8.7 Koufos et al.*

49

George D. Ko u fo s , Katerina V.Va s s i l i a d o u , Kalliopi K. Koliadimou & George E. S y r i d e sAristotle University of T h e s s a l o n i k i

Koufos, G.D., Vassiliadou, K.V., Koliadimou, K.K. & Syrides, G.E., 2001 - Early Pleistocene smallmammals from Marathoussa, a new locality in the Mygdonia basin, Macedonia, Greece - DEINSEA 8 :49-102 [ISSN 0923-9308]. Published 09 November 2001

The fossiliferous site of Marathoussa (Mygdonia basin, Macedonia, Greece) is located in thePlatanochori Formation, which consists mainly of lacustrine sediments. The site includes one level withmicromammals and several levels with fresh-water mollusks. The micromammalian fauna containsseveral species of insectivores and rodents. The faunal elements suggest an early Pleistocene age.Detailed morphological and metrical comparisons of the arvicolids allow the dating of the fauna to theearly Biharian and, more precisely, to the Betfia Phase.

Correspondence: Aristotle University of Thessaloniki, Lab. of Geology and Palaeontology, 54006Thessaloniki, Greece; e-mail: [email protected]

Key words: Pleistocene, Insectivora, Rodentia, biochronology, Macedonia, Greece

I N T RO D U C T I O NThe paleontological and stratigraphicalresearch in the Mygdonia basin started at thebeginning of the eighties by a team led byProf. G.D. Koufos. Since then, several fossi-liferous sites have been found, while a greatamount of fossils has been unearthed. T h i smaterial has been studied and, besides twoPh.D. theses, a number of papers have beenpublished (for the bibliography see Koufos &Kostopoulos 1997). A new fossiliferous siteincluding mammals and mollusks was foundin 1994 by one of the authors (G.E.S.); thelocality is named Marathoussa (MAR). Ourfirst collection from the various beds of Mara-thoussa section indicated the presence of a richmicromammalian fauna in one of them. Collect-ing was continued until 1997 and the collectedfossils are described in the present a r t i c l e .

The material has been measured on a ZeissStereoscope (Dept of Biology, University ofThessaloniki) and on a Reflex Microscope 3-D (Dept of Geology, University ofUtrecht). Measurements are given in mm,unless something else is indicated. T h enomenclature and the way of measuring ofthe Soricidae are given in Reumer (1984), ofthe Muridae in Van de Weerd (1976) andMartin Suarez & Freudenthal (1993) respecti-v e l y, of the Cricetidae in Mein & Freudenthal(1971) and Daams & Freudenthal (1988)r e s p e c t i v e l y, of the Arvicolidae in Van derMeulen (1973), of the Sciuridae in De Bruijn(1966) and Van de Weerd (1976), of theZapodidae in Green (1977). All the materialis stored in the Laboratory of Geology andP a l a e o n t o l o g y, University of T h e s s a l o n i k i .

E a r ly Pleistocene small mammals fro mM a r a t h o u s s a , a new locality in the Mygdoniab a s i n , M a c e d o n i a , G re e c e

GEOLOGY AND STRAT I G R A P H YThe Mygdonia basin is an east to west-north-west trending inland tectonic depression inthe area of Central Macedonia, Greece. T h epre-Neogene basement consists of crystalline,mainly metamorphic rocks of the CircumRhodope belt and the Serbomacedonian mas-sif, as well as granitic rocks. The Neogene-Quaternary sedimentary deposits of the basinwere divided in two Groups, thePremygdonian and the Mygdonian Group.The Premygdonian Group corresponds to thetime interval from the Middle Miocene to theEarly Pleistocene. During the ?MiddleMiocene an inland tectonic depression (thePremygdonian basin) was formed and gradu-ally filled up by fluvial, continental, lacustri-ne and fluvio-lacustrine clastic sediments upto ~350 m thick (Psilovikos 1977). T h ePremygdonian Group has been divided inthree formations: the Chrysavgi, Gerakarou,and Platanochori Formations, respectively(Koufos et al. 1995).

The Mygdonian Group was deposited duringthe Middle Pleistocene to Holocene. Duringthe end of the Early Pleistocene, a new inlandtectonic depression (the Mygdonia basin) wasformed into the central part of the formerPremygdonian basin. The new basin was fil-led up with fresh water and an extensive lake(Mygdonia lake) was formed. Mygdonia lakegradually dried up and the present lakes ofKoronia and Volvi are the relics of it.Lacustrine and fluvio-lacustrine sediments upto ~150m thick were deposited during the

Middle Pleistocene to Holocene, forming theMygdonian Group. Several fossiliferous siteshave been recognized in the basin. The fossi-liferous sites include macro- and micro-mam-malian fossils and fresh-water mollusks(Koufos et al. 1992, Koliadimou 1996,Kostopoulos 1996). The fossil faunas are richand allowed the dating of the formations.The Chrisavgi Formation is the oldest oneand has been dated to the end of the MiddleMiocene (end MN 7+8). The upper part ofthe Gerakarou Fm has been dated to LatePliocene - Early Pleistocene and thePlatanochori Fm to the end of the EarlyPleistocene (Koufos & Kostopoulos 1997).

L O C A L I T YThe locality of Marathoussa is situated 3 kmN W of the village of Marathoussa, in theMygdonia basin (Fig.1A). The locality issituated in the Platanochori Fm, which con-sists mainly of lacustrine deposits. T h eMarathoussa site is close to the locality ofApollonia (APL) (Koufos et al. 1992). A s t r a-tigraphical correlation of the two sections isnot possible, due to a very dense network offaults. The locality includes a rich micro-mammalian fauna in the bed 5, while thereare several beds with fresh-water mollusks(Fig.1B). The mollusks were unknown in theMygdonia deposits. The first indication abouttheir presence comes from the locality ofApollonia (Koufos et al. 1992). T h eMarathoussa molluskan fauna includes thefollowing species: Gastropoda: Va l v a t a s p . ,B i t h y n i a sp., Ly m n a e a sp., P l a n o r b i s c f .

50

DEINSEA 8, 2001

Figure 1 A. S ketch map of Mygdonia basin with the fo s s i l i ferous sites. B. S t ra t i g raphical column of the Marathoussa

s e c t i o n . L i t h o s t ra t i g raphical legend of the beds: 2 2. S o i l ; 2 1. Sand with subangular quartzitic pebbl e s ; 2 0. Greyish sands with i n t e r c a-

lations of grey-green silts and clay s ; 1 9.Ye l l owish coarse cross-bedded sands; 1 8. Grey-green silty sand; 1 7. Grey marly sandstone;

1 6. Grey-green silt/clay with nodules of grey marly sandstone; 1 5. Grey-greenish silty sand; 1 4. Grey-greywhite coarse cross-bed-

ded sands with small sandstone concretions in the lower part ; 1 3. B r ownish grey silty sand; 1 2. Grey coarse sand with grave l s ;

1 1. Grey thin bedded silty sand; 1 0. B r ownish grey massive silt/clay ; 9. G r e y - ye l l owish silty sand; 8. Grey cross-bedded coarse sands

with grave l s , more dark coloured at the base; 7. B r ownish green silt/clay with numerous fragments of fresh-water mollusks;

6. Intercalations of thin (5-20cm) beds of grey nodular marl with dark grey-green massive silt/clay. Four marly intercalations appear

in the lower part ; 5. B r ow n i s h - black silts/clays with numerous fragments of micromammals and fresh-water mollusks; 4. G r e y - g r e e n-

ish fo s s i l i ferous (mollusks) fine silty-clayey sands. In the upper part there are nodules of marly sandstone; 3. B r own silt/clay with

fresh-water mollusks; 2. B r ownish green silty-clayey sands with mollusks; 1.Grey-green silt/sand with mollusks.

51

KOUFOS et al.: Pleistocene small mammals from Marathoussa

A

B

52

DEINSEA 8, 2001

p l a n o r b i s, P l a n o r b i s sp., A r m i g e r a ff. c r i s t a,G y r a u l u s a ff. a l b u s, S e g m e n t i n a s p . ,A c ro l o x u s sp. B i v a l v i a: P i s i d i u m sp., pearlyfragments of Unionidae indet., Ostracodesind., in addition to Charophytes indet. andcalcareous fragments of egg shells, probablyof birds. The lithostratigraphy of theMarathoussa section is given in Fig.1B

S YS T E M ATIC PA L A E O N TO L O G YOrder: Insectivora BO W D I C H, 1821Family: Talpidae GR AY, 1825Subfamily: Desmaninae TH O M A S, 1912Desmaninae indet.

M a t e r i a l :P4, 1 sin.: 130.M1, 2 dex.: 35, 117. 1 sin.: 493.M2, 2 dex.: 503, 510. 1 sin.: 525.M3, 3 dex.: 504, 524, 526.p2, 1 dex.: 551.p3, 3 sin.: 133, 550, 552.p4, 1 dex.: 487.m1, 1 sin.: 42.m2, 1 sin.: 486.m3, 2 dex.: 40, 456. 2 sin.: 41, 39.Mandible: 484, 487.

M e a s u re m e n t s : see Tables 1 and 2.

D e s c r i p t i o n :P 4 (Plate 1, 1): It is strongly molarized andsub-triangular in shape. The paracone is highand the posterocrista starting from it lacks ametacone, bends slightly to the lingual sideand ends to the postero-lingual cingulum. Onthe other hand, the anterocrista is extremelyweak and ends to the parastyle. The linguallobe is very pronounced, with the anteriorside concave and the posterior one straight.The protocone is rather high and well separatedfrom the paracone. The postero-lingual basinis shallow and narrow, bordered by a strongcingulum, which lacks a hypocone. There is avery small postero-labial basin bordered by aweak cingulum, while in the anterior part ofthe tooth the cingula are almost absent.M 1 (Plate 1, 2): The metacone and paracone

are very high, separated by a deep valley. T h eparastyle is well developed, separated fromthe paracone and connected with the protoco-nule through a weak cungulum. The protoco-nule is connected with the protocone, withwhich it is of equal size, and the paraconethrough a low ridge. The posterocrista of theprotocone bears an elongated accessory cusp,which is separated from the metaconule.There is a weak cingulum running along theposterior edge of the molar, starting from themetaconule and ending next to the posteriorarm of the metacone. The metastyle is absent.The labial side of the tooth bears a weak cin-gulum, while the lingual side lacks one.M 2 (Plate 1, 3): The posterior arm of themetacone ends slightly more labially than theanterior arm of the paracone. The parastyle isa small widening on the antero-labial cingul-um. The cusps are as on the M1, except forthe protocone and the protoconule that areseparated. The cingulum that runs along theanterior part of the tooth, from the protoconu-le to the parastyle, is strong, while the one inthe posterior edge of the tooth is reduced onlyin the postero-labial corner. The cingula alongthe labial and the lingual side are absent. T h elingual side of the tooth is slightly assymetri-cal, due to the anterior position of the proto-c o n e .M 3: The parastyle is absent; there is a weakcingulum anteriorly to the parastyle. The pro-toconule is small, connected with the paraco-ne by an anterior and a posterior ridge andseparated from the protocone. There is asmall, elongated accessory cusp on the poste-rocrista of the protocone and a deep valleyseparates it from the metaconule. The metaco-nule and the metacone are separated. In onespecimen there is a valley lingually to theaccessory cusp and a small cuspule in it. T h elingual lobe is rather prominent in one speci-men. p 2: It has an elliptical shape. A very weakparaconid is present. There is a well develo-ped cingulum in the postero-lingual side thatforms a small bulge in the posterior corner, inthe base of the posterocristid.

53

p 3: There is a small paraconid in two speci-mens, while in the third it is absent. The ante-rocristid is well pronounced, while the poste-rocristid is less pronounced. There is a smallbulge at the base of the posterocristid. T h ewell pronounced posterior cingulum curvesupwards and forms narrow basins postero-lin-gually and postero-labially , with the postero-lingual one wider.p 4 (Plate 1, 4a & b): The lingual margin isstraight, while the labial one is strongly con-vex. The anterocristid ends near the antero-lin-gual corner, where a paraconid is formed. Anarrow talonid basin is present. The postero-cristid is very weakly developed and at its enda small hypoconid is formed. The entoconid isabsent. There is a narrow cingulum runningalong the antero-labial side of the tooth.m 1 (Plate 1, 5a & b): The morphology of thetooth is very simple. The oblique cristid endson the metaconid. The entostylid is small andindependent. There is a weak and narrow cin-gulum starting from a very small cuspulid inthe re-entrant valley, running labially to thetrigonid and along the anterior part of thetooth and ending before the parastyle, whichis very small. The cingula along the lingualand posterior part of the tooth and labially tothe talonid are absent. m 2: The morphology is very similar to that ofthe m1. The antero-labial cingulum extentsfar to the antero-lingual corner of the tooth,where the parastylid is formed.m 3: (Plate 1, 6): The cusps are as on them1,2; so are the cingula, only weaker. T h eoblique cristid ends on the protoconid-meta-conid crest. The entostylid and the cuspulidin the re-entrant valley are extremely small,while the parastylid is bigger.

D i s c u s s i o n : Rümke (1985) gave some char-acteristics of the subfamily Desmaninae. T h ep4 often bears a metaconid or a metaconalrib; the oblique cristid of the lower molarsends either against the tip of the metaconid oragainst the protoconid-metaconid crest; them3 is slightly reduced; the upper molars havea divided mesostyle and a strongly developed

lingual part featuring a protoconule, a proto-cone, a metaconule and often a small tubercleon the posterocrista of the protocone. A l lthese features are present in the Marathoussasample. The genera included in the subfamilyare: D e s m a n a, G a l e m y s, D i b o l i a(= R u e m k e l i a) and M y g a l i n i a (Rümke 1985).The genera are distinguished mainly on thebasis of morphological features of the incis-ors and the premolars and two ratios, p2/p3and P2/P3. The scantiness of the material andthe absence of these teeth from theMarathoussa sample makes the identificationof the genus to which the sample belongsi m p o s s i b l e .

Family: Soricidae FI S C H E R V O N WA L D H E I M,1 8 1 7Subfamily: Crocidurinae MI L N E- ED WA R D S,1 8 7 2Genus: C ro c i d u r a WA G L E R, 1832Crocidura kornfeldi KO R M O S, 1934

M a t e r i a l :I sup., 7 dex.: 33, 383, 388, 395, 397, 401,404. 6 sin.: 382, 387, 392, 405, 429, 576.P4, 4 dex.: 541, 542, 544, 546. 2 sin.: 153,5 4 0 .A1 sup., 5 dex.: 556, 558, 560, 563, 564. 3sin.: 557, 561, 562.A2,3 sup., 1 dex.: 546. 4 sin.: 574, 580, 581,5 8 2 .M1, 4 dex.: 490, 495, 508, 544. 1 sin.: 507.M2, 7 dex.: 45, 498, 499, 506, 511, 518, 519.4 sin.: 48, 496, 497, 514.M3, 1 sin.: 536.I inf., 1 dex.: 426. 4 sin.: 412, 413, 415, 423.A1 inf., 2 dex.: 559, 569. 2 sin.: 566, 575.p4, 6 sin.: 472, 475, 570, 571, 572, 573.m1, 5 dex.: 448, 453, 468, 470, 473. 7 sin.:109, 432, 454, 455, 471, 472, 475.m2, 6 dex.: 54, 452, 468, 470, 473, 476. 4sin.: 457, 461, 462, 474.m3, 2 dex.: 464, 476. 3 sin.: 465, 474, 555.

M e a s u re m e n t s: see Tables 3 and 4.


54

DEINSEA 8, 2001

D e s c r i p t i o n :I sup. (Plate 2, 13a & b): There are sevencomplete specimens which can be described.The apex is single (the incisor is not fissi-dent). The cingulum along the postero-buccalm a rgin is narrow but well pronounced, whilethe antero-buccal part of the cingulum isstrong and undulate.A s u p . (Plate 2, 14a & b): The upper antemo-lars are typical for Crocidura, triangular, witha crest directed from the main cusp to the dis-tal part of the tooth. The cingula are wide andwell pronounced. The A1 is larger than theother two antemolars.P 4 (Plate 2, 15): The parastyle is separatedfrom the paracone by a deep valley. The pro-tocone is small and situated buccally in theantero-lingual corner. The hypocone is not adistinct cusp, but it forms a cingulum-likeridge that runs along the lingual margin ofthe tooth. The hypoconal ridge is separatedfrom the protocone by a valley. The posteriore m a rgination is strong.M1 and M2 (Plate 2, 16 & 17): They areboth relatively broad and short and with slen-der cusps. The protocone is connected to theparacone, while between the protocone andthe metacone there is a wide and deep valley.The hypocone is situated postero-lingually tothe protocone. The valley that separates thehypocone from the protocone is rather shal-l o w. In the M1 the posterior emargination isstrong. In one of the M1s , there is a smallcuspule situated lingually to the protocone.M 3 (Plate 2, 18): It is relatively short. T h eprotocone and the hypocone form a ridge thatsurrounds the lingual basin. The protocone isthe best developed cusp.I inf. (Plate 2, 19a & b): The apex is curvedupwards. The upper margin of the two teeththat are complete is slightly bicuspulate in thefirst and tricuspulate in the second. The buc-cal cingulum is narrow but well pronounced.Along the lingual face of the incisor there is agroove that curves downwards below thenotch in the basal border.A i n f .: The lower antemolars are typical forC ro c i d u r a, with two low crests, beginning

from the main cusp and directed the one postero-lingually and the other postero-buc-c a l l y. The cingula are wide and strong, butthe buccal one stops before the anterior endof the tooth.p 4 (Plate 2, 20): The lower premolar has onehigh and pointed cusp. Both the lingual andthe buccal cingula are strong and wide.m1 and m2 (Plate 2, 21): The buccal re-ent-rant valley opens in a large distance abovethe base of the molars. The entoconid crest israther low. The buccal cingulum is very nar-r o w, the lingual one is wider and they areboth undulate and not well pronounced.m 3 (Plate 2, 22a & b): The talonid consistsonly of the hypoconid and, thus, does notform a basin. The buccal cingulum is narrowand well pronounced, but the lingual one isvery weak and it disappears completely underthe talonid.M a n d i b l e: The mandibles, which werefound, consist only of the horizontal ramusthat is high but narrow. The mental foramenis placed underneath the posterior part of thep 4 .

D i s c u s s i o n : The presence of one triangularcusp on the p4, the position of the mentalforamen underneath the posterior-most part ofthe p4 and the unpigmented teeth lead to theassumption that the specimens fromMarathoussa belong to the subfamilyCrocidurinae. Several genera are included inthis subfamily. According to Repenning(1967), the genus C ro c i d u r a is characterisedby a reduced talonid with hypoconid only onthe m3, short lower incisor with upturnedapex, occlusal surface without cuspules and agroove along the lingual length that curvesdownward below the notch in the basal bor-d e r. All these characters have been recogni-zed in the Marathoussa teeth, which must beidentified as C ro c i d u r a.

A number of species are included in thegenus C ro c i d u r a. The Marathoussa samplewill be compared to those found inSoutheastern Europe. The recent species C.

55


ru s s u l a (HE R R M A N N, 1780) is characterisedby the constriction in the buccal cingulum ofthe m2 (Reumer 1996). The absence of thisfeature from the m2s of Marathoussa teethd i fferentiates the latter from C. ru s s u l a. Aspecies of C ro c i d u r a described only from thePleistocene of Crete (Xeros) is C. zimmer -m a n n i (WE T T S T E I N, 1953). The P4 fromXeros has broader lingual basin with a more-developed protocone situated more linguallyand larger distance between the parastyle andthe paracone than the P4 from Marathoussa.The lower incisor from Xeros is much larg e r,with more pronounced buccal cingulum thanthat from Marathoussa. The lower molarsfrom Xeros have a weaker lingual cingulumthan those from Marathoussa (pers. observa-tion). Therefore, the Marathoussa sampledoes not belong to C. z i m m e r m a n n i.

The species C. k o r n f e l d i was described byKormos (1934) and its diagnostic featuresare: the P4, M1 and M2 that are relativelyshort and broad, the upper molars that have ashort degree of posterior emargination andthe lower incisor that is acuspulate. Somemore dental characters of the species are thebuccal position of the protocone in the P4and the presence of a narrow and undulatecingulum in the lower molars. The presenceof these characters in the teeth ofMarathoussa allow us to determine the sam-ple to this species. The C. k o r n f e l d i f r o mMarathoussa is morphologically identical tothat from Ravin Voulgarakis (Koliadimou1996) and that from Osztramos 3/2 (Reumer1984). Moreover, the dimensions of the teethfrom the three localities are very similar (Fig.2).

Figure 2 Diagram with the mean values of the dimensions for the M1 (PE, L L , B L , AW, PW) and the m1 (TRW,TAW, L) of three

samples of C. k o rn fe l d i from Osztramos 3/2 (Reumer, 1 9 8 4 ) , R avin Vo u l g a rakis (Koliadimou, 1996) and Mara t h o u s s a .

Subfamily: Soricinae FI S C H E R V O N WA L D H E I M,1817 Tribe: Soricini FI S C H E R V O N WA L D H E I M, 1817Genus: S o re x LI N N A E U S, 1758Sorex minutus (LI N N A E U S, 1766)

M a t e r i a l :I sup., 3 dex.: 396, 406, 407. 7 sin.: 32, 124,125, 389, 398, 399, 402.P4, 2 sin.: 89, 11 3 .M1, 1 dex.: 101. 5 sin.: 46, 505, 509, 515,5 1 6 .M2, 1 dex.: 99.I inf., 3 dex.: 128, 419, 425. 3 sin.: 410, 414,4 1 7 .m1, 1 dex.: 444. 3 sin.: 53, 110, 459.m2, 4 dex.: 431, 442, 443, 460. 5 sin.: 100,107, 439, 441, 449.m3, 1 dex.: 106.Mandible, 1 sin.: 480.


D e s c r i p t i o n :I sup. (Plate 2, 1a & b): All specimens arefissident. The talon is very large and pointed.There is a buccal cingulum in all of the com-plete incisors.P 4 (Plate 2, 2): The hypocone corresponds toa small elevation of the lingual cingulum, w h i l ea wide valley separates it from the protocone.The protocone is situated buccally and bearsa short crest directed posteriorly that forms asmall cuspule postero-lingually to the para-style. The parastyle is connected to the para-cone by a rather high parastylar crest.M1 and M2 (Plate 2, 3): The metaloph isdivided in two branches: the first is connec-ted to the metacone and the second is direc-ted posteriorly, but stops buccally to thehypocone. The latter is small and situated onthe lingual cingulum, which starts from thepostero-lingual corner of the molar and stopsright under the protocone. The hypocone isseparated from the second branch of themetaloph by a deep valley.I inf. (Plate 2, 4a & b): Two of the six availa-ble specimens are complete. They are both

tricuspulate, but a very small prominence isformed posteriorly to the last cuspule, so thatit gives the impression that they are tetracu-spulate. The central valley is the deepest ofthe three existing. There is a rather wide butnot well-pronounced cingulum on the buccalposterior margin of four specimens. There isalso a weak cingulum along the lingual poste-rior margin in all the incisors.m1 and m2 (Plate 2, 5a & b, 6a & b): T h eentoconid is situated posteriorly and the ento-conid crest is very high. The oblique crestbears a mesoconid on the unworn specimens.The buccal re-entrant valley opens in smalldistance above the buccal cingulum. The lat-ter is narrow but well pronounced, while thelingual cingulum is wide but not pronounced.m 3: The talonid consists of a well formedbasin with hypoconid, entoconid and ento-conid crest. The cingula are like in the otherlower molars.M a n d i b l e: One fragment of the horizontalramus was found. It is short and narrow andthe distance between the alveoli shows rathersmall molars. The area of the mental foramenis absent.

D i s c u s s i o n : The high entoconid crest on them1,2 and the basined talonid of the m3 char-acterise the genus S o re x (Repenning 1967).Their presence in the studied material allowits classification to this genus. Several spe-cies are included in the genus S o re x. The pre-sence of the fissident upper incisors, howe-v e r, characterises only two Eurasian species:S. minutus and S. mirabilis HO F F M A N N, 1971.Also, in S. araneus a fissident upper incisormay occasionally be observed (Reumer1985), but not always as in the assemblagefrom Marathoussa. The teeth fromMarathoussa differ from S. mirabilis m a i n l ybecause of their smaller size. On the otherhand, they have the same morphology as, anda size very close to, those of S. minutus f r o mTegelen described by Reumer (1984). T h e i rmorphology is also identical to the morpholo-gy of the teeth of S o re x cf. m i n u t u s f r o mRavin Voulgarakis (Koliadimou 1996), but

56

DEINSEA 8, 2001

57


the size of the latter is larger (Fig.3). A l lthese allow us to determine the Marathoussasample as S o rex minutus.

Genus: S o re x LI N N A E U S, 1758Subgenus: D re p a n o s o re x KR E T Z O I, 1941S o r e x (Drepanosorex) praearaneus(KO R M O S, 1934)

M a t e r i a l :I sup., 7 dex.: 386, 393, 394, 400, 403, 408,428. 5 sin.: 122, 123, 384, 385, 390.A sup., 12 dex.: 579, 583, 585, 586, 587, 588,590, 600, 601, 602, 603, 604. 2 sin.: 584,5 8 9 .P4, 3 dex.: 50, 118, 543. 3 sin.: 115, 11 6 ,5 3 9 .M1, 2 dex.: 501, 530. 1 sin.: 520.M2, 2 dex.: 43, 97. 6 sin.: 44, 47, 49, 102,451, 502.M3, 1 dex.: 533. 3 sin.: 535, 537, 538.I inf., 2 dex.: 34, 421. 4 sin.: 129, 154, 409,

4 2 4 .A1, 1 sin.: 565.p4, 1 dex.: 549. 5 sin.: 135, 136, 137, 138,5 4 7 .m1, 1 dex.: 52. 5 sin.: 51, 438, 445, 447, 469.m2, 3 dex.: 52, 434, 435. 6 sin.: 51, 436, 450,467, 477, 527.m3, 2 dex.: 105, 466. 4 sin.: 51, 114, 463,4 6 9 .Mandibles, 3 dex.: 52, 479, 481. 7 sin.: 51,467, 469, 477, 482, 483, 485.


D e s c r i p t i o n :I sup. (Plate 3, 1a & b): It is fissident. T h etalon is well developed and sharply pointed.The five complete specimens show a verybroad, undulated cingulum along the buccalposterior marg i n .A s u p . (Plate 3, 2): The upper antemolar isbulbous. The buccal and the lingual cingulum

Figure 3 Diagram with the mean values of the dimensions for the Isup (L, LT, h ) , the Iinf (L) and the m1 (TRW,TAW, L) of

S . m i n u t u s from Tegelen (Reumer 1984) and Marathoussa and S . c f. m i n u t u s from Ravin Vo u l g a rakis (Koliadimou 1996).

are strong. In nine of the studied teeth thereis a ridge connecting the lingual cingulumwith the main cusp, on the antero-lingual partof the tooth. The postero-lingual basin isw i d e .P 4 (Plate 3, 3): The two complete specimenshave a very weak hypocone, separated fromthe protocone by a wide valley that is lingual-ly surrounded by a small cingulum. A c i n g u l-um starts from the protocone, directedtowards the parastyle. In five of the speci-mens, it continues under the latter. On thiscingulum, lingually to the parastyle, there is asmall cuspule on all the specimens.M1 and M2 (Plate 3, 4 & 5): None of theavailable teeth can be certainly identified asM1. However, there are three fragments,which are described as M1, because they areslightly larger than the complete molars,which are identified as M2. All teeth (com-plete and broken ones) have a weakly develo-ped hypocone, separated from the protoconeby a valley. In the lingual side of this valleythere is a cingulum. The metaloph is not welldeveloped and between the latter and themetacone there is a wide and deep valley.M 3 (Plate 3, 6): The hypocone and the meta-cone are united forming a ridge, on the poste-rior part of the tooth, which is separated fromthe protocone by a valley. On the lingualm a rgin of the valley there is a small cingul-um which seems to be an extension of theridge and stops lingually to the protocone.I inf. (Plate 3, 7a & b): Three of the studiedteeth are complete. They are all tricuspulatewith well developed, bulbous cuspules. T h e r eare three valleys separating the cuspules fromeach other and the middle one is the deepest.There is no sign of cingulum on the buccals i d e .A1 inf. (Plate 3, 8a & b): The a1 found isonly slightly smaller than the p4. It is alsobulbous and the cingula on both the lingualand the buccal sides are wide but not pro-nounced. A postero-lingual basin is almostd i s t i n g u i s h a b l e .p 4 (Plate 3, 9a & b): The tooth is bulbous

with lingual and buccal cingula. The postero-lingual basin is wide and, in three out of fivespecimens, bordered postero-lingually by ashort ridge.m1 and m2 (Plate 3, 10a & b): The entocon-id is situated posteriorly, while the entoconidcrest is rather high. On some unworn speci-mens the oblique crest forms a small meso-conid. The buccal re-entrant valley opens in asmall distance above the cingulum. The lin-gual cingulum is wide but not well projected,while the buccal one is narrower, well deve-loped and undulate.m 3 (Plate 3, 10a & b): The talonid forms awide basin with a well-developed hypoconidand entoconid and a high entoconid crest.The cingula are as on the m1 and the m2.M a n d i b l e (Plate 3, 11a & b): There is onecomplete hemi-mandible found, in the assem-blage from Marathoussa. The apex of thecoronoid process is rounded. There is a coro-noid spicule situated very close to the apex.The external temporal fossa is very shallow,with not very distinct borders and reaches theupper condylar facet. The upper and thelower facets of the condyle are almost paral-lel and the interarticular area is relativelybroad all along the condyle. The internal temporal fossa is triangular and extends farupwards to the apex of the coronoid process.The mental foramen is preserved in four frag-ments and is situated under the p4/m1 transi-tion, on three and under the middle of them1, on the fourth of them.

D i s c u s s i o n :D re p a n o s o re x is considered as a subgenus ofthe genus S o re x and is characterised by gene-rally robust teeth, slightly to strongly exoeda-enodont anterior dental elements, a compara-tively large condyle and fissident upper incis-ors (Reumer 1985). The presence of all thesefeatures in the Marathoussa sample allows itsdetermination to D re p a n o s o re x. The subgenusincludes the following species: D. p r a e a r a -n e u s KO R M O S, 1934, D. marg a r i t o d o nKO R M O S, 1930, D. savini HI N TO N, 1911 and

58

DEINSEA 8, 2001

59


D. austriacus KO R M O S, 1937, while D. pachy -odon PA S A, 1947 and D. tasnadii KR E T Z O I,1941 are considered to be synonymous to D .p r a e a r a n e u s and D. savini, respectively.According to Reumer (1984), the species ofD re p a n o s o re x form a morphological sequen-ce with three trends: increase of exoedaeno-d o n t y, size increase and enlargement of thecondyle. The presence of a slight exoedae-nodonty in the antemolars and the small(comparatively to the other species) size ofthe teeth from Marathoussa, indicate that theybelong to D. praearaneus, the most primitivespecies of D re p a n o s o re x known. The meanvalues (Fig. 4) of the dental dimensions fromMarathoussa are slightly higher than thosefrom Tegelen (Reumer 1984) and lower thanthose from Ravin Voulgarakis (Koliadimou1996).

Tribe: Neomyini MAT S C H I E, 1909Genus: A s o r i c u l u s KR E T Z O I, 1959Asoriculus gibberodon (PE T E N Y I, 1864)

M a t e r i a l :I sup., 3 dex.: 119, 120, 391. M1, 6 dex.: 95, 96, 103, 437, 488, 513. 2sin.: 494, 521.M2, 1 dex.: 517. 2 sin.: 491, 512.M3, 2 dex.: 104, 534. 1 sin.: 532.I inf., 1 sin.: 422.m1, 3 dex.: 440, 446, 458. 1 sin.: 111 .m2, 1 dex.: 433. 1 sin.: 108.


D e s c r i p t i o n :I sup. (Plate 2, 7a & b) : Only one upperincisor is complete and it is fissident. T h etalon is rather short. There is a well-develo-ped and undulate cingulum on the postero-

Figure 4 Diagram with the mean values of the dimensions for the Isup (LT, L , H ) , the Iinf (L) and the m1 (TRW,TAW, L) of

D. p ra e a ra n e u s from Tegelen (Reumer 1984) and Marathoussa and D. c f. p ra e a ra n e u s from Ravin Vo u l g a rakis (Koliadimou 1996).

buccal margin of the tooth.M1 and M2 (Plate 2, 8 & 9): The hypoconeis well developed and separated from the pro-tocone by a valley. On the lingual part of thisvalley there is a cingulum. In one specimen,this cingulum forms a cuspule lingually to theprotocone. The hypocone is also separatedfrom the ridge that surrounds the hypoconalflange by a very shallow valley. The metal-oph is present and, in four specimens, it isdivided in two, small branches. The anteriorm a rgin of the molars bends strongly besidethe protocone. The parastyle is very strong.M 3 (Plate 2, 10): The protocone and theparacone are well developed, while the meta-cone forms a ridge on the posterior margin ofthe tooth. The hypocone is united to thisridge, but is very short. There is a cingulumrunning along the lingual margin of them o l a r.I inf. (Plate 2, 11a & b): One broken lowerincisor was found in Marathoussa. The dorsalm a rgin shows a deep valley and another one,s h a l l o w e r, situated anteriorly to the first. T h etooth, thus, is bicuspulate. The apex is uptur-ned. The posterior part of the tooth is not pre-s e r v e d .m1 and m2 (Plate 2, 12a & b): The entocon-id crest is very low, but present. The obliquecrest bears a mesoconid in the unworn speci-mens. The buccal re-entrant valley opens in avery small distance above the cingulum. T h ebuccal cingulum is narrow, undulate and wellpronounced, the lingual one is wide and lessp r o n o u n c e d .

D i s c u s s i o n :The short and bicuspulate lower incisors, thebroad but not pronounced lingual cingulumof the lower molars in the Marathoussa samp-le allow us to recognise the genus A s o r i c u l u s.The genus includes only European fossil spe-cies dated from latest Miocene/EarliestPliocene (Doukas et al. 1995) to the MiddlePleistocene (Koliadimou 1996). These spe-cies were described as E p i s o r i c u l u s, togetherwith the recent species from Asia. However,

Hutterer (1994) distinguished the Europeanfossil species as belonging to A s o r i c u l u s,while the Asiatic resent ones belong toE p i s o r i c u l u s. The species of A s o r i c u l u s, the-refore, are: A. gibbero d o n (PE T E N Y I, 1864), A .castellarini (PA S A, 1947), A. tornensisJA N O S S Y, 1973, A. borsodensis JA N O S S Y, 1973and A. thenii MA L E Z & RA B E D E R, 1984. T h especies A. tornensis a n d A. borsodensis a r enow considered as synonyms of A. gibbero -d o n (see Reumer 1984).The characteristic of A. thenii that diff e r e n t i a-tes this species from the assemblage fromMarathoussa is its longer I inf. with stronglyupturned apex. The species A. castellariniwas initially described by Pasa (1947) asNeomys castellarini, based on material fromCava Sud (Soave, Italy). Later on, A . cf. c a s -tellarini has been reported from Monte Peglia63 (Orvieto, Italy) (van der Meulen 1973),from Vcelare 3,4 a/7 (Slovakia) (Fejfar &Horacek 1983) and from Ravin Vo u l g a r a k i s(Greece) (Koliadimou 1996). The type mate-rial is scanty and not well preserved. T h esamples from Monte Peglia 63 and RavinVoulgarakis are also scanty and fragmentary.The species shares a lot of features in com-mon with A. gibbero d o n, including the size ofthe dental elements. Some differences seemto be in the area of the mandibular articula-tion. In A . cf. c a s t e l l a r i n i from RavinVoulgarakis the coronoid process inclinessomewhat anteriorly, the interarticular area isnarrow and the internal temporal fossaextends upwards. In A . cf. c a s t e l l a r i n i f r o mMonte Peglia 63 the coronoid process incli-nes backwards, the interarticular area is lin-gually deeply excavated and the internal tem-poral fossa extends upwards without reachingthe tip of the coronoid process. Finally, in A .cf. c a s t e l l a r i n i from Slovakia, the coronoidprocess is anteriorly inclined. The variabilityin the morphology of the mandibular articula-tion within the species A. castellarini is notwell defined. On the other hand, in A . g i b b e -ro d o n the interarticular area is also narrow,the coronoid process has concave marg i n s

60

DEINSEA 8, 2001

61


and the internal temporal fossa continuesupwards to just below the tip of the coronoidprocess. The distinction of the two speciesbased on the morphology of the mandibulararticulation is not clear and needs furthers t u d y, which might lead to their synonymy(Reumer 1984).

The species A. gibbero d o n is reported fromseveral localities in Europe (Table 11). T h e r eis a large variability in the morphology of theupper molars, which have been separated intwo morphotypes within the species (Reumer1984). Morphotype A is characterised by anisolated hypocone, separated by a small val-ley from the ridge that surrounds the hypo-conal flange. Morphotype B has a hypoconethat is considerably less developed and con-nected to the ridge surrounding the flange.The PE-index, calculated by means of theformula [(LL+BL)/2*PE]-1 (Reumer 1984),is a number between 0 and 1: higher valuesof the index indicate larger degree of posteri-or emargination. For Morphotype A, the PE-index is moderate, while for Morphotype B,the index is moderate to strong. Despite thevariability in the morphology of the uppermolars, the lower molars do not show anyv a r i a b i l i t y. However, size differences are alsoobserved among the dental remains from thevarious localities. There is not any relationbetween the size or the morphotypes and time(in Osztramos 7 both morphotypes werefound) and these differences cannot be usedfor biostratigraphical correlation.

The fissident upper incisors, the bicuspulatelower incisors, as well as the low entoconidcrest and the buccal re-entrant valley openingslightly above the cingulum in m1,2s indicatethat the Marathoussa sample can be attributedto A. gibbero d o n. The morphology of theupper molars suggests similarities toMorphotype A. The mean value of the PE-index for the M1s is 0.18 and for the M2s0.14 (Tab. 11), which also confirm the simila-r i t y. The Marathoussa teeth are larger than

those described from some European locali-ties (Tab. 12), but they have the same size asA. gibbero d o n from Tourkovounia 1 (Reumer& Doukas 1985) (based on comparison of thetwo assemblages).

Tribe: Beremendiini RE U M E R, 1984Genus: B e re m e n d i a KO R M O S, 1934Beremendia fissidens (PE T E N Y I, 1864)

M a t e r i a l :P4, 1 dex.: 545. 1 sin.: 139.M1, 1 dex.: 545. 1 sin.: 489.M2, 1 dex.: 492.M3, 1 sin.: 531.m1, 1 sin.: 478.Mandible, 1 sin.: 478.

M e a s u re m e n t s : see Table 13.

D e s c r i p t i o n :P 4 (Plate 1, 7): It is almost triangular, with alow protocone, situated close to the parastyle,while the hypocone is placed anteriorly.Between the protocone and the hypoconethere is a very shallow valley. The parastyleis large, with a high parastylar crest.M1 and M2 (Plate 1, 7 & 8): The protoconeis situated on a U-shaped crest, which con-nects it to both the paracone and the metaco-ne. The hypocone is low and separated fromthe protocone by a deep valley.M 3 (Plate 1, 9): The M3 is rather reduced, inrelation to the other upper molars. The proto-cone is low and situated lingually. The lingu-al margin is circular. The hypocone is a cuspseparated from the protocone by a shallowv a l l e y. The crest between the paracone andthe metacone is vertical to the crest that con-nects the paracone and the parastyle.m 1 (Plate 1, 10a & b): The trigonid basin isdeep and wide due to a slight posterior- w a r dshift of the metaconid. The entoconid crest isshort. The lingual cingulum is slightly deve-loped, but the buccal one is well developedand a little undulate.M a n d i b l e (Plate 1, 10b): On the fragment of

one mandible with the m1 the mental fora-men is situated underneath the buccal re-ent-rant valley of the first molar.

D i s c u s s i o n : The tribe Beremendiini includesonly the genus Beremendia. Kormos (1934)did not give a formal diagnosis for it, but hegave descriptions for the diagnostic dental ele-ments of the genus (I sup., upper antemolars, Iinf. and m3). Unfortunately, in the Mara-thoussa samples these teeth were not found.H o w e v e r, the slight posterior-ward shift ofthe metaconid in the m1s and m2s, in relationto other genera, is characteristic for the genus(Reumer 1984). The morphology of the m1found in Marathoussa shows this character.Furthermore, the size of the studied teethfalls within the limits of variation for theteeth of the genus B e re m e n d i a. Therefore, thespecimens from Marathoussa belong to thisgenus.

Four species of the genus B e re m e n d i a h a v e

been described: B. ucrainica (PI D O P L I C H K O,1956), B. minor RZ E B I K, 1976, B. sinensis(ZD A N S K I, 1928) and B. fissidens (PE T E N Y I,1864). The species B. ucrainica is consideredsynonymous to B. fissidens ( R z e b i k -Kowalska 1976). The species B. m i n o r i ssimilar to the specimens from Marathoussa,but the size of its elements is much smaller.

The species B. sinensis, originally describedas Neomys sinensis, was later transferred toB e re m e n d i a (Kretzoi 1956). The elements areonly slightly smaller than those of B. fis -s i d e n s and the main difference between thetwo species is in the area of the mandibularcondyle. Unfortunately, there is no homolo-gous feature in the Marathoussa mandible forcomparison. The dental morphology of thestudied teeth is identical to that of B. fis -s i d e n s from Ravin Voulgarakis, Tegelen andMonte Peglia 63 (Koliadimou 1996; Reumer1984; van der Meulen 1973). However, thedimensions of the Marathoussa M1 and m1

62

DEINSEA 8, 2001

Figure 5 Diagram with the mean values of the dimensions for the M1 (PE, L L , B L , AW, PW) and the m1 (TRW,TAW, L) of three

samples of B . f i s s i d e n s from Tegelen (Reumer 1984), R avin Vo u l g a rakis (Koliadimou 1996) and Mara t h o u s s a .

63


are smaller than those of Tegelen and larg e rthan those of Ravin Voulgarakis (Fig. 5).Taking into account all the above mentioned,the Marathoussa sample of B e re m e n d i a m u s tbe attributed to B. fissidens.

Order: Rodentia BO W D I C H, 1821

Family: Sciuridae GR AY, 1821Subfamily: Marmotinae PO C O C K, 1923Tribe: Citellini GR O M O V, 1965Genus: S p e r m o p h i l u s CU V I E R, 1825S p e r m o p h i l u s s p.

M a t e r i a l :P4, 1 dex.: 156.M1-2, 1 dex.: 378. 1 sin.: 381.m3, 1 dex.: 379.


D e s c r i p t i o n :P 4 : The only tooth found is broken and worn.Only the protocone can be distinguished,which is very prominent.M 1 - 2 (Plate 4, 1): The shape of the tooth ist r i a n g u l a r. The protocone is the most promi-nent cusp. It is connected with the paraconeand the metacone by two low ridges. So, themain cusps form a V on the occlusal surface.There are no metaconule or protoconule. T h emesostyle is very well-developed. There is acingulum running along the anterior marg i nof the tooth. The posterior cingulum startsfrom the mesostyle, ends next to the metaco-ne and is less pronounced than the anteriorcingulum. The roots are not preserved.m 3 (Plate 4, 2): The anterior width is larg e rthan the posterior one. The protoconid is verystrong, while the metaconid and the hypocon-id are rather well-developed and connectedwith the protoconid through the anterior lop-hulid and the ectolophid, respectively. T h e r eare no endoconid or mesostylid. The occlusalsurface is closed and a wide basin. is formedin the middle. The roots are not preserved.

D i s c u s s i o n : The Sciuridae from Marathoussaare scanty. However, the triangular shape ofthe M1-2, the same size of the paracone andthe metacone and the low posterior cingulumon the M1-2, the large width relatively to thelength and the shape of the mesoconid on them3 indicate similarities to the genusS p e r m o p h i l u s. The genus S p e r m o p h i l u s i n c l u-des several recent and fossil European spe-cies with small differences. The main diff e-rences are referred to the number of the rootsand the morphology of the premolars.In the material from Marathoussa the rootsare not preserved and no premolars werefound. The molars from Marathoussa resem-ble to those from Ravin Vo u l g a r a k i s(Koliadimou 1996), which belong to the spe-cies S. nogaici. However, the M1-2 fromMarathoussa are larger than those from RavinVoulgarakis and the m3 from Marathoussa issmaller than that from Ravin Vo u l g a r a k i s .M o r e o v e r, the absence of metaconule andprotoconule and the presence of the mesosty-le distinguish them from the RV L s a m p l e .The scanty material from Marathoussa cannotallow certain comparisons for a specificdetermination and, thus, it is referred asS p e r m o p h i l u s s p .

Family: Zapodidae CO U E S, 1875Subfamily: Sicistinae AL L E N, 1901Genus: S i c i s t a GR AY, 1827S i c i s t a c f . s u b t i l i s

M a t e r i a l: m1 dex.: 597.

M e a s u re m e n t s: L=1.147; W = 0 . 8 9 6

D e s c r i p t i o n :m 1 (Plate 4, 3): The only molar that wasfound is very worn and the morphologicalfeatures are not well distinguished. There is arather large anteroconid that is united to theprotoconid and the metaconid. There is not amesostylid or ectostylid, while the mesoconidis very low. The posterior cingulum is lowand ends on the endoconid.

D i s c u s s i o n : The molar found in Marathoussabelongs to the genus S i c i s t a, which is charac-terised by brachydont molars with simplem o r p h o l o g y. According to Kowalski (1979),S i c i s t a is separated in two evolutionary bran-ches. The first is represented by the species S .b e t u l i n a, with a rather complicated morpholo-gy on the molars, and the second by the spe-cies S. subtilis, with very simple morphology.The molar from Marathoussa, with a simplem o r p h o l o g y, belongs to the second group,which first appeared in Pliocene of Asia, withthe species S. pliocaenica. In Greece and theBalkan region only representatives of thesecond branch have been reported. The molarfrom Marathoussa is quite similar, in morp-hology and size, to the molar from A r n i s s a(Mayhew 1977) that was classified as S. sub -t i l i s. The same species was also reportedfrom Ravin Voulgarakis (Koliadimou 1996),with upper molars so that comparisonbetween the two samples is impossible. T h esole specimen of Marathousa and its simplemorphology allow us to determine this as S . cf. s u b t i l i s.

Family: Muridae GR AY, 1821Genus: A p o d e m u s KA U P, 1829Apodemus sylvaticus/flavicollis

M a t e r i a l :M1, 7 dex.: 28, 142, 191, 193, 194, 195, 223.7 sin.: 143, 144, 192, 196, 197, 198, 222.M2, 10 dex.: 147, 148, 149, 199, 200, 223,354, 356, 357, 368. 3 sin.: 201, 202, 224.M3, 5 dex.: 152, 349, 355, 370, 374. 2 sin.:151, 224.m1, 8 dex.: 25, 146, 204, 205, 206, 208, 358,369. 10 sin.: 145, 203, 207, 209, 210, 211 ,259, 360, 361, 362. m2, 5 dex.: 215, 219, 220, 363, 364. 6 sin.:212, 213, 214, 216, 217, 218.m3, 4 dex.: 155, 221, 350, 367. 8 sin.: 351,352, 353, 365, 366, 371, 372, 373.


D e s c r i p t i o n :M 1 (Plate 4, 4 & 5): The tooth is oval. The t1is situated posteriorly of the t3 in 27% of theteeth and it is always connected with the t2.In one specimen there is a t1 bis and in threespecimens there is a t2 bis. In eight speci-mens the t1 and the t5 are connected througha spur, in three the t1 bears a smaller spurtowards the t5 and in one specimen the t1does not bear a spur. In six specimens the t3is connected with the t5 and in five speci-mens the t3 bears a small spur . In nine speci-mens the t7 and t4 are separated, while in theremaining three (25%) the two cusps are con-nected with each other. The t7 is relativelyl a rge and oval. The t4 and the t6 are almostsymmetrical. The t12 is very small in all thespecimens. It is connected with the t8, but notwith the t9 in eight specimens and it is con-nected with the t8 and t9 in two specimens.The t9 is smaller than the t6 and it is connec-ted with the t8 in eight out of eleven speci-mens. There are four roots.M 2 (Plate 4, 6 & 7): The length and thewidth are almost equal. The t1 is almosttriangular and bears a small spur in two outof nine specimens. The t3 is small, without aposteriorly directed spur. The t7 is large andit is connected with the t4 only in four out often specimens. The t9 is smaller that the t6,but still relatively large in 70% of the teeth.The t12 is very small, almost absent. The t9and the t8 are connected in all the teeth.There are four roots.M 3: The t1 is big and oval and isolated in thetwo unworn specimens. In the rest the t1-t5-t4 and t5-t6-t8 are connected. There are threer o o t s .m 1 (Plate 4, 8-10): The antero-lingual andthe antero-labial cuspids (a.lg.c. and a.lb.c.respectively) are symmetrical in 92.3% ofthe teeth. The antero-central cuspid (a.c.c.) issmall in nine and absent in two specimens,while it is large in the remaining three speci-mens. It is connected with the a.lg.c. and the

64

DEINSEA 8, 2001

65


a.lb.c. except for one specimen. There is al a rge posterior accessory cuspid (c1) connec-ted with the hypoconid; in front of it there areone (41.7%) or two (58.3%) smaller acces-sory cuspids on the labial cingulum. The ter-minal heel is low. It is rounded in eight andelongated in five specimens. There is a smallspur directed from the hypoconid-entoconidjunction towards the protoconid-metaconidjunction in twelve specimens, while the lon-gitudinal crest is well developed only in threes p e c i m e n s .m 2 : It is almost square. The a.lb.c. is larg eonly in one out of eleven specimens. It iselongated and connected with the protoconidby a low ridge in the anterior edge of thetooth. The c1 is very reduced in all teeth andabsent in two specimens. Only one specimenhas a second accessory cuspid on the labialcingulum. There is a short longitudinal spurin six specimens; the spur is even shorter inthe remaining five teeth and it is never con-nected with the protoconid-metaconid junc-tion. The terminal heel is low, elongated andalmost of equal size as, or a little smallerthan, the a.lb.c..m 3 : The tooth is triangular. Half of the speci-mens preserve a small a.lb.c., but in the otherhalf of them it is absent. In three out of ele-ven specimens there is a very small posteriorcuspid (c1).

D i s c u s s i o n :The presence of the t7 and the t12, the con-nection of the t6 with the t9, the anteriorposition of the t1 and its connection with thet2 in all studied M1s, the presence of thea.c.c., the tendency of the cuspid pairs toform chevrons and the rareness of longitudin-al spur in the m1s of the Marathoussa sampleare characteristic for the genus A p o d e m u s.The genus includes many species and theMarathoussa material will be compared tosome European and Western Asian species ofapproximately the same size only.

Apodemus dominans KR E T Z O I, 1959 is a well-

known European species that differs clearlyfrom the Marathoussa sample, by the follo-wing features: its M1 has large posterior cin-gulum and three roots, its M2 has a large t9and the t1 lacks a posteriorly directed spur,its m1 has a mostly elongated posterior cin-gulum and weak accessory cuspids on thelabial cingulum. Moreover, A. dominans t e e t hare slightly larger than those of the Marathous-sa sample (Table 16).

Apodemus caesare a n u s BAT E, 1942, knownfrom the Middle-East, has many similaritieswith the specimens from Marathoussa, inclu-ding their size (Tab. 16), but also some featu-res that clearly distinguish it from theMarathoussa A p o d e m u s. In the M1 of A. cae -s a re a n u s, the t1 and the t4 are of equal sizeand smaller than the t7, the t2 bis is absentand a fifth root is very common. In the M1from Marathoussa the t7 is smaller than thet1 and the t4, the t2 bis is present in three outof eleven molars and there are always fourroots. The buccal cingulum in the m2 of A .c a e s a re a n u s has three distinct, well-develo-ped cuspids, whereas the m2 fromMarathoussa have a very weak cingulum. Inthe m3 of A. caesare a n u s both a.lb.c and pos-terior cuspid exist, but in the m3 fromMarathoussa the a.lb.c. exists in six and theposterior cuspid in three out of elevenmolars. Therefore, the A p o d e m u s f r o mMarathoussa does not belong to this species.

Another species known from Middle-East isApodemus levantinus BAT E, 1942. The m1 ofthis species is longer (Tab. 16) and the M2has a much larger t1 that those fromMarathoussa. The m1 and m2 of this specieshave a very well-developed labial cingulumwith four cuspids. On the contrary, the m1and m2 from Marathoussa have a weak cin-gulum with two or three cuspids in the m1and one or none in the m2. Finally, in the m3of A. levantinus a posterior cuspid is commonin most cases, but in the m3 of Marathoussathis cuspid is very rare. The above morpholo-

gical and metrical differences distinguish theMarathoussa sample from A. levantinus.

Apodemus atavus HE L L E R, 1936 fromGundersheim-4 (Germany) has three roots,the t12 is slightly smaller than the t7 and thet4 is situated posteriorly and is always con-nected with the t7. The M2 has a t12, the t4and t7 are connected and the t3 is rather big.The t9 has the same size as the t12 and is notconnected with the t8. The m1 has a veryl a rge a.c.c. that is always connected with thea.lb.c. and a.lg.c.. The buccal cingulum iswell developed, with two or three cuspids.The protoconid and the metaconid are con-nected with the anterior cuspids. There is alongitudinal spur, but there is no connectionbetween the two posterior chevrons. The ter-minal heel is round and has the same size asthe a.c.c. The m2 has a well developed buccalcingulum with two cuspids. The a.lb.c. andthe terminal heel are of equal size, ratherl a rge. The m3 has a well pronounced a.lb.c..The molars of A. atavus are very close metri-cally to those of Marathoussa (Tab. 16).H o w e v e r, the very small t12 and the fourroots on the M1, the absence of the t12 fromthe M2, the very small a.c.c. on the m1, theweak buccal cingulum on the m2 and thevery small or absent a.lb.c. on the m3 of theMarathoussa sample differentiate the latterfrom A. atavus.

Apodemus etru s c u s EN G E S S E R, 1989 is a pri-mitive species of A p o d e m u s, with almostsimilar dental dimensions to the recent A. syl -v a t i c u s ( Tab. 16). In 40% of the M1s the t7 isabsent, while all M2s have a small t7. The t6and t9 are not connected in 30% of the M1sand in 33% of the M2s. The M1 and M2 havethree roots. All the m1s have a well-develo-ped a.c.c.. Most of the m2s have three acces-sory cuspids on the labial cingulum. In allthese features, the specimens fromMarathoussa differ from A. etru s c u s: M1sand M2s always have a well developed t7,the t6 is always connected with the t9, and

they have four roots, the a.c.c. in the m1s issmall or absent and most of the m2s have onecingulum cuspid only. These differences indi-cate that A. etru s c u s is much older than theMarathoussa sample.

Two of the extant species of A p o d e m u s t h a tare found in Western Anatolia are A. hermo -n e n s i s FI L I P P U C I et al., 1989 and A. uralensisPA L L A S, 1811. The first species has an M1 inwhich both the t1 and the t3 are connectedwith the t5 and, thus, the tooth is somewhatstephanodont. A t1 bis is usually present. Onthe M2 the t1 is large and, in most cases, hasan anterior and a posterior connection withthe t5. On the m1 the buccal cingulum is welldeveloped and the a.c.c. is usually isolatedfrom the a.lb.c. and the a.lg.c.. The M1 fromMarathoussa, however, show no tendency todevelop stephanodonty and a t1 bis is rarelypresent. On the M2 the t1 has not two con-nections with the t5, the buccal cingulum ofthe m1 is less developed and the a.c.c. is usu-ally connected with the a.lb.c. and the a.lg.c..Therefore, the Marathoussa A p o d e m u s d i ff e r sf r o m A. hermonensis, even though they sharemany common features. The second species,A. uralensis, is characterized by a ridge-liket7 on the M1 and M2, a t1 mostly separatedfrom the t2 on the M1, a well-developeda.c.c. and an asymmetrical posterior chevron(the entoconid is displaced anteriorly) on them1. On the M1 and M2 from Marathoussa,the t7 is a well-developed cusp and the t1 andt2 are always connected. On the m1, the a.c.c.is rather small and the posterior chevron isalmost symmetrical. Thus, the specimensfrom Marathoussa cannot be identified as A .u r a l e n s i s .

Many extant A p o d e m u s species differ in exte-rior appearance, but their dental elements ares i m i l a r. The specimens from Marathoussabelong to either A. sylvaticus or A. flavicollis,two species which are found both fossil andextant all over Europe. In A. flavicollis(ME L C H I O R, 1834) the t1 of the M1 is placed

66

DEINSEA 8, 2001

67


posteriorly and the t4 and t7 are rarely con-nected to each other; the t1 of the M2 hasvery often a posteriorly directed spur and thet12 is very reduced or absent; the m1 has avery reduced or small a.c.c. and the longitu-dinal spur is frequently well-developed, tosuch an extent that the two chevrons are con-nected with each other; the labial cingulumon the m2 is reduced (Pasquier 1974). In A .s y l v a t i c u s (LI N N A E U S, 1758) the t1 and t3 onthe M1 are symmetrical and placed at equaldistances from the t4-t6 line, the t4 and t7 areoften connected; on the M2, the t1 is triangu-lar and rarely bears a posteriorly directedspur and a t12 is mostly present; on the m1the a.c.c. is rather large and the longitudinalspur is weakly developed (Pasquier 1974).All the other characters of the dentition of thetwo species are the same and identical tothose from Marathoussa. There seems to bealso a small difference in the size of the twodentitions, as A. flavicollis has a little larg e rdental elements. However, a comparison offossil and extant samples from Eurasian loca-lities (Table 17), indicates that the size of thetwo species varies chronologically andg e o g r a p h i c a l l y, in such a way that no trendscan be detected. The size difference is smalland overlaps, even in the recent samples. T h eteeth from Marathoussa show a mixture ofthe characters of A. sylvaticus and A. flavicol -l i s and cannot be identified at the specieslevel with certainty; therefore, we refer tothem as A. sylvaticus/f l a v i c o l l i s.

Family: Cricetidae RO C H E B R U N E, 1883Subfamily: Cricetinae MU R R AY, 1886Genus: C r i c e t i n u s ZD A N S K Y, 1928Cricetinus koufosi KO L I A D I M O U, 1996

M a t e r i a l :M1, 1 dex.: 345. 2 sin.: 346, 347.M2, 1 dex.: 29. 2 sin.: 348, 380.M3, 1 dex.: 375. 1 sin.: 376.m1, 1 dex.: 150.m2, 2 dex.: 30, 377.


Description: M 1 (Plate 4, 11): The anterocone consists oftwo well-developed, separated cuspules.There is a not very pronounced anterior cin-gulum, which forms a small pre-anteroconein one specimen; the third specimen is veryworn and the features of this area are notc l e a r. There is an anterolophule that connectsthe anterior cuspules to the protocone. On thetwo unworn molars there is a very small spurlabially to the anterolophule directed towardsthe paracone, but not connected with it. T h eentolophe connects the protocone and paraco-ne to the hypocone. The metacone is connec-ted with the posterolophe through a veryshort metalophule. The posterolophe is well-developed and, thus, the hypocone and meta-cone are united on the posterior part of thetooth and the posterosinus is closed. The lin-gual cingulum is absent, while the labial oneis present but not well pronounced. In onespecimen, there is a small parastyle. T h elabial sinuses are wider, but the lingual onesare deeper. There are four roots.M 2 (Plate 4, 12): The morphology is verysimilar to that of the M1. The protocone isunited to the paracone through an anteriorand a posterior ligament and so does thehypocone to the metacone. The posterolopheis well developed and the posterosinus is clo-sed. The posterior metalophule is absent. T h ecingula are not very prominent. There arefour roots. M 3 : The dental pattern is similar to that ofthe M2, but the paracone is very weak, thetooth has a triangular shape. The labial cin-gulum is very strong. There are three roots.m 1 (Plate 4, 13): The only m1 found is veryworn. The anterior margin is smooth, withoutcingulum and convex. The anteroconid isseparated in two small cuspulids. The anterol-ophid is underdeveloped (doe not emengefrom the level of the anterosinusid) and con-nects the antero-labial cuspulid to the meta-conid. The protoconid is united to the ento-

conid through the ectolophid. The posterolop-hid is well-developed and reaches the poste-ro-lingual margin of the tooth. The posterosi-nus is closed. The labial sinuses are wide andthe labial and lingual cingula are well-develo-ped. There are two roots.m 2: The morphology is similar to that of them1, but the molar is wider. The labial ante-rolophid is well-developed, while the lingualone is absent. There are two roots.

D i s c u s s i o n : The genus C r i c e t i n u s has an m1with undivided or slightly divided anterocon-id, a smooth and convex oral surface, missingmesolophid and a very low, I-shaped anterol-ophid, while the upper first and secondmolars are characterised by missing or week-ly developed mesolophs and the M2 rarelybears a posterior metalophule (Hir 1997). T h epresence of these features in the Marathoussasample suggest a determination to the genusC r i c e t i n u s.

The species in this genus are: C. varians(ZH E N G, 1984), C. gritzai (TO PA C H E V S K I &SK O R I K, 1992), C. euro p a e u s (KR E T Z O I,1959), C. bere m e n d e n s i s HI R, 1994 and C .k o u f o s i KO L I A D I M O U, 1996. C. varians i sknown from the Middle Pleistocene of China.The m1 of this species and of theMarathoussa sample have approximately thesame size, while all the other molars of thisspecies are larger than their homologous fromMarathoussa (Figs. 5-9). The M1 of C .v a r i a n s lacks a preanterocone and the m2usually has a mesolophid (Zheng 1984).

The species C. gritzai was described fromOdessa, Ukraine (Topachevski & Skorik,1992). The large size (Figs. 5-9), the rare pre-sence of the preanterocone and the very smallmetalophule in M1, the well-developed ante-rolophid of m1, as well as the presence of amesolophid - always in m1 and usually in m2- distinguish this species from theMarathoussa C r i c e t i n u s.

The species C. euro p a e u s was described fromCsarnota 2 (Kretzoi 1959). The M1 alwayslacks an anterior cingulum or preanterocone,while the M1s found in Marathoussa haveeither an anterior cingulum or a preanteroco-ne. The M2 of C. euro p a e u s sometimes has amesoloph, which is absent in the MarathoussaM2. The posterior part of the M2 fromCsarnota 2 is less wide than the anterior part.On the M2 from Marathoussa the two partshave equal width. On the m1 of C. euro p a e u sthe anteroconid is not always divided in two,and the anterolophid is higher than the one onthe m1 from Marathoussa. Finally, the m2 ofC. euro p a e u s usually bears a mesolophid,which is absent from the Marathoussa one.The dental dimensions of C. euro p a e u s a r erather close to those of Marathoussa (Figs. 5-9). However, the morphological diff e r e n c e scannot allow the determination of theMarathoussa sample to C. euro p a e u s.

The dental morphology of the MarathoussaC r i c e t i n u s resembles that of C. bere m e n d e n -sis from Hungary (Hir 1994). However, theteeth from Marathoussa are slightly larg e rthan those from Beremend 15, the type locali-ty of that species (Figs. 6-10). Furthermore,there are some small differences in the mor-phology that separate the two samples. Onthe M1 of C. bere m e n d e n s i s the metacone isalways connected with the posteroloph wit-hout a metalophule, while on the M1s fromMarathoussa there is a metalophule throughwhich the metacone and the posteroloph areconnected. On the m1 from Marathoussa theposterosinusid is broader than that of the m1from Beremend 15. Finally, on the m2 fromMarathoussa the buccal anterolophulid is bet-ter developed than that from Beremend 15.Thus, the material from Marathoussa diff e r sfrom C. bere m e n d e n s i s.

The dental features of the Marathoussa samp-le are similar to those of C. koufosi, describedby Koliadimou (1996), from Ravin Vo u l g a r a k i s.According to her descriptions, the M1 has a

68

DEINSEA 8, 2001

69


preanterocone, a metalophule, a rare parastyleand a mesoloph or a spur labially to the ante-rolophule. The m1 has an anteroconid dividedin two cuspulids rather high above the roots,a very low, I-shaped anterolophid and nomesolophid. The presence of all these featu-res on the molars from Marathoussa indicatessimilarities with C. koufosi. All the molarsfrom Marathoussa have approximately thesame size as those from Ravin Vo u l g a r a k i s(Figs. 6-10). Thus, the Marathoussa sample ishere described as C. koufosi.

Family: Arvicolidae GR AY, 1821Subfamily: Arvicolinae GR AY, 1821Tribe: Arvicolini KR E T Z O I, 1955Genus: M i m o m y s FO R S Y T H MA J O R, 1902M i m o m y s s p .

M a t e r i a l :M2, 1 sin.: 240.M3, 1 sin.: 611 .

M e a s u re m e n t s :M2: L=2.05, W = 1 . 1 5M3: L=2.20, W=1.15

D e s c r i p t i o n :M 2 (Plate 5, 1): It consists of the AL1, T 2 ,T3 and T4. The enamel-free areas are on thebuccal and the lingual side of the AL1 and onthe posterior edge of the T4. The enamel isthinner on the posterior sides of the trianglesand thicker on the anterior ones ("M i m o m y s-type"). The synclines are closed at the base ofthe tooth and there is abundant crown-cementin the BRA1, BRA2 and the LRA2. There isa linea sinuosa very low under the crown, butthe roots are absent.M 3 (Plate 5, 2): It consists of the AL1, T 2 ,

Figure 6 Scatter diagram comparing the dimensions of the M1 of C. k o u fo s i from Marathoussa and Ravin Vo u l g a rakis (Koliadimou

1996) and the mean values of C. g ri t z a i from Odessa (Topachevski & Sko rik 1992), C. va ri a n s from Choukoutien (Zheng 1984),

C. b e r e m e n d e n s i s from Beremend 15 and C. e u r o p a e us from Csarnota 2 (Hir 1994).

T3, T4 and the PC. There are no enamel-freeareas. The occlusal surface shows threefields: the AL1, the T2 and the rest of thetooth. The enamel-thickness differentiation isof the "M i m o m y s - t y p e". The PC has analmost triangular shape, because of a smallsalient angle on its lingual side. The syncli-nes are closed at the base of the tooth. T h ecrown-cement is abundant in the LRA2,LRA3, BRA1 and BRA2 and a few in theBRA3. The linea sinuosa is in a large distan-ce from the crown, but there are no roots.

D i s c u s s i o n : The material from Marathoussais scanty. The "M i m o m y s- type" enamel-diff e-rentiation, the abundant crown-cement in thesynclines, the presence of the linea sinuosa

imply that the molars belong to the tribeArvicolini and, probably, to the genusM i m o m y s. The fact that the synclines closevery low and the crown of the teeth is high,lead to the assumption that they belong to arather advanced species. However, absence ofthe m1 in the available material makes acomparison with the known M i m o m y s s p e c i e simpossible. The dental features and the sizeof the teeth are very similar to those of M .s a v i n i from Ravin Voulgarakis (Koliadimou1996) and M. savini from Monte Peglia (Va nder Meulen 1973). However, the limitedmaterial does not allow a certain specificd e t e r m i n a t i o n .

70

DEINSEA 8, 2001

Figure 7 Scatter diagram comparing the dimensions of the M2 of C. k o u fo s i from Marathoussa and Ravin Vo u l g a rakis (Koliadimou

1996) and the mean values of C. g ritzai from Odessa (Topachevski & Sko rik 1992), C. va ri a n s from Choukoutien (Zheng 1984),

C. b e r e m e n d e n s i s from Beremend 15 and C. e u r o p a e u s from Csarnota 2 (Hir 1994).

71


Tribe: Microtini SI M P S O N, 1945Genus: M i c ro t u s SC H R A N K, 1797S u b g e n u s : A l l o p h a i o m y s KO R M O S, 1933Microtus (Allophaiomys) pliocaenicus(KO R M O S, 1933)

M a t e r i a l :M1, 15 dex.: 6, 11, 14, 59, 281, 284, 285, 296,292, 295, 296, 298, 299, 301, 342. 9 sin.: 9,13, 79, 282, 287, 297, 300, 310, 311. M2, 14 dex.: 7, 60, 71, 78, 84, 88, 90, 320,324, 325, 332, 338, 340, 344. 14 sin.: 80,313, 314, 315, 316, 317, 318, 319, 321, 322,327, 328, 331, 339.M3, 15 dex.: 4, 5, 73, 83, 257, 258, 259, 260,

263, 266, 267, 268, 275, 592, 595. 14 sin.:63, 77, 86, 92, 93, 190, 261, 262, 264, 265,271, 272, 274, 343. m1, 13 dex.: 57, 62, 66, 67, 68, 82, 242, 243,244, 246, 249, 250, 273. 24 sin.: 1, 2, 3, 55,56, 58, 61, 74, 76, 85, 94, 245, 247, 248, 251,252, 253, 254, 255, 256, 269, 270, 591.m2, 7 dex.: 8, 72, 81, 278, 288, 291, 294.12 sin.: 12, 64, 70, 276, 277, 279, 280, 283,289, 290, 293, 341. m3, 14 dex.: 69, 75, 91, 158, 302, 304, 305,307, 309, 312, 323, 329, 337, 594. 9 sin.: 65,87, 303, 306, 308, 333, 335, 336, 593.

M e a s u re m e n t s: see Tables 19 and 20.

Figure 8 Scatter diagram comparing the dimensions of the M3 of C. k o u fo s i from Marathoussa and from Ravin Vo u l g a ra k i s

(Koliadimou 1996) and the mean values of C. g ri t z a i from Odessa (Topachevski & Sko rik 1992), C. va ri a n s from Chouko u t i e n

(Zheng 1984), C. b e r e m e n d e n s i s from Beremend 15 and C. e u r o p a e u s from Csarnota 2 (Hir 1994).

D e s c r i p t i o n :All the upper molars have closed triangles.There is no differentiation of the enamelthickness between the posterior and the ante-rior part of the anticlines. There is abundantcrown-cement in the synclines and there areno roots.M 1 : It consists of the AL2, T1, T2, T3 andT4. There are four enamel-free areas: two onboth lingual and buccal sides of AL2, one onthe buccal side of T1 and one on the posteriorpart of T4 (in the part of the tooth that con-tacts to the M2).M 2 : It consists of the AL1, T2, T3 and T 4 .There are three enamel-free areas, two on thetwo sides of AL1 and one on the posterior

part of T 4 .M 3 (Plate 5, 6-8): It consists of the AL1, T 2 ,T3, T4, T5 and PC1. There are two lingualand three buccal synclines. The buccal re-ent-rant angles are shallower than the lingualones, while the BRA3 is the most shallowand the BRA2 the deepest re-entrant angle.The LRA2 and LRA3 are deep. The occlusalsurface shows three fields: in eight specimensthe first is AL1, the second T2 and T3 thatare confluent and the third the rest of themolar; in one specimen (MAR 93), T2 andT3 are separated, while T3 and PC1 are conf-luent. In the remaining nineteen specimensthere are four fields: the AL1, the T2, the T 3and the rest of the tooth. The buccal anticli-

72

DEINSEA 8, 2001

Figure 9 Scatter diagram comparing the dimensions of the m1 of C. k o u fo s i from Marathoussa and from Ravin Vo u l g a ra k i s

(Koliadimou 1996) and the mean values of C. g ri t z ai from Odessa (Topachevski & Sko rik 1992), C. va ri a n s from Chouko u t i e n

(Zheng 1984), C. b e r e m e n d e n s i s from Beremend 15 and C. e u r o p a e u s from Csarnota 2 (Hir 1994).

73


nes are smaller than the lingual ones and lessrounded, while the lingual ones are large withrounded ends. The PC is rounded, short and itmay bear an incipient lingual re-entrant angle(LRA4). The BRA1, BRA2, LRA2 and LRA3have always a lot of crown-cement, whileBRA3 and the incipient LRA4 (if present) areusually without crown-cement. Only in twoout of twenty-four specimens there is crown-cement in the BRA3. In a few specimens theenamel is a little thicker in the posterior partof the anticlines. In the following descriptionof the variability of the posterior cap of theM3, Rabeder’s (1981) morphotypes are givenin parentheses. Four specimens (18.2 %) havea very simple and rounded PC1, without anytrace of BRA3 or LRA4 ("p ro s i m p l e x- t y p e " ) .One specimen (4.6 %) has a PC1, as above,

but the fields of T2 and T3 are confluent( "p a r a p ro s i m p l e x-type"). Three specimens(13.6 %) are intermediate between these twotypes ("p ro s i m p l e x - p a r a p ro s i m p l e x- t y p e " ) .Ten specimens (45.5 %) have a longer PC1with BRA3, usually with crown-cement, anda LRA4 without crown-cement ("s i m p l e x-type"). One specimen (4.6 %) is characterisedas "p ro s i m p l e x - s i m p l e x-type", as it is inter-mediate between these two types. One otherspecimen is characterised as "s i m p l e x - p a r a -s i m p l e x-type", because the PC1 is the sameas in the "s i m p l e x-type", but the T2 and theT3 are not completely separated from eacho t h e r. One specimen has a PC1 with BRA3and LRA4 with crown-cement and a smallLSA5 ("c o m p l e x-type"). Finally, one speci-men has a well-developed BRA3, a small

Figure 10 Scatter diagram comparing the dimensions of the m2 of C. k o u fo s i from Marathoussa and from Ravin Vo u l g a ra k i s

(Koliadimou 1996) and the mean values of C. g ri t z a i from Odessa (Topachevski & Sko rik 1992), C. va ri a n s from Chouko u t i e n

(Zheng 1984), C. b e r e m e n d e n s i s from Beremend 15 and C. e u r o p a e us from Csarnota 2 (Hir 1994).

BSA4 and an almost absent LRA4 ("p r a e -o e c o n o m u s- t y p e " ) .

All lower molars have thicker enamel in theanterior and central part of the triangles,while it is thinner in the posterior part of thetriangles ("M i c ro t u s-type"). There are nor o o t s .

m 1 (Plate 5, 3-5): It consists of the AC2, T 5 ,T4, T3, T2, T1 and PL. There are four lingualand three buccal synclines, all with abundantcrown-cement. The occlusal surface showsfour fields consisting of the PL, T1, T2 andT3 and one more of the ACC. There are threeenamel-free areas: two on each side of the PLand one on the anterior part of the ACC. T h emorphology of the ACC shows a wide varia-tion and Chaline (1972) distinguished severalmorphotypes to describe it. Ten specimens(33.3 %) have a very simple AC1 that com-municates with the T5 and T4 ("p re h i n t o n i-type"). Three specimens (10 %) have a moreclosed and rounded AC1 and a very small T 7starts appearing ("h i n t o n i-type"). Twelve spe-cimens (40 %) are intermediate betweenthese two types ("p re h i n t o n i - h i n t o n i- t y p e " ) .Two specimens (6.7 %) have a separated, rat-her elongated AC1, while the T7 is welldeveloped ("h i n t o n i - g re g a l o i d e s- t y p e " ) .F i n a l l y, three specimens (10 %) have a ratherelongated AC1 without any sign of T 7( "l a g u ro i d e s- t y p e " ) .m 2 : It consists of the T4, T3, T2, T1 and PL.There are two lingual and two buccal syncli-nes with abundant crown cement, while innine specimens there is a little crown cementin front of T3 and in one specimen there is alittle crown cement in front of T4. The lingu-al re-entrant angles are deeper than the buccalones and the lingual salient angles are biggerthan the buccal ones. There are three enamel-free areas: two in each side of the PL and onein the anterior part of T4. The occlusal sur-face shows four or five fields: PL, T1 and T 2are closed, while in five out of twenty speci-mens the T3 and T4 communicate with eachother and form one field; in the rest (fifteen

specimens) all the fields are closed.m 3 : It consists of the T4, T3, T2, T1 and PL.In fifteen specimens there are three dentinefields: the PL, the T1 together with T2 andthe T3 together with the T4. In the remainingeight specimens T1 and T2 are closed andonly T3 and T4 communicate to a greater orlesser degree. The position of the PL i sslightly oblique in relation to the longitudinalaxis of the tooth. Other features are as in m2.

D i s c u s s i o n :The specimens from Marathoussa have abun-dant crown-cement in the synclines, lackingof roots and enamel-thickness diff e r e n t i a t i o nof the "M i c ro t u s-type". All these featurescharacterise the genus A rv i c o l a of the tribeArvicolini and all the genera of the tribeMicrotini (Koliadimou 1996). However, theteeth of Arvicola are much longer and theyhave more synclines than the specimens fromMarathoussa. So, it is obvious that the molarsfrom Marathoussa belong to the tribe Microtini.

We will discuss the m1s and M3s, as theother molars are not diagnostic. The morpho-logy of the ACC of the m1 shows a widevariation and it can be used for specific deter-mination. Three main morphotypes wereseparated by Van der Meulen (1973), basedon the number of the closed triangles in theocclusal surface of the m1. The specimensfrom Marathoussa belong to the morphotype1 (AC forms one field): twelve of them (40%) belong to morphotype 1a (parts of the A Cbroadly confluent) and eighteen (60 %) tomorphotype 1b (narrow connections betweenthe parts of the AC). According to Van derMeulen (1973) morphotype 1 molars charac-terise the subgenus A l l o p h a i o m y s. In additionto the three main morphotypes, the same aut-hor distinguished six additional types of theAC2. The m1s from Marathoussa belong tothe following types: "arvicolid-type" 20 % (6molars), "arvicolid-hintonid-type" 63.3 % (19molars), "hintonid-type" 6.7 % (2 molars),"arvicolid-ratticepid-type" 10 % (3 molars).The majority of the specimens (90%) belong

74

DEINSEA 8, 2001

75


to the "arvicolid-hintonid-type" that characte-rises a primitive species of A l l o p h a i o m y s .Van der Meulen (1973) also proposed the useof three ratios to demonstrate the diff e r e n c ebetween and the variation of the main morp-hotypes. These ratios are: A/L, B/W and C/W.The most important ratio is A/L, whichexpresses the relative length of the AC incomparison to the total length of the molar. Itis calculated according to A/L=a/L*100. T h efrequency distribution of A / L is expected toshift to higher values from older to youngerassemblages. Furthermore, he uses the meanvalues of A / L (M-A/L) to illustrate the in-creasing length of the AC. The M- A / L f o rthe assemblage from Marathoussa is 43.6( Table 20), which is only slightly smallerthan the same parameter for the assemblage ofAllophaiomys p l i o c a e n i c u s from Beftia-2 andN a g y h a r-sanyhegy-2 (M-A/L=43.7) (Van derMeulen 1973). The proposed values for M - A / L and M-B/W, by Van der Meulen(1973), for the species M. (Allo-phaiom y s )p l i o c a e n i c u s are: 42.0=M-A/L <44.5 and23.0=M-B/W<28.0. The values of these para-meters of the assemblage from Marathoussafall within these limits (Table 20, Fig. 11 ) .

According to Chaline’s (1972) separation inmorphotypes, the specimens from Marathoussabelong to the following morphotypes: "p re -h i n t o n i" 33.3 % (10 molars), "l a g u ro i d e s" 10% (3 molars), "p re h i n t o n i - h i n t o n i" 40 % (12molars), "h i n t o n i" 10 % (3 molars) and "h i n -t o n i - g re g a l o i d e s" 6.7 % (2 molars). All thesemorphotypes characterise the subspeciesM. (A.) pliocaenicus laguroides (KO R M O S,1932), although Chaline (1972) gives for thesubspecies a length between 2.12-2.52(ML=2.30), which is smaller than that of thespecimens from Marathoussa (ML=2.624).H o w e v e r, Van der Meulen (1973) andRabeder (1981) consider the subspecies M.(A.)pliocaenicus laguroides to be synonymous toM.(A.) pliocaenicus pliocaenicus and, thus,constitute the species M.(A.) pliocaenicus.

According to Rabeder (1981), M3s can also

be separated into morphotypes. In the assem-blage from Marathoussa, the following typesare recognized: "p ro s i m p l e x" 18.2 % (4molars), "p a r a p ro s i m p l e x" 4.6 % (1 molar),"p ro s i m p l e x - p a r a p ro s i m p l e x" 13.6 % (3molars), "s i m p l e x" 45.5 % (10 molars), "s i m -p l e x - p a r a s i m p l e x" 4.6 % (1 molar),"c o m p l e x" 4.6 % (1 molar), and "p r a e o e c o n o -m u s" 4.6 % (1 molar). The above morphoty-pes (except for "c o m p l e x" and the "praeoeco-nomus") belong to the s i m p l e x-Group, whichappears to be the most primitive group. T h etype "c o m p l e x" belongs to the c o m p l e x- G r o u pand the type "p r a e o e c o n o m u s" to the o e c o n o -m u s-Line, but they represent a very smallpercentage in the assemblage fromMarathoussa. More than 50 % (91 %) of theM3s from Marathoussa belong to the s i m p l e x-Group and, thus, according to Rabeder(1981), the assemblage belongs to the speciesM.(Allophaiomys) pliocaenicus.

The primitive species Allophaiomys deuca -l i o n KR E T Z O I, 1969 is a medium-sizedM i c ro t u s species (Table 21). The three para-meters for the m1 are: M-A/L<42 (Fig. 11 ) ,M-B/W=33, M-C/W=20. The m1 has a verysimple and short ACC and it shows a verybroad connection between the AC2 and thetriangles T4 and T5 (morphotype 1a of Va nder Meulen 1973). The three parameters forthe assemblage from Marathoussa sample arenot in accordance with the values that aregiven for this species. Furthermore, the morp-hology of the ACC is not as simple as that ofA. deucalion, since almost two thirds of thespecimens belong to the type 1b and only onethird of them belongs to the type 1a. Becauseof these reasons, the assemblage fromMarathoussa does not belong to the speciesA. deucalion. Since the M-A/L of theMarathoussa m1 is 43.6 (higher than that ofA. deucalion), it seems that these specimensbelong to a younger species.

The species Allophaiomys chalinei AL C A L D EE T A L., 1981 is a large-sized vole. The enamelin the lower molars is undiff e r e n t i a t e d .

M o r e o v e r, the values for the M-A/L are low( Table 21, Fig. 11). The anteroconid complexis very characteristic, with broad and shorte-ned AC1, associated with a very straight ante-roconid ‘neck’, which is reflected in the lowvalue of B/W. On the other hand, the M-C/Wreaches high values. Thus, the m1 displays avery accentuated pitymyoid shape, widelyconfluent T4 and T5 (Van der Meulen &Doukas, in press). The specimens fromMarathoussa cannot belong to this species,because their length is much smaller, there isa clear enamel-thickness differentiation, theAC and the T4-T5 are more confluent and theconnection between T4 and T5 is narrower.The fact that the value of M-A/L of A. chali -n e i is lower leads to the conclusion that spe-cimens from Marathoussa must belong to ayounger species.

The species A. nutiensis CH A L I N E, 1972( =A l l o p h a i o m y s s p .A in Van der Meulen

1973) is a medium-sized vole (Table 21) withelongated and complicated anteroconid com-plex in m1, showing conspicuous narrowingof the area that connects the AC2 with theT4-T5. The values of the three parameters forthe m1s are: 44.5=M-A/L<48.5 (Figure 11 ) ,8.0<M-B/W=14.5, M-C/W=20.0 (Van derMeulen 1973). According to Chaline (1972),the species is characterised by a mixture ofmorphotypes and the presence of two groups.The first group consists of small-sized m1s,belonging to the morphotypes "l a g u ro i d e s" ,"p l i o c a e n i c u s", "h i n t o n i" and "g re g a l o i d e s" .The second group consists of larg e - s i z e dm1s, belonging to the morphotypes "p re r a t t i -c e p o i d e s", "r a t t i c e p o i d e s" and "n i v a l o i d e s" .The complete absence of morphotypes of thesecond group from the Marathoussa sample, aswell as the lower value of M-A/L and M-C/Wand the higher value of the M-B/W indicate thatthe specimens from Marathoussa are diff e r e n tand represent a species older than A. n u t i e n s i s.

76

DEINSEA 8, 2001

Figure 11 Diagram showing the range of the values and the mean value of the parameter M-A/L for seve ral A l l o p h a i o m y s s p e c i e s

( r e ferences as on Ta ble 21).

77


The species A. burg u n d i a e CH A L I N E, 1972( =A l l o p h a i o m y s s p .B in Van der Meulen1973) is a medium-sized vole (Table 21),with elongated and complicated ACC, usuallyshowing narrowing of the area that connectsT4 and T5 and a deep LRA4. The values ofthe parameters for the m1 are: 44.5<M-A/L<46.3 (Figure 11), 19.5<M-B/W<23.0,14.0=M-C/W<20.0 (Van der Meulen 1973).The AC2 is usually "n i v a l i d" or "n i v a l i d - r a t -t i c e p i d" (Chaline 1972). These types areabsent from the molars from Marathoussa.Furthermore, the value of the M-A/L is lowerand the value of the M-B/W is higher thanthose for A. burg u n d i a e . Therefore, the spe-cimens from Marathoussa are distinguishedfrom A. burg u n d i a e and it seems that theybelong to an older species.

The species A. ru f f o i PA S A, 1947 is a mediumto large-sized vole (Table 21). Its m1 has anACC with narrow connection between T4 andT5. The LRA4 and BRA3 are relatively shal-low and, thus, AC2 and the two triangles areconfluent. The m1s belong to the "laguroides"and the "p re r a t t i c e p o i d e s" type. The rangesof the three parameters of the A. ru f f o iassemblage from Cava Sud are: 41<M-A/L<45 (Fig. 11), 19<M-B/W<34 and 12<M-C/W<25 (Van der Meulen & Doukas inpress). The mean values of the parameters forthe Marathoussa sample are within theselimits. However, the dominant morphotype ofthe m1s from Marathoussa is "p re h i n t o n i - h i n -t o n i", which is totally absent in A. ru f f o i. T h i smeans that the connection between the A C 2and the T4-T5 is narrower in the specimensfrom Marathoussa. Furthermore, the meanlength of the m1s from Marathoussa is muchsmaller than that of the m1s of A. ru f f o i. T h especimens from Marathoussa do not belongto A. ru f f o i, but they are very closely relatedto it (based on comparisons to a collection ofA. ru f f o i from Cava Sud, Italy, in UtrechtUniversity).

Rabeder (1981) described two new species of

Allophaiomys. A. praehintoni is a primitivespecies with important morphological varia-bility (Table 21). The two dominant morpho-types are "p re h i n t o n i" followed by "eoratti-ceps" (anteroconid complex elongated, con-fluent T4-T5, very small T7 and LSA5, ratherdeep LSA4 that prevents wide connectionbetween the AC2 and T4-T5). The value ofthe M-A/L is between 44.5 and 45 (Fig. 11 ) .In the sample from Marathoussa there are nomolars of the "e o r a t t i c e p s" type and the valueof M-A/L is lower than those given for thiss p e c i e s .

A. superpliocaenicus is characterised by thepresence of the morphotypes "s u p e r p l i o c a e n i -c u s", "s u p e r l a g u ro i d e s" and "p l i o c a e n i c u s" ,while there are no molars belonging to the"p r a e h i n t o n i" or "h i n t o n i" types (Table 21).Furthermore, the value of the M-A/L i sbetween 44 and 45 (Fig 11) (Rabeder 1981).In the Marathoussa sample there are no speci-mens belonging to these types and the valueof the M-A/L is lower than the minimumvalue given for this species. Therefore, thestudied sample is well distinguished from A. superpliocaenicus.

A species of M i c ro t u s called M. pitymyoideshas been recognised in the locality RavinVoulgarakis (RVL) (Koliadimou 1996). T h em1s of the assemblage belong to the follo-wing types (Chaline' s separation, 1972):"p re h i n t o n i - h i n t o n i" (7.1 %), "h i n t o n i - g re g a -l o i d e s" (35.8 %), "g re g a l o i d e s" (9.5 %), "g re -g a l o i d e s - a rv a l o i d e s" (14.3 %), "a rv a l o i d e s"(21.4 %), "p re n i v a l o i d e s" (2.3 %), "n i v a l o i -d e s" (4.8 %) and "n i v a l o i d e s - r a t t i c e p o i d e s"(4.8 %). The values of the parameters are: M-A/L= 48.13, M-B/W= 7.38 and M-C/W=20.77. It is clear that the assemblages fromMarathoussa and Ravin Voulgarakis are diff e-rent. The former assemblage has m1 of thetypes "p re h i n t o n i", "h i n t o n i" and their inter-mediates, whereas these types are representedin the latter assemblage by a frequency of 7.1% only. The values of the parameters of the

two assemblages also differ: the value of M-A / L for the Ravin Voulgarakis sample ismuch higher than that for the Marathoussasample and the value of M-B/W for the for-mer sample is much lower than that of thelatter sample (Table 21). Considering thesed i fferences, we can assume that M i c ro t u sp i t y m y o i d e s from Ravin Voulgarakis is moreadvanced than the M i c ro t u s f r o mMarathoussa.

The species A. pliocaenicus is characterisedby a simple and short anteroconid complex inthe m1, showing no tendency to subdivide.The BRA3 and LRA4 are of equal and consi-derable depth and the m1s belong mostly to"p re h i n t o n i", "h i n t o n i" and their intermediatetypes. The three parameters have the follo-wing values: M-A/L<44.5, M-B/W=23.0 andM-C/W=20.0 (Van der Meulen 1973). All theabove features are present in the specimensfrom Marathoussa and, thus, the sample isdetermined as A. pliocaenicus.

Tribe: Lagurini KR E T Z O I, 1955Genus: L a g u ro d o n KR E T Z O I, 1956Lagurodon arankae (KR E T Z O I, 1954)

M a t e r i a l :M1, 3 dex.: 185, 236, 238. 4 sin.: 171, 172,181, 602.M2, 4 dex.: 178, 186, 233, 235. 3 sin.: 189,188, 237.M3, 4 dex.: 162, 163, 164, 231. 5 sin.: 23,24, 165, 187, 603.m1, 8 dex.: 19, 20, 159, 160, 161, 225, 229,234. 9 sin.: 16, 17, 18, 21, 22, 226, 227, 228,241. m2, 7 dex.: 166, 167, 169, 170, 182, 184,230. 2 sin.: 168, 180.m3, 4 dex.: 173, 174, 175, 239. 4 sin.: 176,177, 179, 183.


D e s c r i p t i o n : All the molars have no roots, nocrown-cement and the enamel-thickness dif-ferentiation is of the "M i c ro t u s-type"

(thicker on the posterior part of the triangles). M 1 : It consists of the AL2, T1, T2, T3 andT4. In four of the studied specimens there isa small anticline, like a beak, in the LRA2.Each triangle constitutes a closed field, wit-hout any communication with the othertriangles. The enamel-free areas are on thebuccal and lingual side of the AL2 and on theposterior edge of T 4 .M 2 : It consists of the AL1, T2, T3 and T4. Insix specimens (the seventh is broken) there isa small anticline in the LRA2. The fields onthe occlusal surface and the enamel-free areasare as on the M1. M 3 (Plate 5, 9-10): It consists of the A L 1 ,T2, T3, T4 and PC. Among the four speci-mens that retain the AL1, the two have ena-mel-free areas on both lingual and buccalside of AL1, while the other two have onlyone enamel-free area on the lingual side ofAL1. The LRA2 is long but without a micro-anticline as on the M1 and M2. The T2 is, inmost cases, more developed than the T4. A l lmolars belong to the morphotype "a r a n k a e"(PC rounded or oval and not very elongated).The molars consist of only one field. One ofthe molars has a small islet on its PC.m 1 (Plate 5, 11-12): It consists of the PL, T 1 ,T2, T3, T4, T5, T6 and AC1. The fields onthe occlusal surface are five: the PL, the T 1 ,the T2, the T3 and the anteroconid complex.The enamel-free areas are on the anterioredge of the AC1, on the buccal side of the T 6and on both the lingual and buccal sides ofthe PL. Ten specimens (83 %) belong to themorphotype "a r a n k a e" (BRA4 developed)and two specimens (17 %) to the morphotype"p a n n o n i c u s" (BRA4 almost absent). m 2 : It consists of the PL, T1, T2, T3 and T 4 .Each one of these parts form one field, withvery restricted communication among them.The enamel-free areas are on both lingual andbuccal sides of the PL and on the anterioredge of the T 4 .m3: It consists of the PL, T1, T2, T3 and T4,with very narrow openings from one to the o t h e r.The enamel-free areas are as on the m2.

78

DEINSEA 8, 2001

79


D i s c u s s i o n :The absence of crown-cement and roots aswell as the "M i c ro t u s-type" enamel-diff e r e n t -iation indicate the tribe Lagurini. The welldeveloped T6, the absence of a T7, the domi-nance of the "a r a n k a e" morphotype and thepresence of four enamel-free areas on the m1,the presence of a micro-anticline in the LRA2of the M1 and M2 and the dominance of the"a r a n k a e" morphotype for the M3 are thedental features that distinguish the assembla-ge from the genera L a g u ru s, P ro l a g u ru s a n dE o l a g u ru s and suggest similarities withL a g u ro d o n (Koliadimou 1996).

There are two species of L a g u ro d o n: L. aran -k a e and L. praepannonicus ( K o l i a d i m o u1996). The m1 of L . p r a e p a n n o n i c u s has anincipient T6, the BRA4 is completely absentand the enamel-free area of the AC expandsto the buccal edge of the T6 and, thus, theenamel-free areas of the molar are three( Topachevski 1973). The description of themolars, as given above, agrees with thedescription of the molars of the species L .a r a n k a e. The complete absence of m1s of themorphotype "t r a n s i e n s" and the very smallrepresentation of the morphotype "p a n n o n i -c u s", both of which are more advanced thanthe morphotype "a r a n k a e", show the ratherold age of the material (Koliadimou 1996).

B I O C H RO N O L O G YThe micromammalian fauna of Marathoussaincludes several species and some of themcan help with the dating of the fauna. T h eMarathoussa fauna consists of the followingspecies: Desmaninae indet., C rocidura korn -f e l d i, S o rex minutus, S o rex (Dre p a n o s o re x )p r a e a r a n e u s, Asoriculus gibbero d o n,B e remendia fissidens, S p e r m o p h i l u s s p . ,Sicista c f . s u b t i l i s, Apodemus sylvaticus/f l a v i -c o l l i s, Cricetinus koufosi, M i m o m y s s p . ,M i c ro t u s (A l l o p h a i o m y s) p l i o c a e n i c u s a n dL a g u rodon arankae.

The insectivores cannot help with the datingof the fauna, as the stratigraphical range of

most of them is quite large (Fig. 12). On thec o n t r a r y, the rodents of the fauna can giveindications for the age of the Marathoussafauna. The presence of A. sylvaticus/flavicol -l i s, M . ( A . ) p l i o c a e n i c u s and L. arankae s u g g e s t san Early Biharian age (Fig. 12). Moreover,morphological comparisons for these speciesallow more definite conclusions. T h eM a r a t h o u s s a A p o d e m u s represents a formwith a mixture of morphological features thatare present in the species A. sylvaticus and A .f l a v i c o l l i s. Both species appeared at thebeginning of Pleistocene (Early Biharian) andexist until nowadays. The similarity of theMarathoussa A p o d e m u s with that from RavinVoulgarakis indicates that both faunas areclose in age. The latter locality has beendated to Early-Middle Biharian and, moreprecisely at the Nagyharsanyhegy phase(Koliadimou 1996).

The arvicolid M.(A.) pliocaenicus gives moredetails about the age of the fauna. The FirstAppearance Datum (FAD) of M i c ro t u s( A l l o p h a i o m y s ) is considered to mark the endof the Villanyian and the beginning of theBiharian (De Giuli & Torre 1984; Fejfar et al.1997) and this time marker can be used with-in the Holarctic region (Fejfar & Heinrich1989). The type locality of M.(A.) pliocaeni -c u s is Betfia 2, Romania, dated at the EarlyBiharian (Van der Meulen 1973). Since then,it has been found in various Eurasian locali-ties (Fig. 12), but it was unknown in Greece.A more detailed comparison of theMarathoussa M i c ro t u s with other known spe-cies provides more biochronological eviden-ces. The parameter M-A/L (as proposed byvan der Meulen 1973) for the Marathoussam1 is only slightly different from those ofBetfia 2 and Nagyharsanyhegy 2(Romania/Hungary), indicating similar agefor the three faunas. Moreover, most of them1 from Betfia 2 and all of them fromNagyharsanyhegy 2 belong to the morpho-type 1a (Van der Meulen 1973), while theMarathoussa m1 mainly belong to the mor-photype 1b. This suggests a slightly younger

80

DEINSEA 8, 2001

age for Marathoussa than that of the othertwo sites, which are dated at the EarlyBiharian (Fig. 12). The presence of a moreevolved form of Microtini with higher valuesfor the parameter M-A/L in Monte Peglia 63,11, 12 indicates that Marathoussa is olderthan the Nagyharsanyhegy phase, at whichMonte Peglia is dated. In the locality RavinVoulgarakis the more evolved species M .p i t y m y o i d e s has been reported. The fauna ofRavin Voulgarakis has been dated at Middle-Late Biharian (Nagyhars-anyhegy phase)(Koliadimou 1996, and Fig. 12). Taking intoaccount all the above mentioned, theMarathoussa fauna is dated at the EarlyBiharian (Betfia phase) and is slightly youn-ger than Betfia 2 and older than MonteP e g l i a .

The presence of the species L. arankaesuggests a possible Early Biharian age for theMarathoussa fauna as well, since the morpho-types for the m1 of the Marathoussa sampleare more primitive than those of the RavinVoulgarakis sample (Koliadimou 1996).Therefore, it is confirmed that the Mara-thoussa fauna has an Early Biharian age.

AC K N OW L E D G E M E N T SWe wish to thank Dr H. de Bruijn, Dr A . J .van der Meulen, Prof. C.S. Doukas, Dr J. vanDam, Dr J.W. F. Reumer, Ms W. Wessels andDr L. van den Hoek Ostende for having pro-vided us with literature, comparative collec-tions and for their helpful comments and sug-gestions. Mr J. Luteijn made the drawings ofthe Arvicolidae and retouched the plates, MrW. den Hartog did the photographic work, S.Boskamp and E. Vassiliadou made linguisticimprovements on the text and I. Sylvestrouand E. Tsompachidou helped with the washing of the sediments.

R E F E R E N C E SAgusti, J., 1990 - The Plio-Pleistocene arvicolid

succession in Southern and Eastern Spain - Proceedings of the International Symposium Evolution, Phylogeny and Biostratigraphy of Arvicolids (Rodendia, Mammalia): 21-32

Chaline, J., 1972 - Les rongeurs du Pleistocene moyen etsupérieur de France (systematique - biostratigraphie - paléoclimatologie) - Cahiers de Paléontologie, Centre National de la Recherche Scientifique, Paris V I I

Daams, R. & Freudenthal, M., 1988 - Cricetidae (Rodentia) from the type-Aragonian; the genus M e g a c r i c e t o d o n - Scripta Geologica, Special Issue 1: 3 9 - 1 3 2

De Bruijn, H., 1966 - On the mammalian fauna of the H i p p a r i o n-Beds in the Calatayud-Teruel basin (Prov. Zaragoza, Spain) IIA, Part II. The Gliridae - Proceedings of the Koninklijke Nederlandse Akademie van Wetenschappen (B) 69 (3): 1-21

De Giuli, C. & Torre, D., 1984 - A microfauna with Allophaiomys pliocaenicus from Gargano (Southern Italy) - Palaeontographica Italica 73 (n. ser. vol. 43): 129 pp.

Doukas, C.S., Van den Hoek Ostende, L.W., Theocharopoulos, C.D. & Reumer, J.W. F., 1995 - T h eVertebrate Locality Maramena (Macedonia, Greece) at the Turolian-Ruscinian Boundary (Neogene). 5. Insectivora (Erinaceidae, Talpidae, Soricidae, Mammalia) - Münchener Geowissenschaftlichen Abhandlungen (A) 28: 43-64

E n g e s s e r, B., 1989 - The Late Tertiary small mammals ofthe Maremma region (Tu s c a n y, Italy). 2n d part: Muridae and Cricetidae (Rodentia, Mammalia) - Bollettino della Societa Paleontologica Italiana 28 (2-3): 227-252

F e j f a r, O. & Heinrich, W.-D., 1989 - Muroid Rodent Biochronology of the Neogene and Quaternary in Europe - pp. 91-118 in: E. Lindsay, V. Fahlbusch & P.Mein (eds.) - European Neogene Mammal C h r o n o l o g y, Plenum Press, New Yo r k

F e j f a r, O., Heinrich, W.-D., Pevzner, M.A. & Vangengeim, E.A., 1997 - Late Cenozoic sequencesof mammalian sites in Eurasin: an updated correlation- Palaeogeography, Palaeoclimatology, Palaeoecology 133: 259-288

Filippucci, M. G., Storch, G. & Macholan, M., 1996 - Taxonomy of the genus S y l v a e m u s in western Anatolia - morphological and electrophoretic

81


82

DEINSEA 8, 2001

evidence (Mammalia: Rodentia: Muridae) - S e n c k e n b e rgiana biologica 75 (1/2): 1-14

Green, M., 1977 - Neogene Zapodidae (Mammalia, Rodentia) from South Dakota - Journal of Paleontology 51: 996-1015

H i r, J., 1994 - Cricetinus bere m e n d e n s i s n. sp. (Rodentia,Mammalia) from the Pliocene fauna of Beremend 15

(S. Hungary) - Fragmenta Mineralogica et Palaeontologica 17: 71-89

H i r, J., 1997 - A short sketch of the evolution and stratigraphy of the Plio-Pleistocene cricetids (Rodentia, Mammalia) in Hungary - Folia Historico Naturalia Musei Matraensis 22: 43-49

Koliadimou, K.K., 1996 - Palaeontological and biostratigraphical study of the Neogene/Quaternary micromammals from Mygdonia basin - Ph.D Thesis, University of Thessaloniki (in Greek): 612 pp.

Kordos, L., 1988 - Karstocricetus skofleki gen. n., sp. n. and the evolution of the Late Neogene Cricetidae in the Carpathian Basin - Fragmenta Mineralogica et Palaeontologica 13: 65-88

Kostopoulos, D.S., 1996 - The Plio-Pleistocene Artiodactyls from Macedonia, Greece; systematics, p a l a e o e c o l o g y, biochronology - Ph.D Thesis, University of Thessaloniki (in Greek with English summary): 540 pp.

Kostopoulos, D.S., Vassiliadou, K. & Koufos, G.D., in press - The beginning of Pleistocene in the Balkan area according to the mammal record; palaeozoo-geographical approach - Annales Géologiques des Pays Helléniques

Koufos, G.D., Syrides, G.E., Kostopoulos, D.S. & Koliadimou, K.K., 1992 - Apollonia, a new vertebratesite in the Pleistocene of the Mygdonia basin (Macedonia, Greece); the first fossil fresh-water mollusks in the area - Comptes rendus de l’Académie des Sciences Paris 315 (II): 1041-1046

Koufos, G.D., Syrides, G.E., Kostopoulos, D.S. & Koliadimou, K.K., 1995 - Preliminary results about the stratigraphy and the palaeoenvironment of Mygdonia basin, Macedonia, Greece - Geobios MS 18: 243-249

Koufos, G.D. & Kostopoulos, D.S., 1997 - Biochronology and succession of the Plio-Pleistocenemacromammalian localities of Greece - Proceedings BiochromM 97, Mem. Tr a v. E.P.H.E., Inst. Montpellier 21: 619-634

Kretzoi, M., 1959 - A Csarnótai fauna és faunaszint - A

Magyar Allami Földtani Intézet Evi Jelentéze az 1959: 297-385

Kretzoi, M., 1959 - Insectivoren, Nagetiere und Lagomorphen der jüngstpliozänen fauna von Csarnóta im Villanyer Gebirge (Südungarn) - Vertebrata Hungarica 1: 237-246

Martin Suárez, E., 1988 - Une nouvelle espéce d ’A p o d e m u s (Rodentia, Mammalia) du Pliocène de ladépression de Guadix-Baza (Grenade, Espagne) - Geobios 21 (1): 523-529

Martin Suárez, E. & Freudenthal, M., 1993 - Muridae (Rodentia) from the Lower Turolian of Crevillente (Alicante, Spain) - Scripta Geologica 106: 11 - 3 4

M a y h e w, D.F., 1977 - Late Pleistocene small mammals from Arnissa (Macedonia, Greece) - Proceedings of the Koninklijke Nederlandse Akademie van Wetenschappen (B) 81(3): 302-321

Mein, P. & Freudenthal, M., 1971 - Les Cricetidae (Mammalia, Rodentia) du Néogène de Vi e u x -Collonges (Rhône, France). Partie 1: Le genre C r i c e t o d o n Lartet, 1851 - Scripta Geologica 5: 1-51

Michaux, J. & Pasquier, L., 1974 - Recherche sur les rapports espèce actuelle - espèce fossile chez les Rongeurs du groupe Apodemus sylvaticus- Apodemus f l a v i c o l l i s. Intéret pour la connaissance de l ’ Holocène - Bulletin de la Société Languedocienne

de Géographie 8 (3-4): 381-383Montuire, S., Sen, S. & Michaux, J., 1994 - The Middle

Pleistocene mammalian fauna from Emirkaya-2, Central Anatolia (Turkey): Systematics and Paleoenvironment - Neues Jahrbuch für Geologie und Paläontologie Abhandlungen 193 (1): 107-144

M u s s e r, G.G., Brothers, E.M., Carleton, M.D. & H u t t e r e r, R., 1996 - Taxonomy and distributional records of Oriental and European A p o d e m u s, with a review of the A p o d e m u s - S y l v a e m u s problem - Bonner Zoologische Beiträge 46 (1-4): 143-190

P a s q u i e r, L., 1974 - Dynamique évolutive d’un sous-genre de Muridae, A p o d e m u s (S y l v a e m u s). Etude biometrique des caractères dentaires de populations fossiles et actuelles d’Europe occidentale - These, Université des Sciences et Techniques du Languedoc, Académie de Montpellier.

Psilovikos, A., 1977 - Palaeogeographic development of Mygdonia valley and the lake (Langada-Volvi area) - Ph.D. Thesis, University of Thessaloniki (in Greek with English summary): 154 pp.

R a b e d e r, G., 1981 - Die Arvicoliden (Rodentia,

Mammalia) aus dem Pliozän und dem älteren Pleistozän von Niederösterreich - Beiträge zur Paläontologie von Österreich 8: 1-373

Repenning, C.A., 1967 - Subfamilies and Genera of the Soricidae - Geological Survey Professional Paper 565: 1-73

R e u m e r, J.W. F., 1984 - Ruscinian and early Pleistocene Soricidae (Insectivora, Mammalia) from Tegelen (The Netherlands) and Hungary - Scripta Geologica 73: 1-173

R e u m e r, J.W. F. & Doukas, C.S., 1985 - Early PleistoceneInsectivora (Mammalia) from Tourkobounia (Athens,Greece) - Proceedings of the Koninklijke Nederlandse Akademie van Wetenschappen (B) 88 (1): 111 - 1 2 1

R e u m e r, J.W. F., 1985 - The generic status and species of D re p a n o s o re x reconsidered (Mammalia, Soricidae) - Revue de Paléobiologie 4 (1): 53-58

R e u m e r, J.W. F., 1994 - De evolutiebiologie van de spitsmuizen (Mammalia, Insectivora, Soricidae). I. Anatomie, evolutie en biogeografie - Cranium 11 (1): 9-35

R e u m e r, J.W. F., 1995 - De evolutiebiologie van de spitsmuizen (Mammalia, Insectivora, Soricidae). II.De subfamilies Crocidosoricinae, Allosoricinae, Limnoecinae en Crocidurinae - Cranium 12 (2): 5 3 - 6 3

R e u m e r, J.W. F., 1995 - Insectivora (Mammalia: Soricidae, Talpidae) from the Late Pliocene of Frechen, Germany - Deinsea 2: 9-15

R e u m e r, J.W. F., 1996 - Quarternary Insectivora (Mammalia) from Southeastern France - Acta Zoologica cracoviensia 39(1): 413-426

R e u m e r, J.W. F., 1997 - De evolutiebiologie van de spitsmuizen (Mammalia, Insectivora, Soricidae). III. De subfamilie Soricinae en het literatuuroverzicht - Cranium 15 (1): 13-36

Rümke, C.G., 1985 - A review of fossil and recent Desmaninae (Talpidae, Insectivora) - Utrecht Micropaleontological Bulletins, Special Publication 4: 1-241

Schaub, S., 1930 - Quartäre und jungtertiäre Hamster - Abhandlungen der Schweizerischen Palaeontologischen Gesellschaft 49: 1-47

Tc h e r n o v, E., 1968 - A Pleistocene faunule from a karst fissure filling near Jerusalem, Israel - Verhandlungen Naturforschenden Gesellschaft Basel 79 (2): 161-185

Topachevski, V.A., 1973 - Grizguhi tamanskova Faounisticeskora kompleksa Krima - Akademia Nauk Ukranii, C.C.C.P., Naoukova Doumka: 235 pp.

Topachevski, V.A. & Skorik, A . F., 1992 - Neogenovi i Pleistocenovi nezshie chomiaoobrasnie ioga vostochnoi Evropi - Akademia Nauk Ukranii: 235 pp.

Van der Meulen, A.J., 1973 - Middle Pleistocene smaller Mammals from the Monte Peglia, (Orvieto, Italy) with special reference to the phylogeny of M i c ro t u s

(Arvicolidae, Rodentia) - Quaternaria 17: 1-119 Van der Meulen, A.J., 1978 - M i c ro t u s and P i t y m y s

(Arvicolidae) from Cumberland Cave, Maryland, with a comparison of some new and old world species - Annals of the Carnegie Museum 47 (6): 101-145

Van der Meulen, A.J. & Kolfschoten, T. van, 1986 - Review of the Late Turolian to Early Biharian mammal faunas from Greece and Turkey - Mem. Società Geologica Italiana 31: 201-211

Van de Weerd, A., 1973 - Rodentia from two Pleistocene fissure fillings near Athens - Proceedings of the Koninklijke Nederlandse Akademie van Wetenschappen (B) 76 (2): 148-166

Van de Weerd, A., 1976 - Rodent faunas of the Mio-Pliocene continental sediments of the Te r u e l -Alfambra region, Spain - Utrecht Micropaleon-tological Bulletins, Special Publication 2: 1-218

Zheng, S., 1984 - Revised determination of the fossil Cricetine (Rodentia, Mammalia) of Choukoutien

district - Vertebrata Palasiatica 22 (3): 179-197

re c e i ved 29 March 2000

accepted 26 October 2000

83


84

DEINSEA 8, 2001

PLATE 1 Desmaninae indet. 1: P4s (MAR 130), occlusal view. 2: M1d (MAR 35), occlusal view. 3: M2d (MAR 510), occlusalview. 4: p4d (MAR 487); a: occlusal view, b: buccal view. 5: m1s (MAR 42); a: occlusal view, b: buccal view. 6: m3s (MAR 39),occlusal view. Beremendia fissidens 7: P4-M1d (MAR 545), occlusal view. 8: M2d (MAR 492), occlusal view. 9: M3s (MAR 531),occlusal view. 1 0: m1s (MAR 478); a: occlusal view, b: buccal view. ( E n l a rgement x20)

85


P L ATE 1

86

DEINSEA 8, 2001

PLATE 2 Sorex minutus 1: Isup s (MAR 32); a: buccal view, b: occlusal view. 2: P4s (MAR 89), occlusal view. 3: M1s (MAR505), occlusal view. 4: Iinf d (MAR 128); a: buccal view, b: lingual view. 5: m1s (MAR 107); a: occlusal view, b: buccal view.6: m2d (MAR 443); a: occlusal view, b: buccal view. Asoriculus gibberodon 7: Isup d (MAR 391); a: buccal view, b: occlusal view. 8: M1d (MAR 95), occlusal view. 9: M2s (MAR 491), occlusal view. 10: M3d (MAR 534), occlusal view. 11: Iinf s (MAR422); a: buccal view, b: lingual view. 12: m1s (MAR 111); a: occlusal view, b: buccal view. Crocidura kornfeldi 13: Isup s (MAR429); a: buccal view, b: occlusal view. 14: Asup s (MAR 562); a: occlusal view, b: buccal view. 15: P4d (MAR 544), occlusalview. 16: M1d (MAR 544), occlusal view. 17: M2d (MAR 511), occlusal view. 18: M3s (MAR 536), occlusal view. 19: Iinf s(MAR 412); a: buccal view, b: lingual view. 20: p4s (MAR 570), occlusal view. 21: mandible with m1 and m2 d (MAR 468),buccal view. 22: m3d (MAR 464); a: occlusal view, b: buccal view. (Enlargement x20)

87


P L ATE 2

88

DEINSEA 8, 2001

PLATE 3 Drepanosorex praearaneus 1: Isup d (MAR 386); a: buccal view, b: occlusal view. 2: Asup d (MAR 601), occlusalview. 3: P4d (MAR 118), occlusal view. 4: M1s (MAR 49), occlusal view. 5: M2d (MAR 97), occlusal view. 6: M3s (MAR 533),occlusal view. 7: Iinf d (MAR 34); a: buccal view, b: lingual view. 8: Ainf s (MAR 565); a: occlusal view, b: buccal view. 9: p4s(MAR 137); a: occlusal view, b: buccal view. 10: mandible with m1, m2 and m3 s (MAR 51); a: buccal view, b: occlusal view.11: ascending ramus with the condyle s (MAR 620); a: lingual view, b: buccal view. (Enlargement x20)

89


P L ATE 3

90

DEINSEA 8, 2001

P L ATE 4

PLATE 4 Spermophilus sp. 1: M1,2d (MAR 378), occlusal view. 2: m3d (MAR 379), occlusal view. Sicista cf. subtilis. 3: m1d(MAR 597), occlusal view. Apodemus sylvaticus/flavicollis. 4: M1s (MAR 196), occlusal view. 5: M1s (MAR 197), occlusal view.6: M2d (MAR 147), occlusal view. 7: M2d (MAR 199), occlusal view. 8: m1d (MAR 204), occlusal view. 9: m1s (MAR 207),occlusal view. 10: m1s (MAR 209), occlusal view. Cricetinus koufosi. 11: M1d (MAR 345), occlusal view. 12: M2s (MAR 348),occlusal view. 13: m1d (MAR 150), occlusal view. (Enlargement x20)

91


P L ATE 5

PLATE 5 Mimomys sp. 1: M2s (MAR 240), occlusal view. 2: M3s (MAR 611), occlusal view. Microtus (Allophaiomys) pliocaeni -cus. 3: m1s ‘laguroides’ type (MAR 248), occlusal view. 4: m1s ‘prehintoni’ type (MAR 2), occlusal view. 5: m1s ‘prehintoni-hintoni’ type (MAR 251), occlusal view. 6: M3d ‘praeoeconomus’ type (MAR 275), occlusal view (fragment). 7: M3s ‘simplex’type (MAR 261), occlusal view. 8: M3d ‘prosimplex’ type (MAR 268), occlusal view. Lagurodon arankae. 9: M3d ‘arankae’ type(MAR 231), occlusal view. 10: M3s ‘arankae’ type (MAR 187), occlusal view. 11: m1s ‘arankae’ type (MAR 226), occlusal view.12: m1d ‘arankae’ type (MAR 159), occlusal view. (Enlargement x20)

92

DEINSEA 8, 2001

APPENDIX

Table 1 Desmaninae indet., Marathoussa (MAR), Macedonia, Greece. Dimensions of the upper teeth.(L: length, W: width)

Table 2 Desmaninae indet., Marathoussa (MAR), Macedonia, Greece. Dimensions of the lower teeth.(TRW: trigonid width,TAW: talonid width)

93


APPENDIX (continu e d )

Table 3 C. kornfeldi, Marathoussa (MAR), Macedonia, Greece. Dimensions of the upper teeth. (LT: length of the talon, H: height,PE: posterior emargination, LL: lingual length, BL: buccal length, AW: anterior width, PW: posterior width)

Table 4 C. kornfeldi, Marathoussa (MAR), Macedonia, Greece. Dimensions of the lower teeth.(TRW: trigonid width,TAW: talonid width)

94

DEINSEA 8, 2001


Table 5 Sorex minutus, Marathoussa (MAR), Macedonia, Greece. Dimensions of the upper teeth.

Table 6 Sorex minutus, Marathoussa (MAR), Macedonia, Greece. Dimensions of the lower teeth.

95



Table 7 D. praearaneus, Marathoussa (MAR), Macedonia, Greece. Dimensions of the upper teeth.

Table 8 D. praearaneus, Marathoussa (MAR), Macedonia, Greece. Dimensions of the lower teeth.

96

DEINSEA 8, 2001


Table 9 A. gibberodon, Marathoussa (MAR), Macedonia, Greece. Dimensions of the upper teeth.

Table 10 A. gibberodon, Marathoussa (MAR), Macedonia, Greece. Dimensions of the lower teeth.

97



Table 11 PE-index for the M1s and M2s of morphotype A and B of A. gibberodon from various European localities. 1: Csarnota 2, 2: Osztramos 7, 3: Osztramos 9 (Reumer 1984), 4:Tourkobounia 1 (Reumer & Doukas 1985), 5: Marathoussa, 6: Osztramos 1, 7:Villany 3 (Reumer 1984), 8: Maramena (Doukas et al. 1995).

Table 12 Ranges of the dimensions for the lower molars of A. gibberodon morphotypes from various localities. 9: Maritsa (deBruijn et al. 1970) 10: Apolakkia (Van de Weerd et al.1982). All the other references are as for the Table 11.

98

DEINSEA 8, 2001


Table 13 Beremendia fissidens, Marathoussa (MAR), Macedonia, Greece. Dimensions of the teeth.

Table 14 Spermophilus sp., Marathoussa (MAR), Macedonia, Greece. Dimensions of the teeth.

Table 15 Apodemus sylvaticus/flavicollis, Marathoussa (MAR), Macedonia, Greece. Dimensions of the upper and lower teeth.

99



Table 16 Mean values of M1 and m1 dimensions of various Apodemus species. A.dominans (de Bruijn & Van der Meulen 1975), A.caesareanus (Tchernov 1968), A. levantinus (Tchernov 1968), A. etruscus (Engesser 1989), A. atavus (Fejfar & Storch 1990).

Table 17 . Mean values of M1 and m1 dimensions of A. sylvaticus and A. flavicollis from several localities and different ages:1: Camargue, France (Pasquier 1974), 2: Kebara Cave, Israel (Tchernov 1968), 3: Belvédère 3, the Netherlands (Van Kolfschoten1985), 4: Le Lazaret, France ( Pasquier 1974), 5: Monte Peglia 63, Italy (Van der Meulen 1973), 6: Ravin Voulgarakis, Greece(Koliadimou 1996), 7: Kirchdorf, Austria (Pasquier 1974), 8: Gondrenans, France ( Pasquier 1974), 9: Pirro Nord 1, Italy (De Giuli& Torre 1984), 10:Tourkobounia 2, Greece (Van der Meulen in press), 11: Arnissa, Greece (Mayhew 1977), 12:Varkiza 2, Greece(Van de Weerd 1973).

100

DEINSEA 8, 2001


Table 18 C. cf. koufosi, Marathoussa (MAR), Macedonia, Greece. Dimensions of the upper and lower teeth.

Table 19 Microtus (Allophaiomys) pliocaenicus, Marathoussa (MAR), Macedonia, Greece. Dimensions of the upper and lower teeth.

Ta ble 20 Statistical para m e t e rs for the m1 of M . p l i o c a e n i c u s from Mara t h o u s s a .

101



Ta ble 21 The statistical para m e t e rs of the m1s of the species of A l l o p h a i o m y s from seve ral localities. 1, 3, 4, 8:Van der Meulen( 1 9 7 3 ) ; 2, 5:Van der Meulen & Doukas (in press); 6, 7: Rabeder (1981); 1 0: Koliadimou (1996).

102

DEINSEA 8, 2001


Ta ble 22 Lagurodon ara n k a e, M a rathoussa (MAR), M a c e d o n i a , G r e e c e. Dimensions of the upper and lower teeth.

DEINSEA - ANNUAL OF THE NATURAL HISTORY MUSEUM ROTTERDAMP. O. B o x 2 3 4 5 2 , N L - 3 0 0 1 K L R o t t e r d a m T h e N e t h e r l a n d s

Documents

8.7 Koufos et al.*