in layer G3, and bone points from G1 layer represent a new distinctly Upper Paleolithic element that is not seen in any of the older layers13–15. At several Slovenian sites, such as Divje Babe I and Mokri{ka Jama, bone tools similar to those of Vindija have also been found119–120. Similarly »Aurignacian« as-

Fig. 6: Comparison of Vindija 202 (left) and Krapina 4 (right) frontal bones (photo: J.C.M. Ahern).

462


semblage of Poto~ka Zijalka also differs in pattern from the »classical Aurignacian« assemblages120. In fact, this assemblage was previously referred to as Olschewian121. All hominins from the Vindija G complex can be recognized as a part of Neandertal populations on the basis of their overall gestalt. However, most of the commonly noted »Neandertal features« (for a detailed list see64,122–129 and references therein) do not represent autapomorphies, but are instead either plesiomorphic characters inherited from preceding archaic hominins or shared with contemporary and/or post-Neandertal populations60. It is clear that there are many temporal and geographic differences. Several studies have shown that later Neandertals differ in morphological details from earlier »classic« members of this population, for instance in the reduction of facial dimensions and projection8,9,12,49,56,130–131 as well as in other details of their anatomy. This is true for the Vindija G1 Neandertals, as shown by several studies, especially on the supraorbital and mandibular material56,130,132. Analyses reveal the intermediate position of the Vindija supraorbitals, both in projection and shape compared to the older Krapina sample (Figure 6). The Vindija supraorbital tori have relatively greater degrees of pinching above the orbits compared to the earlier Neandertals49,130,133,134. Recent study of a newly reconstructed partial cranial vault from G3 level comprised of supraorbital and frontal fragments (Vi 284, Vi 230, Vi 255, Vi 256) again suggests anatomical change in the direction of anatomically more modern morphology12. Change in the direction toward a more modern human pattern is also seen in the Vindija mandibular sample, suggesting facial reduction, and the Vindija mandibles have more vertical symphyses than earlier Neandertals and exhibit incipient eminences, though not a true modern human chin133,135–136. Observed gracility and change in shape is not due to body size109 or age and/or sex bias in the sample12,131,132,136 and could suggest gene exchange with anatomically modern populations. »Neandertal« traits are not present in earlier anatomically modern humans (samples predating 40 kya from Africa and Asia) that are the likely ancestors of Upper Paleolithic populations that came to Europe. Thus, the appearance of several »Neandertal» traits in the youngest modern groups in Europe (such as Mlade~ or Predmostí)49,56,60,63,137,138 and the later Gravettian child from Lagar Velho139 is easily explained by interbreeding and would best fit within the framework of the Assimilation model of modern human origins36,57–58.

The Impact of Molecular Data on the Modern Human Origins Debate After the field of genetics entered the modern human origins debate with the initial claims for exclusively African origins140, several authors emphasized that the results could be explained in different ways141–145. Moreover, mtDNA results do not seem to be in agreement with results obtained from other parts of genome146,147. Newer analyses of mtDNA isolated directly from Neandertal

bones added another dimension to the debate148–152. Although these sequences are different from those of living humans, various processes (e.g. bottlenecks, selection, drift, populational expansions etc.) could cloud our insight into the past events. Among these specimens, several Vindija fossils were included151–152 and were reported to fall outside both contemporary modern human, as well as Upper Paleolithic hominid ranges. However, ancient DNA was extracted from Vindija fossils that are both undiagnostic and of uncertain context (Vi 77, Vi 80, Vi 75). While a more meaningful insight into the question of whether or not Neandertals and anatomically modern humans interbred could be provided by extraction of DNA from the earliest modern humans in Europe152, alas, problems with extraction and contamination of ancient DNA, as well as with the small size of the available fossil sample of these crucial specimens makes it impossible to answer this question solely based on genetic evidence. In sum, some amount of interbreeding between these two late Pleistocene populations cannot be excluded and distinction of Neandertals at the species level is refuted by the current evidence58,141–143,147,148,151,152,154. Any molecular analysis dealing with the question of Neandertal and anatomically modern human interaction must take into account the complex pattern of population movements, population size, bottlenecks, etc. Even then, known problems such as small sample size and difficulties with extraction and contamination of DNA would make such analyses questionable. Until these questions are answered, the genetic picture drawn from both ancient DNA studies, as well as of models based on contemporary modern human genetic research allows for different explanations and should not be taken as a proof that no interbreeding between these populations took place.

Conclusion Vindija cave in Croatia has yielded the youngest securely dated Neandertal skeletal remains in Central/ Eastern Europe. In addition, these remains have been found in association with archaeological material exhibiting Upper Paleolithic elements. Due to its geographic location and date, the Vindija remains are particularly crucial for the understanding of the initial modern human peopling of Europe and the nature of the Neandertal demise. We argue that the association of an early Upper Paleolithic industry with late Neandertals at Vindija is not likely to be a result of artificial mixing of specimens from different strata, but rather that these artifacts are reasonably considered to be products of the Vindija Neandertals. Although similar archaeological samples in Europe have traditionally been regarded as Aurignacian and automatically assigned to anatomically modern humans, we believe that many of earliest Upper Paleolithic assemblages are in fact derived from the local Mousterian, and the question of which population is responsible for the production of these assemblages remains open. 463


The so-called transitional industries such as Uluzzian of Italy and Szeletian of Hungary and adjacent areas were quite likely a product of local Neandertal groups, as they have their origin in preceeding local Mousterian. In Europe at least, only Neandertals have been associated with Mousterian assemblages. Likewise, the only clear association of hominin remains and the Initial Upper Paleolithic thus far has been Neandertals with the Châtelperronian (at Arcy-sur-Cure and St. Cesaire98,100). Although it can be argued that the anatomically modern newcomers are the likely producers of the earlier distinctly Upper Paleolithic industry of Europe (later Aurignacian, or Aurignacian sensu stricto), this still remains to be proven. However if, as we argue, Aurignacian should no longer be considered a single Pan-European industrial complex, but rather represents a number of local early Upper Paleolithic assemblages, the association of Neandertals and Early Upper Paleolithic is not so surprising. The Upper Paleolithic industry at Vindija is not Aurignacian sensu stricto, but one of many »transitional» industry assemblages. This suggestion is supported by the presence of significant Mousterian types, one bifacial stone point typical of Szeletian, as well by significant differences in the assemblage compared to Western European sites21,90,112,155. While we cannot equal industry with biological populations, the simplest explanation would be that late Neandertals developed at least some of these »transitional» industries. Further, we should reexamine the Aurignacian sequence at various sites, especially in Central and Eastern Europe, and try to detect whether these are in fact Aurignacian sensu stricto, or another »transitional» industry. If the later proves to be the case, the association of the split-base bone point (and therefore the Upper Paleolithic sequence) and late Neandertals at Vindija should not come as a surprise at all. The first modern people to come to Europe might have been small groups and it is unclear how much they contributed to the later modern human groups (e.g.

Gravettians etc.). Therefore we must bear in mind that it is not only the issue of Neandertal genetic contribution to the initial anatomically modern newcomers, but also the relation of these first groups to the later modern humans that needs to be taken into account. Unfortunately the relatively short time frame of the populational overlap between late Neandertals and early moderns, possible differential site use, and numerous factors, including sedimentation rates, preservation of the sediment which is erroding more quickly than forming differences in site use, etc., will result in rare preservation of such evidence. Therefore, the Vindija G1 layer is a rare and important find. Anthropological analyses demonstrate that the late Neandertals at Vindija exhibit a more modern pattern of morphology compared to most other European Neandertals. We believe that both the anatomical and archaeological characteristics of Vindija are best explained by the Assimilation model of modern human origins. The studies on the Vindija cave anthropological, archaeological and paleontological material is by no means over. New dating, DNA and various other skeletal analyses, as well as the recently published newly recognized hominids allow for a better insight into the human evolutionary past. There are many questions still to be answered and still more to be created by these answers. No doubt the material from the Vindija Cave will have a crucial part in answering some of them.

Acknowledgements Authors would like to thank the Ministry of science, education and sports of the Republic of Croatia, the Fulbright foundation, and the University of Wyoming for their financial support over the years. We would also like to thank the SABRE Foundation Croatia, Dr. Helena Pavi}, Dr. Arthur Durband, Dr. Preston T. Miracle, Adam Foster and Matt Kesterke.

REFERENCES 1. MALEZ, M.: Nalazi{ta paleolitskog i mezolitskog doba u Hrvatskoj. In: BENAC, A., (Ed.): Praistorija Jugoslavenskih Zemalja I: Paleolitsko i Mezolitsko Doba. In Croat. (Svjetlost, Sarajevo, 1979). — 2. HIRC, D.: Vindija. In: KLAI], V. (Ed.): Prirodni zemljopis Hrvatske. (C. Albrecht, Zagreb, 1878). — 3. VUKOVI], S.: Istraìvanje prethistorijskog nalazi{ta u spilji Vindiji kod Vo}e. In Croat. (Spomenica vara`dinskog muzeja, Vara`din, 1935). — 4. VUKOVI], S., Hist. Zborn., 2 (1949) 243. — 5. VUKOVI], S., Hist. Zborn., 3 (1950) 241. — 6. MALEZ, M.: Razvoj kvartara, fosilnog ~ovjeka i njegovih materijalnih kultura na tlu Sjeverne Hrvatske. Posebni otisak iz knjige »Vara`dinski Zbornik». In Croat. (JAZU. Vara`din, 1983). — 7. MALEZ. M., F. H. SMITH, J. RADOVÎ], D. RUKAVINA, D., Curr. Anthropol., 21 (1980) 365. — 8. WOLPOFF, M. H., F. H. SMITH, M. MALEZ, J. RADOVÎ], D. RUKAVINA, Am. J. Phys. Anthropol., 54 (1981) 499. — 9. SMITH, F. H., D. C. BOYD, M. MALEZ, M., Am. J. Phys. Anthropol., 68 (1985) 375. — 10. AHERN, J. C., F. H. SMITH, Am. J. Phys. Anthropol., 16 Suppl (1993) 47. — 11. SMITH, F. H., J. C. AHERN, Am. J. Phys. Anthropol., 93 (1994) 275. — 12. AHERN, J. C. M., I. KARAVANI], M. PAUNOVI], I. JANKOVI], F. H. SMITH, J. Hum. Evol., 46 (2004) 27. — 13. KARAVANI], I., Opuscula Archaeol., 17 (1993) 53. — 14. KARAVANI], I., J. Anthropol. Res., 51 (1995) 223. — 15. KARAVANI], I., F. H. SMITH, J. Hum. Evol., 34 (1998) 223. — 16. BLASER,

464

F., D. KURTANJEK, M. PAUNOVI], L’Anthropologie, 106 (2002) 387. — 17. MIRACLE, P., Rad HAZU, 458 (1991) 193. — 18. BRAJKOVI], D., Korelacija tafodema skupine ungulata iz gornjopleistocenskih sedimenata {pilja: Vindija, Velika pe}ina i Veternica u sjeveriozapadnoj Hrvatskoj. PhD Thesis. In Croat. (University of Zagreb, Zagreb, 2005). — 19. MALEZ, M., D. RUKAVINA, Rad JAZU, 383 (1979) 187. — 20. PAUNOVI], M., G. JAMBRE[I], D. BRAJKOVI], V. MALEZ, J. MAUCH LENARDI], Acta Geol., 26 (2001) 27. — 21. MIRACLE, P. T.: The spread of modernity in Europe. In: OMOTO, K., P. TOBIAS (Eds.): The Origins and Past of Modern Humans. Toward Reconciliation. Recent Advances in Human Biology 3. (World Scientific, Singapore, 1998). — 22. CHURCHILL, S. E., F. H. SMITH, Yrbk. Phys. Anthropol., 43 (2000) 61. — 23. SVOBODA, J., 2001., Mlade~ and other caves in the Middle Danube region: early modern humans, late Neandertals, and projectiles. In: ZILHÃO, J., T. AUBRY, A. F. CARVALHO (Eds.): Les premiers hommes modernes de la Péninsule Ibérique. (Actes du colloque de la Commission VIII de lÚISPP, Lisabon, 2001). — 24. MALEZ, M., H. ULLRICH, Paleontol. Jugosl., 29 (1982) 1. — 25. MALEZ, V., Radovi Zavoda znan. rad JAZU, 2 (1988) 31. — 26. PAUNOVI], M., F. SMITH, F., J. Hum. Evol., 42 (2002) A27. — 27. MALEZ, M., A. [IMUNI], A. [IMUNI], Rad JAZU, 411 (1984) 231. — 28. DRAXLER, I., Rad JAZU, 424 (1986) 275. — 29. AITKEN, M. J., C. B.

I. Jankovi} et al.: Vindija Cave and The Modern Human Peopling of Europe, Coll. Antropol. 30 (2006) 3: 457–466 STRINGER, P. A. MELLARS (Eds.): The origin of modern humans and the impact of chronometric dating. (Princeton University Press, Princeton, New Jersey, 1993). — 30. SMITH, F. H., E. TRINKAUS, P. B. PETTIT, I. KARAVANI], M. PAUNOVI], Proc. Natl. Acad. Sci. USA, 96 (1999) 12281. — 31. CONRAD, N. J., P. M. GROOTES, F. H. SMITH, F. H., Nature, 430 (2004) 198. — 32. WILD, E. M., M. PAUNOVI], G. RABEDER, I. STEFFAN, P. STEIER, Radiocarbon, 43 (2001) 1021. — 33. HIGHAM, T., C. BRONK RAMSEY, I. KARAVANI], F. H. SMITH, E. TRINKAUS, Proc. Natl. Acad. Sci., 103 (2006) 553. — 34. MELLARS, P., Nature, 439 (2006) 931. — 35. KOZLOWSKI, J. K. Cultural context of the last Neandertals and early modern humans in the Central-Eastern Europe. In: BAR-YOSEF, O., L. L. CAVALLI-SFORZA, R. J. MARCH, M. PIPERNO (Eds.): The Lower and Middle Paleolithic. (International Union of Prehistoric and Protohistoric Science, Forli, 1996). — 36. TRINKAUS, E., Ann. Rev. Anthropol., 34 (2005) 207. — 37. KEITH, A., Trans. Proc. Torq. Nat. Hist. Soc., 5 (1927) 1. — 38. GARROD, D. A. E., The Upper Paleolithic in Britain. (Oxford University Press, Oxford, 1926). — 39. OAKLEY, K. P., B. G. CAMPBELL, T. I. MOLLESON, Catalogue of fossil hominids. Part II: Europe. (British Museum, London, 1971). — 40. HEDGES, R. E. M., R. A. HOUSLEY, I. A. LAW, C. R. BRONK, Archaeometry, 31 (1989) 207. — 41. STRINGER, C. B., British Isles. In: ORBAN, R. (Ed.): Hominid remains: An update. British Isles and Eastern Germany. (Univ. Libre Bruxelles, Bruxelles, 1990). — 42. JACOBI, R. M., T. F. G. HIGHAM, C. BRONK RAMSEY, J. Quat. Sci. (in press). — 43. HENRY GAMBIER, D., B. MAUREILLE, B., R. WHITE, Bull. Mém. Soc. Anthropol. Paris, 16 (2004) 49. — 44. BAILEY, S. E., J. J. HUBLIN, Bull. Mém. Soc. Anthropol. Paris, 17 (2005) 115. — 45. DUJARDIN, V., Antiq. Natl., 33 (2003) 231. — 46. RICHARDS, M. P., P. B. PETTITT, M. C., STINER, E. TRINKAUS, Proc. Natl. Acad. Sci. USA, 98 (2001) 6528. — 47. OLARIU, A., G. SKOG, R. HELLBORG, K. STENSTRÖM, M. FAARINEN, P. PERSSON, Report Wp1 IDRANAP (2004). — 48. NICOLAESCU-POPSLOR, D., 7th Int Cong. Anthropol. Ethnol. Sci. Moscow, 3 (1968) 381. — 49. SMITH, F. H.: Fossil hominids from the Upper Pleistocene of Central Europe and the origin of modern Europeans. In: SMITH, F. H., F. SPENCER (Eds.): The Origins of Modern Humans: A World Survey of the Fossil Evidence. (Alan. R. Liss. New York, 1984). — 50. PAUNESCU, A., A Paleoliticul si Mezoliticul din Spatiul Transilvan 231. (Editura AGIR, Bucuresti, 2001). — 51. RAINER, F., I. SIMIONESCU, An. Acad. Rom., SIII. TXVIII (1942) 489. — 52. TRINKAUS, E., S. MILOTA, R. RODRIGO, M. GHERASE, O. MOLDOVAN, J. Hum. Evol., 45 (2003) 245. — 53. TRINKAUS, E., O. MOLDOVAN, S. MILOTA, A. BÎLGAR, L. SARCINA, S. ATHREYA, S. BAILEY, R. RODRIGO, G. MIRCEA, T. HIGHAM, C. BRONK RAMSEY, J. VAN DER PLICHT, Proc. Natl. Acad. Sci. USA, 100 (2003) 11231. — 54. SZOMBATHY, J., Die Eiszeit, 2 (1925) 1., 73. — 55. WILD, E. M, M. TESCHLER-NICOLA, W. KUTSCHERA, P. STEIER, E. TRINKAUS, W. WANEK, Nature, 435 (2005) 332. — 56. SMITH, F. H., Curr. Anthropol., 23 (1982) 667. — 57. SMITH, F. H., A. B. FALSETTI, S. M. DONNELLY, Yrbk. Phys. Anthropol., 32 (1989) 35. — 58. SMITH, F. H., I. JANKOVI], I. KARAVANI], Quatern. Intern., 137 (2005) 7. — 59. FRAYER, D. W.: Cranial variation at Mlade~ and the relationship between Mousterian and Upper Paleolithic hominids. In: NOVOTNY, V. V., A. MIZEROVÁ, (Eds.): Fossil Man. New Facts, New Ideas. Papers in honor of Jan Jelínek's life anniversary. (Anthropos, Brno, 1986). — 60. FRAYER, D. W.: The persistence of Neanderthal features in post-Neanderthal Europeans. In: BRÄUER, G., F. H. SMITH (Eds): Continuity or replacement: controversies in Homo sapiens evolution. (A. A. Balkema, Rotterdam, 1992). — 61. FRAYER, D., Perspectives on Neanderthals as ancestors. In: CLARK, G. A., C. M. Willermet (Eds): Conceptual issues in modern human origins research. (Aldine De Gruyter, NewYork, 1997). — 62. FRAYER, D. W., M. H. WOLPOFF, F. H., SMITH, A. G. THORNE, G. G. POPE, Am. Anthropol., 95 (1993) 14. — 63. KIDDER, J., R. JANTZ, F. H. SMITH: Defining modern humans: a multivariate approach. In: BRÄUER, G., F. H. SMITH (Eds): Continuity or replacement: controversies in Homo sapiens evolution. (A. A. Balkema, Rotterdam, 1992). — 64. WOLPOFF, M. H., Paleoanthropology. (McGraw Hill, New York, 1999, 2nd. ed.). — 65. HARROLD, F., Ampurias, 43 (1981) 35. — 66. HARROLD, F. B., Mousterian, Châtelperronian and early Aurignacian in Western Europe: continuity or discontinuity? In: MELLARS, P., C. STRINGER (Eds.): The human revolution: behavioural and biocultural perspectives on the origin of modern humans. (Princeton University Press, Princeton: New Jersey, 1989). — 67. ALLSWORTH-JONES, P., The Szeletian: main trends, recent results, and problems for resolution. In: DAY, M., R. FOLEY, W. RUKANG (Eds): The Pleistocene Perspective. (Papers of the World Archaeological Congress, Southampton, 1986). — 68. ALLSWORTH-JONES, P., The Szeletian and stratigraphic succession in Central Europe and adjacent areas: Main trends. recent results and problems for resolution. In: MELLARS, P., (Ed.): The emergence of modern hu-

mans: An archaeological perspective. (Cornell University Press, Ithica, 1990). — 69. ANIKOVICH, M., J. World Prehist., 6 (1992) 205. — 70. GIOIA, P., Problems related to the origins of Italian Upper Paleolithic: Uluzzian and Aurignacian. In: KOZLOWSKI, J. K. (Ed): La Mutation. (Etudes et Recherches Archeologiques de l’Universite de Liege, Liege, 1988). — 71. BORDES, F., Du Paléolithique moyen au paléolithique supérieur — continuité ou discontinuité? In: BORDES, F. (Ed.): The origin of Homo sapiens. (UNESCO, Paris, 1972). — 72. BORDES, F.: A Tale of Two Caves. (New York, Harper & Row, 1972). — 73. CLARK, G. A., J. M. LINDLY, The case of continuity: Observations on the biocultural transition in Europe and Western Asia. In: MELLARS, P., C. STRINGER (Eds.): The human revolution: behavioural and biocultural perspectives on the origin of modern humans. (Princeton University Press, Princeton: New Jersey, 1989). — 74. CLARK, G. A., Amer. Anthropol., 104 (2002) 50. — 75. GOLOVANOVA, L. V., V. B. DORONICHEV, J. World Prehist., 17 (2003) 71. — 76. RIGAUD, J.-P., From the Middle to the Upper Paleolithic: Transition or convergence. In: TRINKAUS, E. (Ed): The Emergence of Modern Humans: Biocultural Adaptations in the Later Pleistocene. (Cambridge University Press, Cambridge, 1989). — 77. RIGAUD, J.-P., Scenarios for the Middle to Upper Paleolithic tranisition. In: CLARK, G. A., C. M. Willermet (Eds): Conceptual issues in modern human origins research. (Aldine De Gruyter, NewYork, 1997). — 78. PRADEL, L., Curr. Anthropol., 7 (1966) 33. — 79. D'ERRICO, F., P. VILLA, A. PINTO, R. RUIZ IDARRAGA, Antiquity, 72 (1998) 65. — 80. D’ERRICO, F., J. ZILHÃO, M., JULIEN, D. BAFFIER, J. PELGRIN, Curr. Anthropol., 39 (1998) 51. — 81. CABRERA VALDÉS, V., M. HOYOS GÓMEZ, M., F. B. DEQUIRÓS, The transition from the middle to the Upper Paleolithic in the cave of El Castilo (Cantabria, Spain). In: CLARK, G. A., C. M. WILLERMET (Eds): Conceptual issues in modern human origins research. (Aldine De Gruyter, New York, 1997). — 82. STRAUS, L. G., The Iberian situation between 40,000 and 30,000 years B.P. in light of European models of migration and convergence. In: CLARK, G. A., C. M. WILLERMET (Eds): Conceptual issues in modern human origins research. (Aldine De Gruyter, NewYork, 1997). — 83. PALMA DI CESNOLA, A., Riv. Sci. Prehist., 20 (1965) 33. — 84. PALMA DI CESNOLA, A., Riv. Sci. Prehist., 21 (1966) 3. — 85. KOZLOWSKI, J., S. KOZLOWSKI: Upper Paleolithic and Mesolithic in Europe: Taxonomy and paleohistory. (Polska Akademia Nauk, Wroclaw, 1979). — 86. KOZLOWSKI, J. K., Early Upper Paleolithic backed blade industries in Central and Eastern Europe. In: BRANTINGHAM, P. J., S. L. KUHN, K. W. KERRY (Eds): The Early Upper Paleolithic beyond Western Europe. (University of California Press, Berkeley, 2004). — 87. LAPLACE, G., Récherches sur l'origine et l'evolution des complexes leptolithiques. (De Bochard, Paris, 1966). — 88. LEROI-GOURHAN, A., Bull. Soc. Méridion. Spéleol. Préhist., 6 (1963) 75. — 89. SVOBODA, J., The complex origin of the Upper Paleolithic in the Czech and Slovak Republics. In: KNECHT, H., A. PIKE-TAY, R. WHITE (Eds): Before Lascaux. The Complex Record of the Early Upper Paleolithic. (CRC Press, Boca Raton, 1993). — 90. SVOBODA, J. A., Continuities, discontinuities, and interactions in Early Upper Paleolithic technologies. A view from the Middle Danube. In: BRANTINGHAM, P. J., S. L. KUHN, K. W. KERRY (Eds): The Early Upper Paleolithic beyond Western Europe (University of California Press, Berkeley, 2004). — 91. OTTE, M., From Middle to the Upper Paleolithic: the nature of the transition. In: MELLARS, P. (Ed): The Emergence of Modern Humans. An Archaeological Perspective. (Cornell University Press, Ithica, 1990). — 92. OLIVA, M., Archeolog. Rozhledy, 13 (1980) 48. — 93. SKUTÍL, J., Bratislava, 2/1 (1928) 1966. — 94. PRO[EK, F., Slovenská Archeol. 1 (1953) 133. — 95. VALOCH, K., Casopis Moravské Muzea Sc. Soc., 51 (1966) 5. — 96. VALOCH, K., Rapports entre Le Paléolithique Moyen et le Paléolithique Supérieur en Europe Centrale. In: BORDES, F. (Ed): The Origins of Homo sapiens. (UNESCO, Paris, 1972). — 97. KLÍMA, B., Archeolog. Rozhledy, 13 (1961) 84. — 98. LÉVÊQUE, F., B. VANDERMEERSCH, R. C., Acad. Sci., 291 (1980) 187. — 99. HEDGES, R. E. M., R. A. HOUSLEY, C. BRONK-RAMSEY C., G. J. VAN KLINKEN, Archaeometry, 36 (1994) 337. — 100. HUBLIN, J. J., F. SPOOR, M. BRAUN, F. ZONNENVELD, S. CONDEMI, Nature, 381 (1996) 224. — 101. LEROI-GOURHAN, A., Annal. Paléontol., 44 (1958) 87. — 102. GAMBIER, D., Fossil hominids from the Early Upper Paleolithic (Aurignacian) of France. In: MELLARS, P., C. STRINGER (Eds.): The human revolution: behavioural and biocultural perspectives on the origin of modern humans. (Princeton University Press, Princeton, New Jersey, 1989). — 103. GAMBIER, D., Modern humans at the beginning of the Upper Paleolithic in France. In: CLARK, G. A., C. M. Willermet (Eds): Conceptual issues in modern human origins research. (Aldine De Gruyter, NewYork, 1997). — 104. KOZLOWSKI, J. K., S. K. KOZLOWSKI: Praziejie Europy od XL do IV tysiaclecia p.n.e. (Panstwowe wydawniciwo naukowe, Warsaw, 1975). — 105. KOZLOWSKI, J. K., A multiaspectual approach to the origins of the Upper Paleolithic in Europe. In: MEL-

465

I. Jankovi} et al.: Vindija Cave and The Modern Human Peopling of Europe, Coll. Antropol. 30 (2006) 3: 457–466 LARS, P. (Ed): The Emergence of Modern Humans. An Archaeological Perspective (Cornell University Press, Ithica, 1990). — 106. MELLARS, P., Technological changes across the Middle-Upper Paleolithic transition: Economic, social and cognitive perspectives. In: MELLARS, P., C. STRINGER (Eds.): The human revolution: behavioural and biocultural perspectives on the origin of modern humans. (Princeton University Press, Princeton, New Jersey, 1989). — 107. HILLEBRAND, E., Eiszeit und Urgescichte, 5 (1928) 99. — 108. OLIVA, M., The Aurignacian in Moravia. In: KNECHT, H., A. PIKE-TAY, R. WHITE (Eds): Before Lascaux. The Complex Record of the Early Upper Paleolithicm, (CRC Press, Boca Raton, 1993). — 109. TRINKAUS, E., F. H. SMITH, J. Hum. Evol., 28 (1995) 201. — 110. GRAVINA, B., P. MELLARS, C. BRONK RAMSEY, Nature, 438 (2005) 51. — 111. OLIVA, M., Anthropologie, 27 (1989) 251. — 112. KARAVANI], I., F. H. SMITH, Curr. Anthropol., 41 (2000) 838. — 113. ZILHÃO, J., F. D'ERRICO, Curr. Anthropol., 40 (1999) 355. — 114. STRAUS, L. G., Curr. Anthropol., 40 (1999) 352. — 115. MALEZ, M., D. RUKAVINA, Rad JAZU, 371 (1975) 245. — 116. TRINKAUS, E., S. E. CHURCHILL, Am. J. Phys. Anthropol., 75 (1988) 15. — 117. CHURCHILL, S. E., Human Upper Body Evolution in the Eurasian Later Pleistocene. PhD Thesis. (University of New Mexico, Albuquerque, 1994). — 118. KURTANJEK, D., V. MARCI, Rad Jugosl. Akad. Znan. Umjetn., 449 (1990) 227. — 119. TURK, I., B. KAVUR, Survey and description of Paleolithic tools, fireplaces, and hearths. In: TURK, I. (Ed.): Divje Babe I Cave Site in Slovenia. (Operni inst. Archaeol. Sloven. 2, Znanstvenoraziskovalni Centar SAZU, Ljubljana, 1997). — 120. BRODAR, M., F. OSOLE, Paleolitske i mezolitske regije i kulture u Sloveniji. In: BENAC, A. (Ed.): Praistorije jugoslavenskih zemalja 1. (Svjetlost, Sarajevo, 1979). — 121. BAYER, J., Eiszeit und Urgesch, 6 (1929) 83. — 122. TRINKAUS, E., The Shanidar Neandertals (Academic Press, London, 1983). — 123. TRINKAUS, E., The evolutionary origins of Neandertals or, why were there Neandertals? In: TRINKAUS, E. (Ed): L’Homme de Neandertal, Vol. 3: L’ Anatomie. (Etudes et Recerches Archéologiques de l’ Université de Liège, Liège, 1988). — 124. TRINKAUS, E., The Upper Pleistocene transition. In: TRINKAUS, E. (Ed): The Emergence of Modern Humans: Biocultural Adaptations in the Later Pleistocene. (Cambridge University Press, Cambridge, 1989). — 125. GORJANOVI]-KRAMBERGER, D.: Der Diluvijale Mensch von Krapina in Kroatien. Ein Bertrag zur Paläoanthropologie. (Kreidel, Wiesbaden, 1906). — 126. SMITH, F. H., F. SPENCER (Eds): The Origin of Modern Humans: A World Survey of the Fossil Evidence. (Alan R. Liss, New York, 1984). — 127. STRINGER, C. B., C. GAMBLE: In Search of the Neanderthals. (Thames and Hudson, London, 1993). — 128. NITECKI, M. H., D. V. NITECKI (Eds): Origins of Anatomically Modern Humans. (Plenum Press, New York, 1993). — 129. JANKOVI], I., Coll. Antropol., 28 Suppl. 2 (2004) 379. — 130. SMITH, F. H., G. C. RANYARD, G. C., Am. J. Phys. Anthropol., 53 (1980) 589. — 131.

AHERN, J. C. M., S.-H. LEE, J. D. HAWKS, J. D., J. Hum. Evol., 43 (2002) 419. — 132. AHERN, J. C. M., Late Pleistocene frontals of the Hrvatsko Zagorje: an analysis of intrapopulational variation among south central European Neandertals. PhD Thesis. (University of Michigan, 1998). — 133. SMITH, F. H.: Samples, species and speculations in the study of modern human origins. In: NITECKI, M. H., D. V. NITECKI (Eds.): Origins of Anatomically Modern Humans. (Plenum Press, New York, 1994). — 134. SMITH, F. H., J. F. SIMEK, M. S. HARRILL, Geographic variation in supraorbital torus reduction during the Later Pleistocene. (c. 80000– 15000 B.P.). In: MELLARS, P., C. STRINGER (Eds.): The human revolution: behavioural and biocultural perspectives on the origin of modern humans. (Princeton University Press, Princeton, New Jersey, 1989). — 135. AHERN, J. C. M., F. H. SMITH, Homo, 55 (2004) 1. — 136. KESTERKE, M., J. C. M. AHERN, Coll. Antropol. (in press). — 137. TRINKAUS, E., M. LEMAY, Am. J. Phys. Anthropol., 57 (1982) 27. — 138. AHERN, J. C. M., Non-metric variation in recent humans as a model for understanding Neandertal-early modern human differences: Just how »unique« are Neandertal unique traits? In: HARVATI, K., T. HARRISON (Eds.): Neanderthals Revisited: New Approaches and Perspectives. (Kulwer, New York, 2006). — 139. DUARTE, C., J. MAURICIO, P. B. PETTITT, P. SOUTO, E. TRINKAUS, H. VAN DER PLICHT, J. ZILHÃO, Proc. Natl. Acad. Sci., 96 (1999) 7604. — 140. CANN, R., M. STONEKING, A. L. WILSON, Nature, 325 (1987) 31. — 141. RELETHFORD, J. H., L. B. JORDE, Am. J. Phys. Anthropol., 108 (1999) 251. — 142. RELETHFORD, J. H., Am J. Phys. Anthropol., 115 (2001) 95. — 143. RELETHFORD, J., Genetics and the earch for modern human origins. (Wiley, New York, 2001). — 144. HARPENDING, H., A. ROGERS, Annu. Rev. Genomics. Hum. Genet., 1 (2000) 361. — 145. HARPENDING, H., V. ESWARAN, Science, 309 (2005) 1995. — 146. TEMPLETON, A., Nature, 416 (2002) 45. — 147. ESWARAN, V., H. HARPENDING, A. R. ROGERS, J. Hum. Evol., 49 (2005) 1. — 148. KRINGS, M., A. STONE, R. W. SCHMITZ, H. KRAINITZKI, M. STONEKING, S. PÄÄBO, Cell, 90 (1997) 19. — 149. KRINGS, M., C. CAPELLI, F. TSCHENTSCHER, H. GEISERT, S. MEYER, A. VON HAESELER, K. GROSSSCHMIDT, G. POSSNERT, M. PAUNOVI], S. PÄÄBO, Nature Genet., 26 (2000) 144. — 150. OVCHINNIKOV, I. V., A. GÖTHERTRÖM, G. O. ROMANOVA, V. M. KHARITONOV, K. LIDÉN, W. GOODWIN, Nature, 404 (2000) 490. — 151. KRINGS, M., H. GEISERT, R. W. SCHMITZ, H. KRAINITZKI, S. PÄÄBO, Procl. Natl. Acad. Sci. USA, 96 (1999) 5581. — 152. SERRE, D., A. LANGANEY, M. CHECH, M. TESCHLER-NIKOLA, M. PAUNOVIC, P. MENNECIER, M. HOFREITER, G. POSSNERT, S. PÄÄBO, PLPS Biology, 2 (2004) 313. — 153. NORDBORG, M., Am. J. Hum. Genet., 63 (1998) 1240. — 154. GUTIÉRREZ, G., D. SÁNCHEZ, A. MARIN, Mol. Biol. Evol., 19 (2002) 1359. — 155. MONTET-WHITE, A.: Le Paléolithique en ancienne Yougoslavie. (Jérôme Millon, Grenoble, 1996).

I. Jankovi} Institute for anthropological research, Amru{eva 8, 10000 Zagreb, Croatia e-mail: [email protected]

[PILJA VINDIJA I DOLAZAK ANATOMSKI MODERNIH LJUDI NA PROSTORE EUROPE

SA@ETAK U gornjopleistocenskim sedimentima {pilje Vindije u SZ Hrvatskoj, u vertikalnoj sukcesiji s ostacima modernih ljudi, na|eni su ostaci najmla|ih neandertalaca sredi{nje i isto~ne Europe. Sedimentolo{ki, paleontolo{ki i arheolo{ki sadràji u autohtonim su startigrafskim relacijama s ostacima hominida, {to Vindiju odre|uje kao klju~no nalazi{te, koje se ve} nekoliko desetlje}a interdisciplinarno istraùje. U potrazi za rasvjetljavanjem interakcije i sukcesije neandertalskih i modernih hominidnih populacija, dosada{nje spoznaje prezentirane su u svjetlu suvremenih saznanja i teorijskih pomaka. Izneseni su novi podaci o faunskim asocijacijama pojedinih stratigrafskih ~lanova koji mijenjaju ranije spoznaje o prehrambenim resursima paleolitskih lovaca Vindije. Nagla{en je zna~aj novih kronolo{kih podataka u repozicioniranju nekih klju~nih gornjopleistocenskih nalaza i nalazi{ta Europe.

466


The Prehistoric Hillfort at Grad (Pelje{ac, Dalmatia) – Preliminary Results of Intensive Surface Survey Sta{o Forenbaher and Petra Raji} [ikanji} Institute for Anthropological Research, Zagreb, Croatia

ABSTRACT An intensive surface survey, covering an area of approximately 10.000 m2, was carried out at Grad, a hillfort in southern Dalmatia. Its aims were to obtain information relevant for determining the spatial extent of the site, its function(s), periods of occupation, degree of preservation, and potential for further investigation. Research strategy included mapping of the visible structural remains and systematic recovery of all surface finds. Majority of the collected finds are coarse Hellenistic ceramics followed by Bronze Age and Iron Age pottery. The central area of the hillfort was intensively used during the last few centuries B.C., while its origins can be traced back to the Bronze Age. The recorded structures and the recovered finds hint at a residential and defensive function of the site, while its central, elevated area may have been a focus of special activities. Key words: hillfort, Pelje{ac, Adriatic, Bronze Age/Iron Age pottery, Hellenistic ceramics, surface survey

Introduction Hillforts are among the most prominent man-made features of the late prehistoric landscape of the eastern Adriatic and its hinterland. Locally known as gradine, the majority of those monumental structures date from the last two millennia B.C. (Bronze Age and Iron Age). Strategically located on hilltops or other elevated locations, typically they consist of an oval area enclosed by a drystone wall or multiple concentric drystone ramparts, except in places where the steep terrain by itself provides ample protection. Traditionally, the eastern Adriatic hillforts were considered as remains of fortified settlements1,2. A few decades ago, however, several authors3–5 noted the great variability in their size and shape, the kind and extent of their enclosing structures, as well as the kind and quantity of domestic and other debris that they contained. They interpreted this as evidence for functional variability and proposed that, aside from fortified settlements, hillforts may have served as refugia, cattle corrals, elite residences, or ritual foci. More recently, the simplistic »fortified settlement« concept was further challenged by Gaffney et al.6,7 who propose that many of these enclo-

sures may be better interpreted as public monuments, associated with control of land through common rituals, and reflecting the power of the local potentates. It is clear that many different kinds of hillforts were constructed and maintained during the periods in question. The main reason why we still know so little about them, in spite of their great number and conspicuous presence in the landscape, is that only a very few, such as Monkodonja8, Varvara9 or O{ani}i10 have been extensively excavated. The main reason for this is the high cost of large-scale excavations, which require long-time commitment, as well as conservation of the recovered structural remains. On the other hand, small-scale test excavations, carried out on numerous hillforts, sometimes produce valuable results, but they can yield only limited information about a few selected spots within those large and often complex sites. Intensive surface survey, consisting of systematic recovery of removable finds, as well as mapping of structural remains and other features visible at the surface, is an alternative, low-budget approach that can easily cover

Received for publication August 28, 2006

467

S. Forenbaher and P. Raji} [ikanji}: Hillfort at Grad Surface Survey, Coll. Antropol. 30 (2006) 3: 467–473

the entire site. Regardless of it many shortcomings, such survey can be a serious analytical tool when used judiciousy11–13. In the eastern Adriatic region, this kind of survey was attempted only on a few hillforts that were investigated in the course of the »Adriatic Islands Project«14. We present here the preliminary results of another such investigation, carried out recently on a hillfort in southern Dalmatia. For our case-study we chose Grad, a major hillfort located near the western tip of Pelje{ac peninsula. Our choice was guided by the fact that the micro-region around Grad is well covered archaeologically. The area has been a focus of a long-term field project that included general surface survey and mapping, as well as excavation of another major prehistoric site, Nakovana Cave15. Data from that survey offer comprehensive information about over a hundred prehistoric sites in Grad’s immediate neighborhood16, while the deep excavation trench at Nakovana Cave provides reliable temporal controls17,18. Centrally located within this microregion and dominating the relatively fertile Nakovana plateau, the hillfort clearly represents its most prominent prehistoric site. Grad is a natural fortress, located on an almost impregnable rocky hilltop that is surrounded on all sides by vertical or overhanging cliffs some 20 meters high. The

only approach to its high plateau is from the southwest, where scant traces remain of a drystone enclosure wall. From there, one can easily ascend to its barren, rocky summit (Figure 1). Grad has been mentioned in scientific reports since the late 19th century19, and was described as an important prehistoric site by Nik{a Petri}20.

Methodology and Techniques Applied Our intensive survey of Grad, carried out in May 2005, had multiple aims. The immediate ones were to roughly determine the spatial extent of the site and the periods during which it had been occupied. We also hoped to gain an initial insight into its function, estimate the degree of its preservation, and identify potential dangers posed to it by natural elements and human agency. Our further intention was to gather information that would allow a reliable estimate of the site’s potential for future research, and could serve as a secure base for planning a more extensive field investigation. We covered only the central area of the hillfort, its high plateau bounded by cliffs, which extends over an area of approximately 10.000 m2. Our research strategy required mapping of all visible structural remains and systematic collection of all surface finds. In order to ac-

Fig. 1. View of the Grad hillfort, rising above the Nakovana plateau.

468


complish this, a 10 x 10 meters square grid was laid out over the area to be surveyed, using compass and measuring tapes, and taking into account the ground slope wherever necessary. Each 10 x 10 meter square was documented individually on a standardized recording sheet prepared for the purpose. We estimated and recorded relative proportions of soil and bedrock exposed at the surface, the relative area covered by vegetation, and kinds of vegetation present. Ground visibility, which critically influences survey data, was estimated and recorded on ordinal scale, ranging from 0 (no visibility) to 5 (excellent visibility). A rough plan of each square, drawn to a scale of 1:100, included all structural remains and other features, together with their descriptions. This served as the base for the composite plan of the site, which was produced cumulatively in the field (Figure 2). All removable artifacts visible at the surface were collected, bagged by class (pottery, lithics, other), and tagged for later laboratory treatment, and their presence/ absence was noted on the recording sheet. Majority of the finds were small, heavily weathered pottery fragments, damaged by long exposure at the surface. A total of 5461 potsherds, weighing almost 60 kilograms, made up by far the largest of all the recovered artifact assemblages. Since most of them were non-diagnostic due to fragmentation and weathering, they were classified into four rough technological categories: (1) sherds of Bronze Age/Iron Age hand-made vessels; (2) sherds of thick-walled hand-made vessels with a characteristic wall section (red-gray-red); (3) sherds of Hellenis-

tic wheel-made vessels; and (4) glazed sherds. Each category was counted and weighed by the square, and the data were used as the base for further analysis of their spatial distribution across the site. As expected, distribution diagrams plotted from raw data revealed that ground visibility had a major impact on the apparent distribution of finds. A linear and an exponential correction for visibility were therefore applied, the latter apparently producing more realistic plots. Three of those plots are reproduced in Figures 3–5, which are further discussed below.

Results of the Survey The highest, eastern part of the hillfort, and the western end of the high plateau, are eroded to bedrock in most places, while its central part is covered by soil. This area was under cultivation in recent historic times, but the fields are now abandoned, and there seems to be little danger that modern agricultural activities might damage the site. Concentration of finds and structural remains is highest in the central area, and there is a distinct possibility that intact archaeological deposits are preserved underneath the plowzone. The site is being actively eroded away, as indicated by denuded areas along all cliff edges, which are strewn by numerous potsherds washed out form the sediment. Earlier researchers20 noted that the site was not confined to the high plateau. In the course of our fieldwork, it became abundantly clear that the artifact scatters extended well beyond it and encompassed large areas on

Fig. 2. Plan of the central area of the Grad hillfort.

469


Grad’s slopes. Since our systematic survey was not extended over the high plateau’s edges, we could not identify the full spatial extent of the site.

Periods of Occupation The nine flaked stone artifacts and a couple of ground stone fragments (probably, an axe and a grindstone) are

Fig. 3. Spatial distribution of Bronze Age/Iron Age potsherds.

Fig. 4. Spatial distribution of Hellenistic potsherds.

470


Fig. 5. Spatial distribution of thick-walled hand-made vessel sherds with red-gray-red wall section.

among the earliest finds recovered by the survey. One may add to them a single typologically early Hvar-style potsherd, published by Petri}20. These finds indicate that the hill was occasionally visited or ephemerally occupied already during the Late Neolithic or the Copper Age (fifth or fourth millennia B.C.). A considerable assemblage of later prehistoric pottery was recovered, containing mostly small, non-diagnostic fragments of coarse, hand-made vessels. Occasional fragments of everted-rims, strap handles, massive horizontal handles, and flat bases, suggest that the greatest part of this assemblage belongs to the Late Bronze Age and the Early Iron Age4,21, and indicate that Grad was more permanently occupied during the last two millennia B.C. Coarse Hellenistic potsherds represent the majority of all collected finds (two thirds by number, three fourths by weight). Among them, amphorae fragments are the most common, especially those belonging to the type Lamboglia II22. Only eight fragments of the fine, black-slipped Hellenistic pottery known as »Gnathia ware« were recovered23, closely similar to the examples found in great quantity at the nearby Nakovana Cave. Abundant Hellenistic finds indicate that intensive human activity on Grad was contemporaneous with the use of Nakovana Cave as a sanctuary during the last few centuries B.C15. Extremely rare post-Hellenistic finds suggest that there was little activity at Grad after the first century B.C. Among the scarce evidence for later building activities are a few lumps of mortar that almost certainly post-date the Hellenistic period. Three sherds of glazed

pottery, each from a different vessel, probably belong to recent historic times. We also collected, recorded, and appropriately discarded a relatively small quantity of modern waste, such as plastic and glass bottles.

Spatial Distribution of the Surface Finds Distribution diagrams were produced for all classes of recovered archaeological materials. Of particular interest are spatial distributions of the three main classes of pottery. The plots indicate relative weight densities of potsherds collected from the surface, calculated in g/m2 and corrected for ground visibility. A thin scatter of Bronze Age/Iron Age pottery extends across much of the central, southern and eastern parts of the surveyed area, with a minor concentration at its western end, and a major concentration in the south, near the main point of access to the plateau (Figure 3). Hellenistic pottery is distributed quite differently. It is spread across the central, western and northern parts of the surveyed area, roughly coinciding with the visible structural remains, with a major concentration in the north, but it is virtually absent from its highest, eastern part (Figure 4). Finally, the distribution of thick-walled hand-made sherds with red-gray-red wall section is much closer to that of the Hellenistic pottery then to the Bronze Age/Iron Age pottery (Figure 5). This distinctively fired ware thus seems to be associated with, and contemporary to, the Hellenistic finds. 471


Spatial distributions of these three classes of pottery suggest that most of the structural remains visible at the surface should be associated with the Hellenistic period – either the structures were built during that period, or were still being used at the time. Differential distributions of Bronze Age/Iron Age and Hellenistic finds may indicate that different parts of the high plateau were used differently during those two periods. Alternatively, they may indicate that Bronze Age/Iron Age deposits are capped by a Hellenistic layer in the central and northern parts of the site, while in its southern part, the Hellenistic layer has been eroded away. While it is too early to discuss the function and internal organization of Grad, a few general remarks can be made. Character and quantity of the remains (mostly, coarse pottery, including many large vessels suitable for cooking and storage) hints at generalized residential activities. So do the lumps of burned clay, scattered almost everywhere across the site. On the other hand, the location itself suggests a defensive purpose, which is further supported by the remains of an enclosing structure that controlled the only feasible access to the high plateau. A fairly large drystone structure of elongated rectangular ground plan (13 x 7 m) occupies a central location and faces west; its purpose, for the moment, remains unknown. Other recorded structural remains are even less informative, consisting mainly of collapsed drystone walls. We collected only a very few fragments of ceramic roof tiles, which suggests that these were not widely used for covering roofs. Grad is a large and complex site, and each one of its parts probably was characterized by a specific set of func-

tions and activities. A particularly interesting issue is the relationship between the elevated, naturally protected central part of the site and its peripheral parts. Discussion of that issue must be postponed until intensive survey is extended to the surrounding slopes. For the moment, one can only speculate about a »special role« of the central plateau – its possible use as a fortress, a local elite residence, an area assigned for ceremonies, or a combination of those uses.

Conclusion Analysis of the data recovered by the intensive survey of Grad indicates that the central area of the hillfort was intensively used during the last few centuries B.C., at the time when Dalmatia was being incorporated into the world of the classical Mediterranean civilization. Its origins can be traced back to the Bronze Age, while a few artifacts testify of an even earlier episode of occupation. The recorded structures and the recovered finds hint at a residential and defensive function of this large site, while its central, elevated area may have been a focus of special activities. A continuation of intensive survey beyond that central area, augmented by test excavation at selected locations, is a prerequisite for resolving some of the issues raised in this report.

Acknowledgements This research was supported by the Ministry of Science, Education and Sports of the Republic of Croatia, project no. 0196004.

REFERENCES 1. MARCHESETTI, C.: I castellieri preistorici di Trieste e della regione Giulia. (Museo Civico di Storia Naturale, Trieste, 1903). — 2. MIROSAVLJEVI], V., ARR, 7 (1974) 259. — 3. ÔVI], B.: Eneolitski supstrat. In: ÔVI], B. (Ed.): Praistorija jugoslavenskih zemalja, vol. 4. (Academy of Sciences of Bosnia and Hercegovina, Sarajevo, 1983). — 4. ÔVI], B.: Srednjodalmatinska grupa. In: GABROVEC, S. (Ed.): Praistorija jugoslavenskih zemalja, vol. 5. (Academy of Sciences of Bosnia and Hercegovina, Sarajevo, 1987). — 5. BENAC, A.: Utvr|ena ilirska naselja (I). (Centre for Balkan Studies, Sarajevo, 1985). — 6. GAFFNEY, V., S. ÂÊ, B. KIRIGIN, P. LEACH, N. VUJNOVI], K. WARDLE, D. WARDLE: Enclosure and Defence: the Context of Mycenaean Contact with Central Dalmatia. In: KARAGHEORHGIS, V., C. E. MORRIS (Eds.): Defensive Settlements of the Aegean and the Eastern Mediterranean after c. 1200 B.C. (The A. G. Leventis Foundation, Nicosia, 2001). — 7. GAFFNEY, V., S. ÂÊ, J. HAYES, B. KIRIGIN, P. LEACH, N. VUJNOVI]: Secret Histories: The Pre-Colonial Archaeological Context for Greek Settlement of the Central Adriatic Islands. In: CAMBI, N., S. ÂÊ, B. KRIGIN (Eds.): Greek Influence along the East Adriatic Coast. (Knjièvni krug, Split, 2002). — 8. TER@AN, B., K. MIHOVILI], B. HÄNSEL, Archäologische Forschungen in Urgeschichtlichen Siedlungslandschaften, Regensburger Beiträge zur Prähistorischen Archäologie, 5 (1988) 155. — 9. ÔVI], B., Glasnik Zemaljskog muzeja BiH u Sarajevu, 32 (1978) 5. — 10. MARI], Z., 1976. Glasnik Zemaljskog muzeja BiH u Sarajevu, 30–31 (1976) 5. — 11. YORSTON, R. M., V. L. GAFFNEY, P. J. REYNOLDS, J. Archaeol. Sci., 17 (1990) 67. — 12. GAFFNEY, C. F., V. L. GAFFNEY, M. TINGLE: Settlement,

472

Economy or Behaviour? Micro-regional Land Use Models and Interpretation of Surface Artifact Patterns. In: HASELGROVE, C., M. MILLETT, I. SMITH (Eds.): Archaeology from the ploughsoil. (Department of Archaeology and Prehistory, Sheffield, 1985). — 13. GAFFNEY, V.: Ceramics ant the Site: Is Survey Enough? In: FRANCOVICH, R., H. PATTERSON (Eds.): Extracting Meaning from Ploughsoil Assemblages: The Archaeology of Mediterranean Landscapes. (Oxbow, Oxford, 2000). — 14. GAFFNEY, V., B. KIRIGIN, J. HAYES, T. KAISER, P. LEACH, Z. STANÎ^: The Adriatic Islands Project: Contact, Commerce and Colonization 6000 BC – AD 600. In: FRANCOVICH, R., H. PATTERSON (Eds.): Extracting Meaning from Ploughsoil Assemblages: the Archaeology of Mediterranean Landscapes. (Oxbow, Oxford, 2000). — 15. FORENBAHER, S., T. KAISER, Antiquity, 75 (2001) 677. — 16. FORENBAHER, S., B. KIRIGIN, N. VUJNOVI], Obavijesti Hrvatskog arheolo{kog dru{tva, 33 (2001) 46. — 17. FORENBAHER, S., T. KAISER, Obavijesti Hrvatskog arheolo{kog dru{tva, 34 (2002) 53. — 18. FORENBAHER, S., T. KAISER: Spila Nakovana: An Illyrian Sanctuary on the Pelje{ac Peninsula. (VBZ, Zagreb, 2003). — 19. VULETI]-VUKASOVI], V., Starinar 9 (1892) 90. — 20. PETRI], N.: Gradina Grad u Nakovani na Pelje{cu. In: PETRI], N. (Ed.): Novija i neobjavljena istraìvanja u Dalmaciji. (Croatian Archeological Society, Split, 1978). — 21. MARIJAN, B., VHAD, 93 (2000) 7. — 22. KIRIGIN, B., T. KATUNARI], L. [E[ELJ, Vjesnik za arheologiju i povijest dalmatinsku 98 (2005) 7. — 23. FORTI, L.: La ceramica di Gnathia. (G. Macchiaroli, Napoli, 1965).


S. Forenbaher Institute for Anthropological Research, Amru{eva 8, Zagreb, Croatia e-mail: [email protected]

PRETPOVIJESNA GRADINA GRAD NA PELJE[CU U DALMACIJI – PRELIMINARNI REZULTATI INTENZIVNOG PREGLEDA POVR[INE

SA@ETAK Intenzivni pregled povr{ine proveden na Gradu, gradini na Pelje{cu u ju`noj Dalmaciji, obuhvatio je otprilike 10.000 m2. Cilj pregleda bio je prikupiti podatke koji bi omogu}ili odre|ivanje veli~ine nalazi{ta, njegove funkcije, razdoblja naseljenosti, stupnja o~uvanosti, te potencijala za daljnja istraìvanja. Pregled je obuhva}ao kartiranje vidljivih ostatka arhitekture i sustavno sakupljanje svih povr{inskih nalaza. Ve}ina nalaza su ulomci grube helenisti~ke keramike, te ne{to manji broj ulomaka bron~anodobne i èljeznodobne lon~arije. Sredi{nji dio gradine intenzivno se koristio u zadnjim stolje}ima prije Krista, dok njeni po~eci seù u bron~ano doba. Vidljivi ostaci arhitekture i prikupljeni nalazi ukazuju na stambenu i obrambenu funkciji nalazi{ta, dok su se na njegovom uzdignutom sredi{njem dijelu vjerojatno odvijale posebne aktivnosti.

473


An Example of Demographic Anthropology, the Study of Matrimonial Exchanges – Endogamy, Choice of Spouse and Preferential Marriage Marie-Hélène Cazes National Institute of Demographic Studies, Paris, France

ABSTRACT The development of demographic studies in anthropology is directly linked to the success of population genetics. The anthropodemographic or anthropogenetic approach is thus underpinned by questions of genetics. While demographers focus on population dynamics and renewal in quantitative terms, population geneticists refer not to individuals but to the sets of genes carried by individuals in a population. Their aim is to detect the factors and processes which influence the genetic evolution of a group, i.e. which modify gene frequencies from one generation to the next. Among them are the factors which affect modes of reproduction. To illustrate the association of these three approaches, i.e. demographic, anthropological and genetic, I use here the example of matrimonial exchanges – which lie at the heart of the population renewal process – among the Dogon of Boni, a Malian ethnic group living in the southern Sahel. We can see how successive analyses – starting with endogamy at macroscopic level and moving down to the individual with choice of spouse and preferential marriage – combining both quantitative and qualitative approaches, can be used to obtain a detailed description of matrimonial exchanges which shed light upon and complement the three different viewpoints. Key words : demographic anthropology, genetic anthropology, matrimonial exchanges, endogamy, preferential marriages, Dogon of Roni, Mali

Introduction The development of demographic studies in anthropology is directly linked to the success of population genetics (sometimes qualified as »qualitative demography«, notably when it concerns the study of hereditary diseases). The anthropodemographic or anthropogenetic approach is thus underpinned by questions of genetics. While demographers focus on population dynamics and renewal in quantitative terms, population geneticists refer not to individuals but to the sets of genes carried by individuals in a population. Their aim is to detect the factors and processes which influence the genetic evolution of a group, i.e. which modify gene frequencies from one generation to the next. Alongside population size and migrations, these are the factors which affect modes of reproduction. Hence, for example, endogamy, family size distribution, choice of spouse, the existence or otherwise of preferential marriages, formation of marriage cycles, etc. are important parameters to be taken into account. And the concepts of anthropology are often borrowed for the study of these practices.

To illustrate the association of these three approaches, i.e. demographic, anthropological and genetic, I will use the example of matrimonial exchanges – which lie at the heart of the population renewal process – among the Dogon of Boni, a Malian ethnic group living in the southern Sahel. The topography of their territory is highly specific, comprising a series of raised massifs around 15 km apart that emerge above the plain. The Dogon population totalled around 5,000 at the time of the study, distributed between 4 massifs, each comprising 3 to 4 villages.

Matrimonial Exchanges among the Dogon Endogamy Various levels of analysis are possible. Exogamy (or its opposite endogamy) are located at the macroscopic level and call for a quantitative approach. Endogamy may exist in varying degrees, at the scale of the ethnic group or

Received for publication June 11, 2006

475

M.-H. Cazes: Endogamy, Choice of Spouse and Preferential marriage, Coll. Antropol. 30 (2006) 3: 475–478

on a finer, intra-ethnic scale. Among the Dogon of Boni for example, strong ethnic endogamy is observed1 : 96% of marriages take place between Dogon, and only 4% with a different ethnic group. If the analysis is pursued within the ethnic group, by cross-tabulating the origins of spouses by their village of birth, we obtain a table in which the majority of marriages are on the diagonal. We observe that 84% of marriages take place within the same massif, and only 16% between massifs (Table 1).

This type of macroscopic analysis, though specific to demography, also seeks to interpret the phenomena observed. It highlights the importance of factors such as population size, geographical location or the weight of tradition in the practice of endogamy. The demographic and genetic consequences of this endogamy are important, since it divides the population into »islands«, thus implying that the population may be biologically heterogeneous.

It is interesting to look for an explanation for this massif endogamy, using log-linear models for example, which can be used to test, from a statistical viewpoint, the pertinence of certain criteria2.

Choice of Spouse Beyond endogamy, there is a second level of analysis involving the study of preferential marriage. This is a much finer »microscopic« level. Do matrimonial customs recommend the choice of spouse based on a criterion of kinship? Since most Dogon marriages take place within the massifs, the second analysis level focuses on a local level4. Within the massif, do people marry by chance (i.e. with no constraints other than those linked to age difference between spouses or to the existence of durable unions in the population)? Or does preferential marriage exist?

Among the Dogon, these endogamy criteria are variable3. When the village is large enough (as is the case for Tabi, a village of 1,200 inhabitants), significant lineage endogamy is observed. On the Sarnyéré massif, where customs are strongly adhered to, spouses are sought either in the same lineage or, failing that, in the same village. On the two other massifs, Ella and Loro, where the villages are smaller and much further apart, the criteria of geographical distance alone is sufficient to explain the exchanges.

TABLE 1 DISTRIBUTION OF MARRIAGES CONCLUDED SINCE THE FOUNDATION OF THE GROUP BY SPOUSES’ VILLAGE OF BIRTH

Husband’s village Wife 's village

Sarnyéré massif Nem

Dja

Tan

Tabi massif

Koyo

Tabi

Tup

Ella massif

Téga

E.Bu E.Bo Mom

Loro massif Ban

Loro Yuna K.Bo Prin

Nemgéné

349

72

67

72

12

2

1

5

0

1

0

0

0

0

Djamaga

60

124

60

40

0

0

0

0

2

1

0

0

0

1

0

Tandi

54

56

98

29

1

0

1

2

1

0

0

0

0

0

0

Koyo

41

26

20

34

0

0

0

0

0

0

0

0

0

0

0

504

278

245

175

13

2

2

7

3

2

0

0

0

1

1

10

0

2

1

746

39

60

1

1

0

0

0

1

0

10

Tupéré

0

1

2

0

31

164

19

0

0

0

0

2

1

3

5

Téga

0

0

6

0

49

35

276

0

0

0

0

4

1

3

1

10

1

10

1

826

238

355

1

1

0

0

6

3

6

16

Ella-Buli

1

2

4

0

0

0

0

10

6

6

0

0

0

1

1

Ella-Boni

2

0

0

0

3

1

1

6

20

13

10

0

0

4

1

Momni

2

1

1

0

1

0

0

8

11

22

10

5

0

4

0

Banaga

0

0

0

0

0

1

0

3

10

11

20

3

0

4

1

Total Ella

5

3

5

0

4

2

1

27

47

52

40

8

0

13

3

Loro

0

0

0

0

0

0

1

1

7

3

5

85

9

39

1

Yuna

0

0

0

0

1

2

3

0

0

1

0

15

7

10

5

Koyo-Boni

0

0

0

1

0

0

0

2

2

3

1

26

2

35

0

Total Sarnyéré Tabi

Total Tabi

1

Pringa

0

0

0

0

11

5

2

0

2

0

0

1

4

0

88

Total Loro

0

0

0

1

12

7

6

3

11

7

6

127

22

84

94

All massifs

519

282

260

177

855

249

364

38

62

61

46

141

25

104

114

Non dogon vill.

0

0

0

0

0

0

0

0

2

2

5

3

0

0

0

Unknown vill.

60

36

22

17

57

19

26

9

24

17

24

41

4

9

42

579

318

282

194

912

268

390

47

88

80

75

185

29

113

156

TOTAL

476

M.-H. Cazes: Endogamy, Choice of Spouse and Preferential marriage, Coll. Antropol. 30 (2006) 3: 475–478 TABLE 2 DISTRIBUTION OF MARRIAGES BETWEEN COUSINS OR OTHERWISE AMONG THE DOGON OF BONI

Marriages between first cousins

Other marriages

Marriage type

TOTAL MoBrDa

FaSiDa

MoSiDa

FaBrDa

Total

Unkn. A

N. C.

Total

First marriages

57

Next marriages

31

25

9

36

127

264

259

523

15

17

31

94

352

329

681

All marriages

88

775

40

26

67

221

616

588

1,204

1,425

650

MoBrDa – mother’s brother’s daughter (matrilateral cross-cousin), FaSiDa – father’s sister’s daughter (patrilateral cross-cousin), MoSiDa – mother’s sister’s daughter (matrilateral parallel cousin), FaBrDa – father’s brother’s daughter (patrilateral parallel cousin), Unkn.A – unknown ancestry, N. C. – non cousin spouses

This question is an anthropological one, though geneticists are also highly interested in the answers obtained. For them, the most important point is to know whether preferential marriage rules are actually applied or whether they are no more than »theoretical«. If preferential marriage is indeed widely practiced, this will affect the genetic structure of the group. Analysis of this kind brings in anthropological concepts, taking into account the population’s kinship terminology (Iroquois among the Dogon) but also, in many cases, its history. Traditionally among the Dogon, preferential marriage was with the matrilateral cross-cousin (mother’s brother’s daughter, MoBrDa), and this tied in with the traditional system of inheritance whereby Ego inherited from his maternal uncle. But over the last century, the introduction of Islam has transformed the customary system. In certain villages, the preferential spouse has become the patrilineal cross-cousin (FaSiDa). From a methodological viewpoint, the possibility of obtaining the family trees of the population over several generations is a considerable advantage. A database of this kind makes it possible to compare the types of marriage actually concluded with the rules laid down by custom. Thus calculations drawn from Dogon family trees show that, for a young man who wishes to marry, only an average of 6% of young women of the right age are cousins. Yet Table 2 shows that 19.5% of marriages are between first cousins for first marriages, though this proportion falls to 15.5% for all marriages. If the four types of marriage between cousins are ranked in order of decreasing proportion among first marriages, the following result is obtained:

This is why researchers always seek to compare actual observations with what one would expect to observe in the case of random unions. Moreover, we know that the mean age difference between spouses is in itself a structural constraint which may be decisive in the choice of spouse. Among the Dogon, wives are, on average, 6 years younger than their husbands. In a pioneering article in 1963, Hajnal showed that certain marriages between cousins reproduced the traditional age differences between spouses more easily than others (Figure 1). Here again, genealogical data, taking account of individuals’ age, were used to calculate the theoretical percentages of marriage with each type of cousin in the Tabi village, taking account of the age difference between spouses (Table 3, 1st line). They can be compared with the proportions observed in the population (Table 3, 2nd line). We observe a very high proportion of marriages with the matrilateral cross-cousin (MoBrDa) compared with the expected average and strong avoidance of the matrilateral parallel cousin (MoSiDa), probably linked to the incest taboo.

45% MoBrDa 28% FaBrDa 20% FaSiDa 7% MoSiDa The most frequent marriage is indeed that recommended by traditional custom (MoBrDa) and not the type of marriage newly »officialized« over the last century (FaSiDa) which arrives in third position. The type of marriage classically recommended by Islam (FaBrDa) ranks in second position only. So these observations contradict the declarations relating to the »new« preferential marriage. For geneticists, the key question is to know whether the observed distribution of marriages differs from that which would be obtained in the case of random marriage distribution.

Fig. 1. The different types of Ego’s cousins. The age difference observed between Ego’s mother and father is observed again between Ego and his matrilateral cross-cousin only. MoBrDa – mother’s brother’s daughter (matrilateral cross-cousin), FaSiDa – father’s sister’s daughter (patrilateral cross-cousin), MoSiDa – mother’s sister’s daughter (matrilateral parallel cousin), FaBrDa – father’s brother’s daughter (patrilateral parallel cousin).

477

M.-H. Cazes: Endogamy, Choice of Spouse and Preferential marriage, Coll. Antropol. 30 (2006) 3: 475–478 TABLE 3 COMPARISON BETWEEN EXPECTED AVERAGE PROPORTIONS UNDER THE ASSUMPTION OF RANDOM UNIONS, TAKING ACCOUNT OF AGE DIFFERENCE BETWEEN SPOUSES, AND OBSERVED PROPORTIONS

Cross-cousins

Parallel cousins Total

Marriages with MoBrDa

FaSiDa

MoSiDa

FaBrDa

Average proportions expected under the assumption of random unions

0.32

0.18

0.27

0.23

1

Proportions observed

0.49

0.23

0.03

0.25

1

MoBrDa – mother’s brother’s daughter (matrilateral cross-cousin), FaSiDa – father’s sister’s daughter (patrilateral cross-cousin), MoSiDa – mother’s sister’s daughter (matrilateral parallel cousin), FaBrDa – father’s brother’s daughter (patrilateral parallel cousin)

Hence, family trees can be used to control for the divergence between »words« and »deeds«. Among the Dogon, the most frequently observed form of marriage is with the matrilateral cross-cousin. And, given the age difference between spouses, this is the only cousin who structurally reproduces the right customary age difference. It is also the spouse who was recommended by traditional custom before the arrival of Islam. So Islamization in the early 20th century has not modified matrimonial practices, despite what the Dogon say.

Conclusion We can see how successive analyses, starting at macroscopic level and moving down to the individual, combining both quantitative and qualitative approaches, can be used to obtain a detailed description of matrimonial exchanges which shed light upon and complement the three different viewpoints: demographic, but also and above all, anthropological and genetic. They simultaneously highlight the complexity of these exchanges, which are always very difficult to interpret.

REFERENCES 1. CAZES, M. H., E. BROWN, B. FLOURY, A. JACQUARD, C. SAUVAIN-DUGERDIL: Les Dogon de Boni, approche démo-génétique d’un isolat du Mali, Cahier Travaux et Documents n°132. (Ined-PUF, Paris, 1993). — 2. BISHOP, Y. M. M., S. E. FIENBERG, P. W. HOLLAND: Dis-

crete Multivariate Analysis: Theory and Practice. (MIT Press, Cambridge MA, 1975). — 3. CAZES, M. H., J. Biosoc. Sci., 22 (1990) 85. — 4. CAZES, M. H., Soc. Biol., 28 (1981) 281.

M. H. Cazes Institut National d’Etudes Démographiques, 133 boulevard Davout, 75980 Paris cedex 20, France e-mail: [email protected]

PRIMJER DEMOGRAFSKE ANTROPOLOGIJE, ISTRA@IVANJE ODABIRA BRA^NOG DRUGA – ENDOGAMIJA I ODABIR SUPRU@NIKA

SA@ETAK Razvoj demografskih istraìvanja u antropologiji izravno je povezan s uspjehom populacijske genetike. Tako su antropodemografski ili antropogeneti~ki pristup poduprti pitanjima iz genetike. Dok su demografi usmjereni na dinamiku populacije i reprodukciju u kvantitativnim odre|enjima, populacijski geneti~ari ne okre}u se pojedincima, ve} skupu gena ~iji su nosioci pojedinci u populaciji. Njihov cilj je otkriti faktore i procese koji utje~u na geneti~ku evoluciju grupe, tj. one koji modificiraju frekvencije gena od jedne do druge generacije. Me|u njima su faktori koji utje~u na reprodukciju. Kako bi se prikazala povezanost ova tri pristupa, tj. demografskog, antropolo{kog i geneti~kog, bit }e kori{ten primjer odabira bra~nog druga – koji predstavlja sredi{te procesa produènja populacije– izme|u Dogona iz Bonija, malijske etni~ke skupine koja ìvi u ju`nom Sahelu. Moèmo vidjeti kako uzastopne analize koje kombiniraju kvantitativne i kvalitativne pristupe i to po~ev{i od endogamije na makroskopskoj razini do odabira supru`nika te poèljnog braka na individualnoj razini mogu biti kori{tene za postizanje preciznijeg opisa bra~nih izmjena, {to dodatno rasvjetljava i upotpunjuje tri pristupa problemu.

478


Trend of Mortality Observed in a Cohort of Drug Addicts of the Metropolitan Area of Bologna, North-Eastern Italy, During a 25-Year-Period Roberto Manfredi1, Sergio Sabbatani1 and Daniele Agostini2 1

2

Department of Clinical and Experimental Medicine, Division of Infectious Diseases, University of Bologna »Alma Mater Studiorum«, S. Orsola-Malpighi Hospital, Bologna, Italy Department of Public Health, Bologna, Italy

ABSTRACT The aim of our study is to evaluate the temporal trend of deaths in a cohort of i.v. drug users (IVDU) followed in a city of Northen Italy (Bologna), and to assess its relationship with HIV infection and AIDS, and availability of potent antiretroviral therapy. One thousand and 214 IVDUs (mainly heroin addicts), 916 males and 298 females, attending an outpatient service for treatment and prevention of substance abuse between 1977 and November 1996, were enrolled into our observational cohort, and their vital status was ascertained up to December 31, 2002. The large majority of enrolled subjects were born in the Bologna metropolitan area and surroundings; no extra-European immigrants were present. During the observation period, 271 IVDUs (22.3%) died, 211 males (23.0%), and 60 females (20.1%). No death was recorded before 1984. Main death causes result as follows: AIDS (52.8% of episodes), heroin overdose (22.1%), street accidents (7.4%), decompensated liver cirrhosis (6.3%), and suicide (2.9%). The highest absolute number of deaths was observed between years 1991 and 1996. Crude mortality rate caused by AIDS was 10.0 per 1000 for males and 13.2/1000 for females; the rate of death due to other causes proved 11.1/1000 among males and 5.2/1000 among females. In most recent years, a sharp decrease in the number of AIDS-related deaths, attributable to the increased use of potent antiretroviral regimens, was recorded among IVDUs, although overall mortality rate remained appreciable. Key words: epidemiology, drug users, HIV infection, overdose, mortality, temporal trend, antiretroviral therapy

Introduction A remarkable decrease of HIV-associated disorders, as well as of notified cases of AIDS, has been observed in industrialized countries during the last seven years. Such significant drop of HIV-related morbidity and mortality (mostly occurred thanks to the introduction and diffusion of highly active antiretroviral therapy, or HAART), was also observed among i.v. drug users (IVDUs)1–3. However, a significant number of potential Italian patients, still unaware of their HIV serostatus4, do not yet make use of antiretroviral treatment: in our previously reported series regarding the year 1997, 29% of active or prior IVDUs were never tested for HIV infection5. A European multicentre study6 shows how IVDUs with recent HIV seroconversion are subject to an increased risk of mortality due to pneumonia, endocarditis, sepsis, meningitis, encephalitis, and decompensated liver

cirrhosis. Before the availability of HAART, an increased rate of mortality due to heroin overdose and suicide was also recorded among HIV-infected IVDUs7,8. Just the use of elevated heroin dosages associated to a suicidal behavior could be partially responsible for this phenomenon in patients with a newly diagnosed HIV disease and lacking of any social and psychological support, especially before the introduction of effective antiretroviral drug combinations (HAART)9–11. In a cohort study of 4,962 IVDUs carried out in Bologna (Italy) between years 1980 and 1990, 332 deaths were observed: 150 of them occurred because of AIDS, 64 of drug overdose, and 39 of accidents8. The HIV infection serostatus (28.2% of 1214 subjects were positive) proved to be significantly related also to deaths caused by trauma, heroin overdose, bacterial infections, and acute and

Received for publication August 29, 2005

479

R. Manfredi et al.: Temporal Trend of Mortality Rate in a Cohort of Drug Addicts, Coll. Antropol. 30 (2006) 3: 479–488

chronic liver disease. The mortality rate of IVDUs proved significantly higher among those subjects who did not receive a HIV serodiagnosis with respect to those subjects who underwent a HIV infection test with negative result (12.1% versus 2.5%)8. Again in Bologna, at the end of nineties the risk of a fatal heroin overdose remained remarkably important12, while between years 1996 and 1998 Italian national cumulative data showed a progressive decrease of deaths due to this last cause13. In a cohort of 11,432 IVDUs enrolled in Rome (Italy) between years 1980 and 1995, 1,734 deaths were registered (15.2%). The peak of AIDS-associated mortality was observed in 1991 and 1992 (13.2 per 1000 personyears), followed by a decrease occurred in both male and female population since 1993–1994, even earlier than expected according to the availability of (before HAART became available in mid-1996)14. In the Emilia-Romagna region of Italy (of which Bologna is the main administrative centre and largest town), a retrospective study conducted on 4,260 IVDUs followed in Piacenza, Modena and Ferrara during two decades (1975–1995), pointed out a dramatic increase in the rate of deaths caused not only by AIDS; but also by overdose and other causes, especially accidents. The mortality rate observed in this last IVDUs cohort proved particularly high: the calculated standard mortality rate (SMR) was 16.7 for males, and 33.4 for females. The overall probability of survival after 15 years of follow-up was 65%. After drug overdose and AIDS, the other relevant causes of deaths were street accidents, decompensated liver disease, other infectious disorders and malignancies among male IVDUs, while accidents (in particular, homicide), and gastrointestinal tract disorders had a greater prevalence among females15. Finally, in an Italian multicentre study carried out in Turin, Rome, Naples, and Cagliari on IVDUs enrolled between 1980 and 1992, a highly variable mortality rate was detected, ranging from 11.0 per 1000 persons-years in Naples, up to 20.5 per 1000 persons-years in Rome. Again, the most frequent causes of death were represented by drug overdose, AIDS, and accidents, although relevant differences were observed among the four considered cities. A higher mortality rate among IVDUs compared with that of the general population was consistently observed in all considered geographical cohorts16,17. In a very recent 7-year survey from the US, active drug use proved temporally linked to HIV disease progression and overall mortality17. Our present study aims at: i) investigating the past and present mortality trend of a cohort of IVDUs based in Bologna, in an attempt to identify its causes; ii) investigating whether HIV infection plays any influence on the rate of mortality due to causes other than AIDS; iii) evaluating the AIDS-associated fatality rate among subjects with missed or delayed HIV serodiagnosis, hen480

ce with a possible late diagnosis of the infection only at an advanced stage of the condition.

Materials and Methods Patients An open cohort of 1,214 IVDUs has been assessed for a number of epidemiological and clinical variables. The people involved were mainly i.v. heroin addicts, and included 916 males (75.5%) and 298 females, referring to a specialized outpatient Centre for the treatment and prevention of substance abuse of Bologna (Italy), from the year 1977 up to November 1996. The great majority of enrolled patients at the time of their first visit lived in South-Western districts of the Bologna conurbation (400,000 inhabitants in the city, half a million with the suburbs). The place of birth is a missing datum for one minor part of the people reached by our study; as to the others, most of them were born within the Province of Bologna and no one comes from abroad. By matching anagraphical data with the registries of the abode or birth municipalities, a retrospective research focusing on whether subjects were still alive on December 31, 2002, has been carried out. In some cases, neither the year of birth nor the subject’s age were available due to patients’ refusal to provide their own personal data. Referring to population registries allowed us to retrieve missing information regarding almost all individuals in our cohort: in 60 cases, still assisted by our outpatient services, data are still lacking, but for deceased patients figures were adjusted when causes of death were researched.

HIV infection At the time of patients’ first contact with our dedicated outpatient centre, HIV testing and specific counselling were always offered, but 355 subjects out 1,214 (29.2%) refused this examination. HIV serology became recommended in late 1984, so that the first recognition of a possible underlying HIV infection is related to the first visit at our outpatient Centre, starting from 1984–1985. IVDUs enrolled before 1985 had HIV testing performed starting from this year either at the outpatient units, or at one of the two Infectious Diseases Divisions of Bologna, or local prison facilities. At the time of their next contact with the above-mentioned health care structures, HIV testing and specific counselling were always offered. Data regarding HIV serostatus for IVDUs enrolled after 1985 dates back mainly to the time of first access to IVDA outpatient Centres. Usually, the seroconversion time for patients who were HIV-negative at their first control is not definable, because of the limited availability of patients to be tested or to furnish results.

Statistical analysis As to the calculations of person-years and rates, the beginning of observation is identified with the year of the first visit at our outpatient Centre for IVDUs. The first

R. Manfredi et al.: Temporal Trend of Mortality Rate in a Cohort of Drug Addicts, Coll. Antropol. 30 (2006) 3: 479–488

access date of 41 individuals is unknown; for each of them, the enrollment was identified with the first date available in his/her clinical records. The end of observation period was matched with that of death for deceased patients, while it was fixed at December 31, 2002 for living individuals; for two persons who were lost to follow-up, the end-of-observation time is identified with 1997: the year of their last access to the Centre. The cause of death of deceased subjects was drawn by local registries, as expressed by the ICD-IX code18,19. Population data were obtained from the official annual statistical update of Bologna; mortality figures were drawn by public health notification registries. Statistical analysis included specific mortality rates per age, gender and HIV-status, and standardized mortality rate (SMR)20, obtained by indirect standardization performed on Bologna rates per each considered year. Chi-square has been used for the analysis of frequency distribution; Student’s t test and Analysis of variance were employed to evaluate differences among mean values. The rates per HIV serostatus were calculated by attributing the years of observation of each single subject to his/her own category (HIV-positive, HIV-negative, HIV-unknown), without focusing on the time of seroconversion (which was not always available). As a consequence, not each person’s time of HIV-positive or HIV-negative serology, but only the patient serostatus was considered. The considered study time was analyzed according to a wide range of variables, by dividing it into four different periods, characterized by prominent epidemiologicals, treatment, prognosis, type of medical assistance, and outcome. The first period (years 1977–1983) encompassed the initial service offer, opposition and resistance by patients, no deaths, and absence of HIV-AIDS cases; the second period (1984–1990) was characterized by the rapid spread of AIDS, spontaneous patient access, and early HIV-associated deaths; during the third period (year 1991–1996) the patient access remained sustained, AIDS had the greatest fatality rate, and many deaths occurred; the fourth period (years 1997–2002) was characterized by the availability of highly active antiretroviral therapy (HAART), followed by a sharp drop of AIDS mortality. Our cohort was closed to enrollments at the end of previous period.

Results Series description – general features Overall, 1,214 patients were considered: 916 males and 298 females. The distribution per age classes at the time of enrollment is summarized in Table 1A, where also enrollment period is reported. The age of 60 subjects was not available. Table 1B shows the subjects’ enrollment age in different periods. One hundred and 10 IVDUs were enrolled from year 1977 to 1983. At the end of year 1990, patients still alive in our cohort were 607, and in the period 1991–1996 enrolled and alive patients became 1,174. The characterizing features of the above- -mentioned time periods are recognizable in the cumulative Table 6.

Age and sex Overall follow-up accounted for 10,030.50 person-years for male subjects, and 3,249.75 person-years for females: cumulated follow-up accounted for 13,280.25 person-years. Mean age did not differ between enrolled males and females, when considering both the entire follow-up, and single time periods. A temporal trend to increased patients’ age occurred through time: the difference among mean age at enrollment during different time periods turned out to be significantly different (p< .0001; Table 1B). The age at which each person started to take up IVD was asked, but not provided for nearly one half of involved subjects;.

HIV serostatus When assessing HIV serostatus, IVDUs HIV-positive were 426, and HIV-negative subjects were 433 (Table 2). In particular, HIV infection was found in 297 men (32% of males), and in 298 women (43% of females). Moreover, among tested individuals, HIV-positive male patients accounted for 46%, while females were 61%, leading to a statistically different distribution (2=14,2667; p= 0.0002). The record of HIV positive serostatus was documented in 8% of patients early HIV-negative; the great majority of them was deducted from death certifications.

Analysis of deaths Overall IVDUs subjects dead on December 31, 2002 were 271 (22.3% of all enrolled individuals): 211 males (23% of enrolled males), and 60 females (20.1% of the female cohort). The mean patient age at the time of death was comparable by gender: 33.2±6.6 years among males, and 33.5±6.0 years among females (2=1,091, p=n.s.). The overall 1.214 enrolled IVDUs accounted for 13,280.25 years-person (a mean of nearly 11 years per patient), while the 271 deceased individuals contributed for 1,258 years-person (an average of less than five years per patient). Table 3 resumes all data concerning deaths, classified per age group and sex, with person-years and rates per 1000. The overall death distribution does not show any significant difference between males and females. The time of initial i.v. drug addiction, when available, does not provide evidence of significant relation between age of start of IVDUs and eventual lethal outcome. The cause of each single death has been retrieved for 267 of the 271 deceased patients (98.5%). No deaths were registered in this cohort in the years preceding 1984. Figure 1 shows the different causes of death per year, in the period 1984–2002. Overall AIDS-related deaths total 143 (52.8%), followed by those due to heroin overdose (60: 22.1%), liver cirrhosis (17: 6.3%), road accidents (20: 7.4%), suicide (8: 2.9%), homicide (3: 1.1%), infectious endocarditis and other heart disorders (6: 2.2%), neoplasm (lung's or lymphoma, 5 deaths, 1.8%), and pneumonia (2: 0.7%). Three causes of death occurred only once: wasting syndrome, bleeding endocranic aneurism, and fire accident. The cause of death could not be found 481

R. Manfredi et al.: Temporal Trend of Mortality Rate in a Cohort of Drug Addicts, Coll. Antropol. 30 (2006) 3: 479–488 TABLE 1 BOLOGNA IVDU COHORT. CASE DISTRIBUTION ACCORDING TO SEX, ENROLLING PERIOD, AND AGE AT ENROLLMENT (SECTION A); COMPARISON OF DIFFERENT PERIODS AND ENROLLMENT AGES (SECTION B)

Section A Enrolling Sex Women

Age groups

I

II

0–14 15–19 20–24 25–29 30–34 35–44 45–49 50–54 Nn

1 6 7 4 2 1

1 9 45 38 20 10

Total women Mean age Standard error Men

Period

0–14 15–19 20–24 25–29 30–34 35–44 45–49 50–54 >55 nn

Total men Mean age Standard error

1 22 22.71 1.342

III

3 126 26.06 0.487 2 18 130 125 55 22 2 1

21 41 16 6 1

1 2 88 22.88 0.609

19 374 26.17 0.287

All

5 33 49 26 27 2 1 7 150 29.01 0.557

2 20 85 91 48 38 2 1 11 298 27.29 0.377

10 95 143 97 67 10 3 1 28 454 29.28 0.324

2 49 266 284 158 90 12 4 2 49 916 27.37 0.219

Section B Enrolling Sex Total Mean age Standard deviation Standard error 95% confidence limits

Age groups

lower upper

Median Minimum Maximum Anova variation among periods within periods total

Period I 110 22.85 5.7196 0.553 21.7542 23.9467 22 14 56

500 26.14 5.4080 0.247 25.6562 26.6283 25 10 53

III 604 29.21 6.6763 0.280 28.6629 29.7624 28 17 62

All 1214 27.35 0.189

Deviance (SS)

gf

Variance (MS)

p (F)

4837.661 42735.203 47572.864

2 1151 1153

2418.831 37.129 41.260

0.05). The obtained slopes of the regression lines correspond to allometric exponents b needed to assess body mass in609

G. Markovi} and D. Sekuli}: Modeling Lifting Performance for Body Size, Coll. Antropol. 30 (2006) 3: 607–613 TABLE 1 ALLOMETRIC MODEL RELATIONSHIPS FOR WEIGHTLIFTING (WL) AND POWERLIFTING (PL) PERFORMANCE (S) TO BODY MASS (M)

N

b

TABLE 2 SECOND ORDER POLYNOMIAL MODEL RELATIONSHIPS FOR WEIGHTLIFTING (WL) AND POWERLIFTING (PL) PERFORMANCE TO BODY MASS (M)

95% CI

Equation

Equation Men’s WL

S= –0.037M2 + 8.68M – 71.77

70

0.61±0.02

0.58 – 0.64

S=25.56M0.61

Women’s WL 70

0.68±0.04

0.62 – 0.76

S=13.85M0.68

Women’s WL

S= –0.068M2 + 10.76M – 169.11

Men’s PL

0.69±0.02

0.65 – 0.74

S=36.31M0.69

Men’s PL

S= –0.081M2 + 20.04M – 232.37

0.74 – 0.87

S=19.11M0.80

Women’s PL

S= –0.070M2 + 15.59M – 151.16

Men’s WL

90

Women’s PL 80

0.80±0.03

N – number of subjects, 95% CI – 95% confidence interval, b – allometric parameter (X±standard error)

S – normalized performance

dependent indices of lifting performance (see parameter b in eq. 3). These allometric exponents b (Table 1) represent the main result of the study. The exponents of 0.68 and 0.69 for women’s WL and men’s PL data are similar to the theoretical prediction of 0.67. In contrast, 95% confidence intervals of the mass exponents b for men’s WL and women’s PL performance do no not include the value of 0.67, suggesting that these lifting performance scale with negative and positive allometry, respectively. It must also been pointed out that women’s mass exponents b were significantly higher (p0.4). In addition, no linear relationship was found between residual scores and body mass for allometric modelling (correlation coefficients r ranged between –0.11 and 0.09, p>0.05), suggesting that the model errors displayed homoscedasticity. The presented data suggest that both models demonstrated excellent fit to the lifting performances analyzed.

a)

b)

0.1

0.1

0.05 Raw residual

Raw residuals

0.05

0 R2 = 0.59 p < 0.05

–0.05

–0.1 3.9

c)

4.1

4.3 ln body mass

4.5

0

–0.05 –0.1 3.8

4.7

d)

4

4.1

4.2

4.3

4.4

0.15 0.1

Raw residuals

0.1

Raw residuals

9.3

ln body mass

0.15

0.05 0 –0.05 R2 = 0.23

–0.1 –0.15 3.8

R2 = 0.20 p < 0.05

4

4.2

4.4

ln body mass

4.6

0.05 0 –0.05 R2 = 0.02 p > 0.05

–0.1

4.8

–0.15 3.7

3.9

4.1 ln body mass

4.3

Fig. 2. Scatter plot of raw residuals vs. ln body mass for allometric model applied to: (a) men’s weightlifting results, (b) women’s weightlifting results, (c) men’s powerlifting results, and (d) women’s powerlifting results. Second-order polynomial curve fits are displayed (solid line) together with associated coefficient of determination (R2).

610

G. Markovi} and D. Sekuli}: Modeling Lifting Performance for Body Size, Coll. Antropol. 30 (2006) 3: 607–613

However, a detailed inspection of scatter plots of residuals vs. ln M (Figure 2) shows a curvilinear relationship for all allometric models, except for women’s PL results. In particular, a moderate to strong non-linear relationship was found between raw residuals and ln M (R2 =0.20–0.59, p0.05). That means that the differences between arithmetic means of NPS and PS athletes are not statistically significant in applied system of variables for these two groups of athletes. Multivariate analysis of variance for spirometric variables is showed in Table 4. Wilks’ Lambda and Rao R

(which tests value of lambda) are statistically significant for group B and altogether athletes. That means statistically significant differences between arithmetic means of NPS and PS athletes in these two groups in applied system of variables. Differences at each observed spirometric variable between unexposed (NPS) and exposed (PS) athletes to passive smoking are presented in Table 3. Results analyses (ANOVA-Table 3) clearly show significant statistical differences in whole system of spirometric variables between NPS and PS athletes for group B and altogether athletes. Significant statistical differences (p