The Csajág mammoths ( Mammuthus primigenius): Late Pleniglacial finds from Hungary and their chronological significance

lable at ScienceDirect

Quaternary International 255 (2012) 130e138

Contents lists avai

Quaternary International

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

The Csajág mammoths (Mammuthus primigenius): Late Pleniglacial findsfrom Hungary and their chronological significance

Lajos Katona a, János Kovács b,*, László Kordos c, Bálint Szappanos d, Isván Linkai e

aNatural History Museum of Bakony Mountains, Rákóczi tér 3-5, H-8420 Zirc, HungarybDepartment of Geology, University of Pécs, Ifjúság u. 6, H-7624 Pécs, HungarycGeological Institute of Hungary, Stefánia u. 14, H-1143 Budapest, Hungaryd Institute of Geography and Earth Sciences, Eötvös University, Pázmány P. s. 1/C, H-1117 Budapest, HungaryeMenyhárt és M. Kft., Mónus Illés u. 47-49, H-9024 Gy}or, Hungary

a r t i c l e i n f o

Article history:Available online 24 February 2011

* Corresponding author.E-mail addresses: [email protected] (L. Katon

(J. Kovács).

1040-6182/$ e see front matter � 2011 Elsevier Ltd adoi:10.1016/j.quaint.2011.01.048

a b s t r a c t

The Csajág mammoths were discovered during road construction work in June 2006. The skeletalremains are well preserved in an Upper Pleistocene loess deposit. This revealed the skeletons were anadult female woolly mammoth (Mammuthus primigenius) of estimated age 24e25 years, largely completeexcept for the skull; and the partial skeleton of a juvenile of age 6e7 years at death. A tooth sample hasbeen radiocarbon dated (AMS) and is of Late Pleniglacial (MIS 2) age (16.9e15.9 ka cal BP). This newradiocarbon evidence fits into the well-known colonization pattern of M. primigenius in East CentralEurope and confirms a continuous distribution at the end of the Late Pleistocene.

� 2011 Elsevier Ltd and INQUA. All rights reserved.

1. Introduction

During the Last Cold Stage the woolly mammoth occurredwidely across Northern Eurasia, including nearly all of Europe,mostly in association with regional treeless steppe-tundra vegeta-tion (Kahlke, 1994; Lister et al., 2005; Stuart, 2005; Álvarez-Laoet al., 2009). Fossil remains of Mammuthus primigenius (Blu-menbach 1799) are common in Hungary; ca. 400 specimens havebeen recovered, among these are six complete skeletons (Vörös,1981; Konrád et al., 2010).

The Csajág mammoths were discovered during road construc-tion (Road 710) work in June 2006. The mammoths are beingexcavated by colleagues from BakonyMuseum and somemotivatedvolunteers. The Csajág mammoths represent the youngest knownand most complete records of M. primigenius in Hungary (Katonaet al., 2010). This paper describes the Csajág remains, assessesthem in terms of number of individuals represented, age and sex,and discusses the taphonomy and palaeoenvironment. It alsoshows the radiocarbon ages of other woolly mammoth remainsfound in Hungary and the surroundings.

The morphometry of the Csajág mammoths remains wereexamined to characterize the individuals represented and confirmwhether their anatomy differs from that of contemporaneous

a), [email protected]

nd INQUA. All rights reserved.

mammoths of other regions. The palaeobiogeographic position ofthe Csajág finds is then placed in a regional context to give newrecords and chronology of these animals’ Central European rangeduring the Late Pleistocene (MIS 2).

2. Regional setting and site characteristics

The excavation site (47�0105100N; 18�1001600E) is located in anopen area near Lake Balaton and the village of Csajág (Fig. 1). Thesouthern foreland of the Transdanubian Range includes one largerrange and some smaller hills. These hills are made of Palaeozoicformations exposed through Upper MioceneeQuaternary sedi-ments. The area between the Velence Hills and Lake Balaton iscalled Balatonf}o. The territory of the Csajág area is covered byUpper Pleistocene loess, formed in MIS 3 and 2. The loess occurs ina few large outcrops and it is mainly exposed in smaller road cuts.The typical material is mainly aeolian, yellow, and structureless.Local redeposition is suggested by the presence of stratification andextraclast sand and pebbles (Gyalog, 2004).

The mammoths occurred at a depth of ca. 170 cm withina leached, yellow loess unit (Fig. 2). Total loess thickness at the siteisw8 m, which is homogenous without any paleosol interbedding.The upper 4 m thick layer is rich in terrestrial molluscs, and dips1e3� towards Lake Balaton. The fossil-bearing layer is not typicalloess sediment (loamy loess), and it indicates a muddy depositionalenvironment. The surroundings of Lake Balaton were a marshlandat that time.

Fig. 1. Location of the mammoths remains near Csajág, Hungary.

L. Katona et al. / Quaternary International 255 (2012) 130e138 131

3. Materials and methods

The excavated bones (148 pieces), although in some casesbroken by the actions of the mechanical digger, were in extremelygood condition. Refitting pieces broken during excavation werejoined using Paleo-Bond glue. The remains are now conserved andexhibited at the Natural History Museum of Bakony Mountains,Zirc, Hungary (Fig. 3).

The palaeobiological analysis considered the gender of theremains using the tusks and postcranial material as a basis. Ageranges were determined on the basis of the dentition and tusks.Measurements were takenwith sliding calipers or flexible tape andquoted in millimeters. Bones were identified using an adult skel-eton of woolly mammoth (Mammuthus primigenius, MátraMuseum, Gyöngyös) for reference. The tusks were studied by OASAas described by Theodorou and Agiadi (2001). The broken, flatsurface of the tusk, as a transverse section was photographed inhigh-resolution, and the images were processed with a photoprocessing program. The Schreger angles were then measured athigh magnification. For the taphonomic analysis of the availablematerial, the isolated skeletal elements and articulated skeletalparts were counted, with consideration of the breakage, weath-ering, abrasion, and plant root traces on the bone material.

Pleistocene terrestrial gastropod fauna were used for palae-oenvironmental reconstruction. The section of loess was sampledat every 4 cm intervals for Quaternary malacological analyses,following sediment cleaning. Mollusc shells for malacologicalstudies were derived from sediment samples of 1 dm3 in all theprofiles in order to gain comparable results regarding the compo-sition of the fauna, and species and specimen numbers. Samplingwas carried out according to the methods advocated by Krolopp(1983). Sediment samples were screen-washed using a mesh of0.5 mm, and the retrieved shells were taxonomically identified.Teeth samples were submitted for radiocarbon (AMS) dating (Gli-wice Radiocarbon Laboratory) and the host sediment was dated byOSL method in the Geological Institute of Hungary.

Abbreviations: dex. ¼ right, sin. ¼ left; dPx ¼ lower deciduousmolar x, Mx ¼ lower permanent molar x; A ¼ adult; J ¼ juvenile;MIS ¼ Marine Isotope Stage; OSL ¼ optically stimulated lumines-cence; OASA ¼ optical analysis of the Schreger angles.

4. Radiometric dating

Loess samples from the fossil-bearing layer were dated in theGeological Institute of Hungary using OSL. The measured age is9970 � 890, which seems too young (Katona and Magyari, 2008).

Fig. 2. The excavation site in the road construction area.

L. Katona et al. / Quaternary International 255 (2012) 130e138132

A sample of tooth was sent to Gliwice Radiocarbon Laboratory for14C dating (AMS). A modified Longin method was used for theextraction of collagen (Longin, 1971), and graphitized samples weremeasured on a single-stage AMS instrument. Radiocarbon ages arequoted both uncalibrated and in calendar years, calibrated usingOxcal v4.1.7 software (Bronk Ramsey, 2009) and calibration curveIntcal09 (Reimer et al., 2009). The results are given in Table 1. Themeasurement gave an age of 13,315 � 35 BP which calibrates to16,249 � 413 calendar years ago from today. Calibration givesa 94.5% age probability of 16,865e15,915 cal BP (MIS 2).

5. Csajág mammoth finds

5.1. Skeletons

The major parts of the bone remains from Csajág are those of anadult mammoth. Most of the major long bones were recovered: leftand right scapulae, humeri, radii, ulnae, femora, tibiae and fibulae(Figs. 4 and 5). Juvenile bones are fewer than those of the adult, andrepresent only one juvenile individual. In addition, for both indi-viduals, the pelvic girdle, patellae, and about 30% of the small bonesof the front and back feet were recovered (Fig. 4). Of the skull, thelower jaws (with all molars) of the juvenile mammoth wererecovered in their entirety (Fig. 6C). Of the vertebral column, onlya third part was found for both mammoths. From the vertebralcolumn of the adult, only two (3rd, 4th) out of 7 cervical (neck)vertebrae, only 3 of 20 thoracics, 3 of 4 lumbars, and the completesacrum, as well as fragments of vertebrae were found. To these areadded 11 of 20 ribs and the presternum. The juvenile vertebralcolumn: the atlas and 4 of the thoracics (20the24th, Fig. 5D), andseveral ribs and fragments of ribs were found.

Missing from the recovered remains are most of the tail andabout half of the foot bones, and most notably, the entire bonycranium for both individuals. One lower right molar and the lefttusk of the adult mammoth are all that remains of the skull (Fig. 6A,B). Themolar is identifiable as anM2 in latewear. Both little tusks ofthe younger mammoth were found. In the juvenile mandible, boththe M1 is preserved in the alveolus behind the dP4 in wear, and 10(sin.) and 12 (dex.) enamel lamellae (plates) can be counted up to

the posterior part which is hidden by a thin bony capsule (theossified sheath around the enamel organ). This sheath hidesthe posteriormost plates of M1, but the length of the sheath behindthe last observable plate suggests that approximately two plates arehidden, giving a total of 14(?) plates (excluding talons).

The Schreger pattern (Trapani and Fisher, 2003; Ábelová, 2008)is clearly visible on the section surface of the adult tusk (Fig. 6b).TheSchreger outer angles range from 74� to 81� suggesting an attri-bution to Mammuthus. Measurement data for the teeth and majorbones are given in Tables 2e4.

5.2. Age determination and gender

TheM2molar of the adult mammoth is in latewear (Fig. 6A). Theanterior roots have been lost through wear. This indicates that thepreserved plate count of 12 was originally at least four higher.Applying the aging criteria for African elephants of Laws (1966),and normalizing for the greater number of enamel plates inmammoth molars compared to those of African elephants (Lister,1999) indicated an age at death of approximately 24e25 years.The juvenile teeth are dP4 andM1 (Fig. 6C). The dP4 in latewear, thefirst 4 plates in M1 are inwear. Thewidths of the left and right teeth(dP4) in this specimen are 53 and 54 mm. Its age is Laws’ categoryVII (6e7 years old).

The pelvic ratio (Lister, 1996) in the adult mammoth is 2.65; and2.69 for the Csajág juvenile. Based on body and tusk size, and pelvicratio the Csajág adult is a female. The pelvic ratio of the juvenilemammoth suggests that it is also a female.

5.3. Taphonomy

The almost complete skeletons, tusks and mandible of juvenile,together with the horizontal orientation of the slightly dissociatedpostcranial remains, indicate that the mammoths were buriedwhere they died. Stratigraphic and geomorphologic studies indicatethat thebonesweredeposited in avery lowenergyenvironment andare in their primary context, with no significant post-depositionalmovement.

Fig. 3. The Csajág mammoths, exhibited in the Natural History Museum of Bakony Mountains, Zirc, Hungary.

L. Katona et al. / Quaternary International 255 (2012) 130e138 133

The environment within which the mammoths died and werepreserved was an actively subsiding muddy pool (or marsh).Probably, the mud was sufficiently dense and tenacious to preventescape. The margins of the pool must have been sloping forslumping to have occurred, and for water to accumulate. Miring inthe loamy deposits, possibly combined with starvation if theanimals survived in this situation long enough, are likely factors.

Table 1Radiocarbon dating results of mammoth findings from Hungary and Central Europe for

Field number or location Sample Laboratory code Conventional 14C

Cs-1 (H) tooth* GdA-2011 13,315 � 35Zók-m1 (H) tusk* AA-80678 17,760 � 200SzegedeÖthalom (H) bone Deb-3344 15,916 � 168EsztergomeGyurgyalag (H) bone Deb-1160 16,160 � 200Austria bone GrN-6586 19,520 � 120Austria bone GrN-6585 19,340 � 100Czech Republic tusk* GrA-20002 14,820 � 120Czech Republic tusk* GdA-459 15,650 � 70Czech Republic bone GrN-14828 12,670 � 80Czech Republic tusk GrN-28450 11,270 � 80Poland tooth* Poz-10135 13,180 � 60Poland tooth* LuS-7739 12,585 � 70

Fractured edges of broken bones are sharp and exhibit no traces ofrounding. Traces of gnaw-marks are absent on the bones.

There is evidence that the skeletons had been at least partlyassociated, if not articulated, in the deposit. Series of six adultvertebrae (25the27th thoracic, 28the30th lumbar and sacrum) andfive juvenile (20the24th thoracic) were recovered in articulation(Fig. 5D). All themajor bones of both individuals recovered are from

comparison (ca. 23e13 ka cal BP), H ¼ Hungary, * ¼ AMS measurement.

age (BP) Calendar age 1s (68.3%) (cal BP) Reference

16,249 � 413 This paper21,250 � 450 Konrád et al., 201019,099 � 248 Krolopp et al., 199519,354 � 354 Vörös, 199123,313 � 343 Salcher-Jedrasiak et al., 201023,101 � 274 Salcher-Jedrasiak et al., 201018,148 � 285 �Skrdla et al., 200418,923 � 229 �Skrdla et al., 200414,939 � 173 Svoboda et al., 200013,173 � 112 Valoch and Neruda, 200516,108 � 392 Nadachowski et al., 201014,935 � 301 Arppe and Karhu, 2010

Fig. 4. Bones recovered of juvenile and adult individuals. The shaded parts show which bones were unearthed.

Fig. 5. Map of bones. A ¼ left femur and tibia of adult mammoth; B ¼ vertebral and pelvis parts of adult mammoth; C ¼ left and right scapulae of adult mammoth; D ¼ the20the24th thoracic vertebrae of juvenile mammoth; 36 ¼ mandible, 40e41 ¼ dorsal spines of thoracic vertebra.

L. Katona et al. / Quaternary International 255 (2012) 130e138134

Fig. 6. The mandible, teeth and tusk of the Csajág mammoths. A ¼ occlusal view of adult M2 tooth; B ¼ polished section of the adult tusk, b ¼ section of the tusk showing theSchreger pattern; C ¼ occlusal and left lateral view of the juvenile mandible.

L. Katona et al. / Quaternary International 255 (2012) 130e138 135

the left side. It is possible that the mammoths were lying on theirleft sides, which had sunk into themud, so that the right-side boneswere exposed and subject to erosion or removal.

There is no clear evidence that the construction work did notcause damage to the skeletons. The road construction workexposed the fossil remains and probably some bones disappearedby accident. The removed cover sediment was searched for boneswithout success. The rescue excavation had enough time to recoverthe skeletons in the site.

The failure to recover the adult skull and the juvenile cranium,whereas the atlas vertebra and mandible of the juvenile werefound, is interesting. The only finds relating to an adult mammothskull are the left tusk and a left lower tooth. It cannot be known forcertain whether the two preserved mammoths became entrappedat roughly the same time. The age and sex estimates on the indi-viduals make it likely they were a family, mother and daughter. The

juvenile may entered into the muddy pool and her mother tried torescue her.

5.4. Palaeoenvironment of the Csajág mammoths

From early studies, it has been accepted that mammoths existedunder conditions of cold, dry climate and preferred open steppes,locally with trees and bushes (Kahlke, 1999; Lister and Sher, 2001;Velichko et al., 2002). During the glacial stages of the Pleistocene,the Pannonian basin was subject to a cryogenic environment thatproduced a variety of periglacial features (Kovács et al., 2007). Fromthe loess, 19 gastropod species and 735 specimens were collectedand identified. Among these Succinea oblonga (Draparnaud, 1801)and Pupilla muscorum (Linnaeus, 1758) were the dominant speciesfrom the mammoth bearing layer, both Holarctic, hygrophilous/mesophilous, cold-resistant, steppe fauna (Sümegi and Krolopp,

Table 2Dental measurement of mammoth tusks.

Measurementsof the tusks (cm)

A Incisivussin. (2007.1.61)

J Incisivussin. (2007.1.47)

J Incisivusdex. (2007.1.46)

Maximum length,measured alongthe outer curvature

116.0 37.0 26.5

Maximum length,measured alongthe inner curvature

100.0 33.3 24.5

Maximum diameterat the end of thealveolus

7.7 4.93 4.57

Table 4Measurements of mammoth limb bones, scapulae, and sternum.

Specimen number Description Length (cm)

2007.1.6 A tibia (sin.) 52.02007.1.140 A fibula (sin.) 34.02007.1.50 A femur (sin.) 97.52007.1.12 A humerus (sin.) 54.02007.1.68 A radius (sin.) 60.52007.1.67 A ulna (sin.) 95.02007.1.35 J femur (sin.) 57.02007.1.34 J ulna (sin.) 63.52007.1.59 J ulna (dex.) - broken 32.02007.1.70 A scapula (sin.)a 70.02007.1.69 J scapula (sin.)a 46.52007.1.53 J scapula (dex.)a 38.52007.1.99 A presternum 19.5

a Total length parallel to spine (including dorsal epiphysis).

L. Katona et al. / Quaternary International 255 (2012) 130e138136

2002). The fauna suggest a relatively mild and humid climate, witha mean July palaeotemperature of 15 �C.

According to Rudner and Sümegi (2001), two cool treeless-steppe periods existed between 26 and 12 ka BP. In this perioda ‘microinterstadial’ was described by Rudner and Sümegi (2001)dated to 18e16 ka BP and indicated by biogeochemical proxies(plant-derived biomarkers and stable C and N isotopes) by Schatzet al. (2011). This short phase of warmer, wetter climate wascharacterized by an increase in the amount of intermediate andwoodland-dwelling molluscs, indicating the development ofa forest-steppe environment with significantly more trees than inthe preceding periods.

Based on these data, the landscape in the Pannonian basin was:extremely continental arid climate with a small amount of solidprecipitation; firm soil; and dominance of open plant communities(herb, grass and low shrubs), locally with trees. However, therewere also intervals (e.g. 18e16 ka BP) in which trees and shrubsexpanded, a tree-steppe or ‘open parkland’ environment (Schatzet al., 2011).

6. Chronological significance

Woollymammoths returned to central Europe by ca. 19 ka cal BP(Stuart et al., 2004). Their presence in Europe after this date, rela-tive to the large number of mammal-bearing (mainly archaeolog-ical) sites, is sporadic compared to MIS 3 (Lister, 2009). The adulttooth dates extended into the end of Greenland Stadial 2 (MIS 2).This data are tabulated, together with others from central and east-central Europe post- 12,500 14C BP (ca. 15,000 cal BP), in Table 1. TheCsajág date of 13,315 � 35 BP (16,249 � 413 cal BP) is important inproviding the latest secure record of mammoths in Hungary. Thenew date on adult tooth is the youngest mammoth date from

Table 3Dental measurements of the molars.

Measurements M2 (adult) dP4 sin. dP4 dex. M1 sin. M1 dex.

Specimen number 2007.1.63 2007.1.60a 2007.1.60a 2007.1.60a 2007.1.60a

Crown width 71 53.7 53.6 e e

Crown length 131 105 103 e e

Height (unworntrue maximum)

75 e e e e

Plate width 6 5 5 e e

Enamel thickness 1.2 0.9 0.9 0.9 0.9Cement width

between plates5 5 5 5 5

Plate formula(lamellae)

�12x x11x x11x x14x x14x

Plates in use 9 all all 3 5Plate (lamellar)

frequency10.9 12 12 e e

a specimen number of juvenile mandible.

Hungary; the others are an older date of 15,916 � 168 BP(19,099 � 248 cal BP) from SzegedeÖthalom (Krolopp et al., 1995),and an almost similar date of 16,160 � 200 (19,354 � 354 cal BP)from EsztergomeGyurgyalag (Vörös, 1991). However, these areboth only pieces of bones from Late Upper Gravettian (Epi-gravettian) culture layer (Dobosi, 2005). The Csajág remains(together with Zók remains, 21,250 � 450 cal BP; Konrád et al.,2010), skeletons in stratified context, provide clear evidence thatthe mammoth was native to Hungary at this time. Close to theCsajág site (ca. 30 km to the south) a Gravettian reindeer hunters’campsite were discovered at Ságvár, dated to 21e20 ka cal BP(Vörös,1982). Elsewhere in central and east-central Europe, recordsfrom the Czech Republic and Poland (Table 1) indicate the broaderdistribution of mammoths in the first half of Greenland Intersta-dial-1. Mammoths do not appear to survive in Europe into the laterpart of GI-1, probably because the spread of Allerød forests drasti-cally reduced their grassland habitat (Lister, 2009).

6.1. Central and east central European extent of woolly mammothduring the Late Pleistocene

In the Last Glacial Maximum (LGM), the mammoth was widelydistributed throughout most of Europe except the Iberian Penin-sula, the Apennine Peninsula, the Balkan Peninsula, and the CrimeaPeninsula (Álvarez-Lao et al., 2009; Markova et al., 2010). Thelowlands of Lower AustriaeMoraviaeSouth Poland form animportant natural corridor in Central Europe, allowing migrationsof both animals and humans between the Danube valley and theNorth European Plain (Svoboda et al., 2005). Woolly mammothprobably spread into Hungary from the North European Plainthrough this corridor. The first dated appearance of M. primigeniusis ca. 50e36 ka in Hungary (Vörös, 1981).

The remains of woolly mammoth from Poland have beendiscovered mainly from Middle and Upper Palaeolithic archaeo-logical sites (24.7e12.5 ka 14C BP) of Spadzista Street (B, C, E and F)and Zwierzyniec on the St. Bronis1awa hill in Kraków, andDzier _zys1aw near Kraków and represent material found nearly insitu (Wojtal and Sobczyk, 2005). The radiocarbon dates ofmammoths range from ca. 52 ka to 12.5 ka 14C BP and can becorrelated with MIS 3 and MIS 2. The severe climatic circumstanceslasting in conventional radiocarbon dates from ca. 22 ka BP to ca.17 ka BP probably reduced the number of animals or even causedthe withdrawal of woolly mammoth from the area north from theSudetes and the Carpathian Mountains for 4e5 millennia. Only onedate ca. 20.3 ka BP (ca. 24.2 cal ka BP) is available from this time-span (Nadachowski et al., 2010, in press). Mammuthus primigeniusreappeared in southern Poland ca. 14.6 ka BP (ca. 17.9 cal BP) but

L. Katona et al. / Quaternary International 255 (2012) 130e138 137

soon had disappeared from this region because of marked reduc-tion in open habitats at the beginning of Late Glacial Interglacialwarming (Greenland Interstadial I or Bølling and Allerød). Thelatest available record from Poland is from Dzier _zys1aw:12,585 � 70 BP (14,935 � 301 cal BP; Arppe and Karhu, 2010).

The pattern of mammoth site location in Czech Republic seemsto be related to the exploitation of large mammals following theriver valleys, and this presupposition is also supported by themammoth bone deposits, located either inside the prehistoricsettlements or in the adjacent side gullies, or, as individual pieces,scattered in the river floodplain. The largest dumps were found atP�redmostí, Dolní V�estonice, Milovice with the age range of30e21 ka 14C BP (Svoboda et al., 2005). The existence of the woollymammoth in Moravia is significant in that, together with southernPoland, it represents the youngest record of Late glacial mammoths.A fossil tusk from K�ulna, Czech Republic, has been recently dated to11,270 � 80 (13,173 � 112 cal BP) which seems to be the youngestrecord from Central Europe (Valoch and Neruda, 2005).

Woolly mammoth is known as hunted game from Gravettiansites in Slovakia (Vla�ciky, 2005). Direct radiocarbon dating ofa mammoth tusk is only recorded from Liptovský Mikulá�s(20,000 � 1000 BP; Holec and Kernátsová, 1997), althoughmammoth finds are common in the country (Prista�s and Schmidt,1977; Vörös, 1983; Vla�ciky, 2005).

The Epiaurignacian Langmannersdorf locality in the W-Eoriented Perschling valley in Austria is broadly agreed to be theyoungest known residential camp site with abundant mammothbones within the Middle Danube region (Salcher-Jedrasiak et al.,2010). The Full Glacial age at around 20.6 ka BP is confirmed byfour dates from mammoth bones: 20,130 � 80 BP (OxA-16318),19,520 � 120 BP (GrN-6586), 20,580 � 170 BP (GrN-6659), and20,590 � 110 BP (GrN-16567). In the 18.2 ka Epigravettian Grub-graben site (Lv-1660: 18,170� 300; GrN-21893: 18,820 � 160) onlyvery rare mammoth bones have been observed (unpublishedanalysis by Fladerer; Terberger and Street, 2002). Mammoth bonesaccumulations are known from other sites along the Danube andfrom cave deposits as well (Döppes and Rabeder, 1997; Fladerer,2000).

The mammoths migrated to the south (former Yugoslavia) andeast (Transylvania; Romania) from the Pannonian basin (Hungary).In the territories of the former Yugoslavia many fossil proboscideanremains have been excavated (Vörös, 1983; Malez, 1986). The mostnumerous findings are isolated teeth (especially in Slovenia; Poharand Recek, 1995), but other postcranial skeletal remains also exist(Lenardi�c and Poje, 1994; Pohar and Recek, 1995). Some knownsites in Croatia are Vinkovci (eastern Slavonia) and Viljeva�cki �Cretnear Donji Miholjac (Lenardi�c and Poje, 1994). In these profiles thedeposits are Late Pleistocene terrestrial and lake-swampy loess. Theremains of mammals including woolly mammoths are typicalrepresentatives of the Upper Palaeolithic or Hazaran complexfound in the loess of Srem, Titel Plateau and near Belgrade(Markovi�c et al., 2006; Nenadi�c et al., 2009). Upper Pleistocenelarge mammal remains were found in cave deposits of southernSerbia and Montenegro (central Balkans) as well (Dimitrijevi�c,1997; Forsten and Dimitrijevi�c, 2003).

The wooly mammoth is by far the commonest large herbivorefound in the Romanian Pleistocene formations (Vörös,1983; Codreaand M�arginean, 2007). It is reported from several dozen sites,located in Transylvania and on the Moesian Platform. The majorityoriginate from loess and loess-like deposits, as well as from graveland sand accumulated in various river terraces (Munteanu et al.,2008). Several teeth (tusk fragments and cheek teeth) wereunearthed from the Danube riverbed, not far from C�al�arasi, due tothe drag works commonly carried on for deepening the riverchannel. All belong to evolved wooly mammoth specimens

(Curlisc�a and Codrea, 2007). A half mandible found in southernDobrogea at Ostrov, on the Danube right bank terrace gravel, inPleistocene deposits (Codrea, 2008). Ostrov is a new locality for thisspecies in this region (Codrea, 2008).

7. Conclusions

The discovery of almost complete skeletons from a stratified,dated context contributes strong evidence for the survival ofmammoths in the Pannonian basin and Central Europe. Theoccurrence ofMammuthus primigenius between 50 and 16 cal ka BPin Hungary was not a singular biogeographic event. During thistime-span, the animals repeatedly occupied steppe-like landscapesof the Carpathian (Pannonian) basin, when sufficient amounts ofsuitable plants allowed their subsistence. The Csajág mammothsdid not differ morphometrically from individuals of contempora-neous populations of other parts of Europe, but represent thecentral part of a continuous, Holarctic extended belt of mammothdistribution.

The central European migrations of woolly mammoths correlatewith periods of particularly dry and cold climatic conditions, asdocumented in terrestrial sediment sequences of the region.Moreover, the temporal coincidence of the central Europeanoccurrences of Late Pleistocene M. primigenius in Europe with coldclimate phases in Greenland Stadial 2 (see Section 6) classifies theCsajág mammoth finds as evidence of transregional palae-oecological processes.

New radiocarbon evidence fits into thewell-known colonizationpattern of Mammuthus primigenius in Central Europe and confirmsa continuous distribution in the Würm/Weichselian Late-glacial ca.16.9e15.9 ka cal BP, within Greenland Stadial 2. The Morava Rivervalley runs in a roughly south-to-north direction, connecting, viathe Danube and Tisa River valleys, the Hungarian Pannonian Plainthrough Serbian via the Vardar River valley in Macedonia tonorthern Greece in the south. The large mammals probably usedthis route for dispersal in the Late Pleistocene.

Acknowledgements

WethankFerenc Szántai, the original discovererof thebones; thestaff of the Bakony Museum for their help with excavation andpreparation of the remains. Our thanks are also due to Árpád Mag-yari and Edit Thamó-Bozsó (Geological Institute of Hungary) forsupervising the OSL dating. Special thanks to Adam Nadachowski(Poland), Florian Fladerer (Austria), Martina Moravcová (Slovakia),and JadrankaLenardi�c (Croatia) forup-to-date informationofwoollymammoth research and data in their countries. Themanuscript wasimproved by helpful comments from Ralf-Dietrich Kahlke and ananonymous reviewer.

References

Ábelová, M., 2008. Schreger pattern analysis of Mammuthus primigenius tusk:analytical approach and utility. Bulletin of Geosciences 83, 225e232.

Álvarez-Lao, D.J., Kahlke, R.-D., García, N., Mol, D., 2009. The Padul mammoth findse On the southernmost record of Mammuthus primigenius in Europe and itssouthern spread during the Late Pleistocene. Palaeogeography, Palae-oclimatology, Palaeoecology 278, 57e70.

Arppe, L., Karhu, J.A., 2010. Oxygen isotope values of precipitation and the thermalclimate in Europe during the middle to late Weichselian ice age. QuaternaryScience Reviews 29, 1263e1275.

Bronk Ramsey, C., 2009. Bayesian analysis of radiocarbon dates. Radiocarbon 51,337e360.

Codrea, V., M�arginean, R., 2007. A catalogue of fossil vertebrates from Auid NaturalSciences Museum. Oltenia. Studii si comunic�ari. Stiintele Naturii 23, 177e186.

Codrea, V., 2008. Wooly mammoth scattered findings from the Danube riverbednear C�al�arasi. Oltenia. Studii si comunic�ari. Stiintele Naturii 24, 267e270.

Curlisc�a, A., Codrea, V., 2007. A new mammoth finding in Dobrogea. Oltenia. Studiisi comunic�ari. Stiintele Naturii 23, 187e189.

L. Katona et al. / Quaternary International 255 (2012) 130e138138

Dimitrijevi�c, V., 1997. Upper Pleistocene mammals from cave deposits of Serbia.Geolo�ski anali Balkanskoga poluostrva 61, 179e370.

Dobosi, V.T., 2005. Cadastre of palaeolithic finds in Hungary. CommunicationesArchaeologicae Hungariae 2005, 49e81.

Döppes, D., Rabeder, G., 1997. Pliozäne und pleistozäne Faunen Österreichs. Mit-teilungen der Kommission für Quartärforschung. Österr. Akad. Wiss., Wien.

Fladerer, F.A., 2000. Late Quaternary Vertebrate Taphocoenoses from Cave Depositsin Southeastern Austria: Responses in a Periglacial Setting. In: GeologicalSociety, London, Special Publications, 181 199e213.

Forsten, A., Dimitrijevi�c, V.M., 2003. Pleistocene horses (genus Equus) in the centralBalkans. Geolo�ski anali Balkanskoga poluostrva 65, 55e75.

Gyalog, L., 2004. Pannonian and Quaternary formations in the detailed maps of theBalatonf}o Area. In: Gyalog, L., Horváth, I. (Eds.), Geology of the Velence Hills andthe Balatonf}o. Geological Institute of Hungary, Budapest, pp. 217e218.

Holec, P., Kernátsová, J., 1997. Cicavce (Mammalia) a ulitníky (Gastropoda) vrchnéhopleistocénu mladopaleolitického táboriska v Tren�cianskych Bohuslaviciach.Mineralia Slovaca 29, 234e236.

Kahlke, R.-D., 1994. Die Entstehungs-, Entwicklungs- und Verbreitungsgeschichtedes Oberpleistozänen Mammuthus-Coelodonta-Faunencomplexes in Eurasien(Großsäuger). Abandlungen Senckenbergischen Natuforschenden Gesellschaft546, 1e64.

Kahlke, R.-D., 1999. The History of the Origin, Evolution and Dispersal of the LatePleistocene Mammuthus-Coelodonta Faunal Complex in Eurasia (LargeMammals). Fenske Companies, Rapid City, p. 219.

Katona, L., Kovács, J., Kordos, L., Linkai, I., Magyari, Á, 2010. Two young woollymammoths findings near Lake Balaton, Hungary. Quaternaire Hors-série 3,75e77.

Katona, L., Magyari, Á, 2008. Csajági mamutok. Természet Világa 139, 34e35.Konrád, G., Kovács, J., Halász, A., Sebe, K., Pálffy, H., 2010. Late Quaternary woolly

mammoth (Mammuthus primigenius Blum.) remains from south Transdanubia,Hungary. Comptes Rendus Palevol 9, 47e54.

Kovács, J., Fábián, S.Á., Schweitzer, F., Varga, G., 2007. A relict sand-wedgepolygon site in north-central Hungary. Permafrost Periglacial Processes 18,379e384.

Krolopp, E., 1983. Biostratigraphic division of Hungarian Pleistocene formationsaccording to their Mollusc fauna. Acta Geologica Hungarica 26, 62e89.

Krolopp, E., Sümegi, P., Kuti, L., Hertelendi, E., Kordos, L., 1995. Palaeoecologicalreconstruction of formation of the Szeged-Öthalom area loess formations.Földtani Közlöny 125, 309e361.

Laws, R.M., 1966. Age criteria for the African elephant, Loxodonta a. africana. EastAfrican Wildlife Journal 4, 1e37.

Lenardi�c, J., Poje, M., 1994. Elephant remains and attendant malacological faunafrom some localities in Slavonia, Croatia. Neogene and quaternary mammals ofthe palaearctic. In: Conference in Honour of Professor Kazimierz Kowalski.Krakow, Poland, p. 41.

Lister, A.M., 1996. Sexual dimorphism in the mammoth pelvis: an aid to genderdetermination. In: Shoshani, J., Tassy, P. (Eds.), The Proboscidea: Trends inEvolution and Paleoecology. Cambridge University Press, Cambridge, pp.254e259.

Lister, A.M., 1999. Epiphyseal fusion and postcranial age determination in thewoolly mammoth, Mammuthus primigenius (Blum. Deinsea 6, 79e88.

Lister, A.M., 2009. Late-glacial mammoth skeletons (Mammuthus primigenius) fromCondover (Shropshire, UK): anatomy, pathology, taphonomy and chronologicalsignificance. Geological Journal 44, 447e479.

Lister, A.M., Sher, A.V., 2001. The origin and evolution of the woolly mammoth.Science 294, 1094e1097.

Lister, A.M., Sher, A.V., van Essen, H., Wei, G., 2005. The pattern and process ofmammoth evolution in Eurasia. Quaternary International 126e128, 49e64.

Longin, R., 1971. New method of collagen extraction for radiocarbon dating. Nature230, 241e242.

Malez, M., 1986. Die quartären Vertebraten-Faunen in der SFR Jugoslawien. Quar-tärpaläontologie 6, 101e117.

Markova, A.K., Puzachenko, A.Y., van der Plicht, J., van Kolfschoten, T.,Ponomarev, D.V., 2010. New data on the dynamics of the mammoth Mammu-thus primigenius distribution in Europe in the second half of the Late Pleisto-ceneeHolocene. Doklady Earth Sciences 431, 479e483.

Markovi�c, S., Oches, E., Sümegi, P., Jovanovi�c, M., Gaudenyi, T., 2006. An introductionto the Upper and Middle Pleistocene loess-paleosol sequences of Ruma section(Vojvodina, Serbia). Geologica Carpathica 56, 545e552.

Munteanu, M.T., Munteanu, E., Stiuc�a, E., Macalet, R., Dumitrascu, G., 2008. Someaspects concerning the quaternary deposits in South Dobrogea. Acta Palae-ontologica Romaniae 6, 229e236.

Nadachowski, A., Lipecki, G., Wojtal, P., 2010. Radiocarbon chronology of woollymammoth (Mammuthus primigenius) from Poland. Geophysical ResearchAbstracts 12, EGU2010e3646.

Nadachowski, A., Lipecki, G., Wojtal P., Miekina, B. Radiocarbon chronology ofwoolly mammoth (Mammuthus primigenius) from Poland. Quaternary Inter-national, in press.

Nenadi�c, D., Kne�zevi�c, S., Bogi�cevi�c, K., 2009. Stratigraphical and palaeogeographicalcharacteristics of Pleistocene series in the Sava riparian area at Belgrade(Serbia). Bulletin of the Natural History Museum in Belgrade 2, 63e80.

Pohar, V., Recek, H., 1995. Zobje fosilnih slonov iz osteolo�ske zbirke Oddelka zageologijo Univerze v Ljubljani. Geologija 37/38, 63e85.

Prista�s, J., Schmidt, Z., 1977. Find of Mammuthus primigenius (Blumenbach) 1799from the environs of Chrámec (Rimavská kotlina basin, Southern Slovakia).Západné Karpaty 2/3, 241e248.

Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., BronkRamsey, C., Buck, C.E., Burr, G., Edwards, R.L., Friedrich, M., Grootes, P.M.,Guilderson, T.P., Hajdas, I., Heaton, T.J., Hogg, A.G., Hughen, K.A., Kaiser, K.F.,Kromer, B., McCormac, F.G., Manning, S.W., Reimer, R.W., Richards, D.A.,Southon, J., Turney, C.S.M., van der Plicht, J., Weyhenmeyer, C., 2009. IntCal09and Marine09 radiocarbon age calibration curves, 0e50,000 years cal BP.Radiocarbon 51, 1111e1150.

Rudner, E., Sümegi, P., 2001. Recurring Taiga forest-steppe habitats in the Carpa-thian basin in the Upper Weichselian. Quaternary International 76/77, 177e189.

Salcher-Jedrasiak, T., Nývltová Fi�sáková, M., Fladerer, F.A., 2010. New warm seasondata from the full-glacial mammoth hunters camp site at Langmannersdorf ander Perschling (Lower Austria). Quaternaire Hors-série 3, 142e145.

Schatz, A.-K., Zech, M., Buggle, B., Gulyás, S., Hambach, U., Markovic, S.B., Sümegi, P.,Scholten, T., 2011. The Late Quaternary loess record of Tokaj, Hungary: recon-structing palaeoenvironment, vegetation and climate using stable C and Nisotopes and biomarkers. Quaternary International 240 (1-2), 52e61.

�Skrdla, P., Nývltová Fi�sáková, M., Sedlá�cková, L., Zapletalová, D., 2004. Paleolithic,Brno (k. ú. �Stý�rice, okr. Brno-m�esto). P�rehled výzkum�u 46, 173e177.

Stuart, A.J., 2005. The extinction of woolly mammoth (Mammuthus primigenius)and straight-tusked elephant (Palaeoloxodon antiquus) in Europe. QuaternaryInternational 126/128, 171e177.

Stuart, A.J., Kosintsev, P.A., Higham, T.F.G., Lister, A.M., 2004. Pleistocene to Holoceneextinction dynamics in giant deer and woolly mammoth. Nature 431, 684e689.

Sümegi, P., Krolopp, E., 2002. Quatermalacological analyses for modeling of theUpper Weichselian palaeoenvironmental changes in the Carpathian Basin.Quaternary International 91, 53e63.

Svoboda, J., Horá�cek, I., Lo�zek, V., Svobodová, H., �Silar, J., 2000. The Pekárna Cave.Magdalenian stratigraphy, environment, and the termination of the loessformation in the Moravian Karst. Anthropozoikum 24, 61e79.

Svoboda, J., Péan, S., Wojtal, P., 2005. Mammoth bone deposits and subsistencepractices during Mid-Upper Palaeolithic in Central Europe: three cases fromMoravia and Poland. Quaternary International 126/128, 209e221.

Terberger, T., Street, M., 2002. Hiatus or continuity? New results for the question ofpleniglacial settlement in Central Europe. Antiquity 76, 691e698.

Theodorou, G.E., Agiadi, K., 2001. Observations on the microstructure of fossil tusksfrom the Charkadio cave (Tilos, Dodecanese, Greece). The World of Elephants.In: Proceedings of the 1st International Congress. Consiglio Nazionale delleRicerche, Rome, Italy, pp. 663e667.

Trapani, J., Fisher, D.C., 2003. Discriminating proboscidean taxa using features of theSchreger pattern in tusk dentin. Journal of Archaeological Science 30, 429e438.

Valoch, K., Neruda, P., 2005. K chronologii moravského magdaleniénu. On thechronology of the Moravian Magdalenian. Archeologické rozhledy 57, 459e476.

Velichko, A.A., Catto, N., Drenova, A.N., Klimanov, V.A., Kremenetski, K.V.,Nechaev, V.P., 2002. Climate changes in east Europe and Siberia at the LateglacialeHolocene transition. Quaternary International 91, 75e99.

Vla�ciky, M., 2005. Mamuty (Proboscidea, Mammalia) z mladopaleolitického sídliskav Tren�cianskych Bohuslaviciach. Master thesis, Archive of DGP, Príf UK, Bra-tislava, 1e155.

Vörös, I., 1981. The skulls of mammoth in Hungary. Fragmenta Mineralogica etPalaeontologica 10, 97e106.

Vörös, I., 1982. Faunal remains from the Gravettian reindeer hunters’ campsite atSagvar. Folia Archaeologica 33, 43e71.

Vörös, I., 1983. Elephantiden-Reste aus dem Karpathenbecken. Fragmenta Miner-alogica et Palaeontologica 11, 61e84.

Vörös, I., 1991. Large mammal remains from the Upper Palaeolithic site at Eszter-gom-Gyurgyalag. Acta Archaeologica Hungarica 43, 261e263.

Wojtal, P., Sobczyk,K., 2005.Manandwoollymammothat theKrakówSpadzistaStreet (B)e taphonomy of the site. Journal of Archaeological Science 32, 193e206.