Original article

Palaeogeographic significance of the giraffid remains (Mammalia,Arctiodactyla) from Cessaniti (Late Miocene, Southern Italy)§

Importance paléogéographique des girafes de Cessaniti (Miocène supérieur, Italie méridionale)

Antonella Cinzia Marra a,*, Nikos Solounias b,c, Giuseppe Carone d, Lorenzo Rook e

a Dipartimento di Scienze dell’Antichità, Università di Messina, Polo Universitario dell’Annunziata, 98168 Messina, Italyb Department of Anatomy, New York College of Osteopathic Medicine, Old Westbury, New York 11568, USA

c Department of Paleontology, American Museum of Natural History, Central Park West at Seventy-Ninth Street, New York 10024, USAd Gruppo Paleontologico Tropeano, Via Vittorio Veneto 7, 89861 Tropea (VV), Italy

e Dipartimento di Scienze della Terra, Università di Firenze, Via La Pira 4, 50121 Firenze, Italy

Received 10 May 2010; accepted 11 November 2010

Available online 19 February 2011

Abstract

The late Tortonian – early Messinian shallow marine sands of Cessaniti area (Monte Poro, Vibo Valentia, Southern Italy) yield marine andcontinental vertebrates. The best represented taxon is the Sirenian Metaxytherium serresii, while the terrestrial mammal assemblage includes aboselafine bovid, an hexaprotodontid hippopotamus, the giraffids Samotherium cf. boissieri and Bohlinia cf. attica, a rhino and the elephantidStegotetrabelodon syrticus. Until now, the latter was a species with an exclusive Afro-Arabian distribution and the record of Cessaniti is the firstoutside Afro-Arabia. Our attention is here focused on the occurrence of Samotherium cf. boissieri and Bohlinia cf. attica, both being species wellrepresented in the Pikermian Biome. Although evidences of the distribution of the genus Samotherium in Late Miocene African assemblages areweak, it is reported at several sites, while a new species of Bohlinia reported in Chad is still debated. At Cessaniti, the co-occurrence of two giraffidtaxa typical for the Pikermian biome together with a frankly Afro-Arabic species (S. syrticus), further marks the existence of a land connectionbetween the Cessaniti area and North Africa as well as the evidence of a phase of expansion of the Pikermian Biome into the African continent.# 2011 Elsevier Masson SAS. All rights reserved.

Keywords: Samotherium; Bohlinia; Perissodactyla; Proboscidea; Pikermian Biome; Afro-Arabic distribution; Palaeobiogeography

Résumé

Les sables marins peu profonds d’âge Tortonien supérieur – Messinien inférieur de la région de Cessaniti (Monte Poro, Vibo Valentia, Italieméridionale) livrent des vertébrés marins et continentaux. Le taxon le plus abondant est le sirénien Metaxytherium serresii, tandis que l’assemblagede mammifères terrestres comprend un bovidé bosélaphiné, un hippopotame exaprotodontidé, les giraffidés Samotherium cf. boissieri et Bohliniacf. attica, un rhinocérotidé et l’éléphantidé Stegotetrabelodon syrticus. Ce dernier avait jusqu’à présent une répartition exclusivement afro-arabique ; son occurrence à Cessaniti est donc la première hors de cette zone. Nous nous focalisons ici sur la présence de Samotherium cf. boissieriet Bohlinia cf. attica, deux espèces typiques du biome Pikermien. Bien que la présence du genre Samotherium dans des assemblages du Miocènesupérieur africain demande confirmation, elle est notée dans plusieurs sites ; une nouvelle espèce de Bohlinia identifiée au Tchad reste discutée. ÀCessaniti, la présence de deux taxons typiques du biome Pikermien (Samotherium cf. boissieri et Bohlinia cf. attica) associés à une espècefranchement afro-arabique (S. syrticus) confirme l’existence d’une connection continentale entre la région de Cessaniti et l’Afrique du Nord,documentant une phase d’expansion du biome pikermien sur le continent africain.# 2011 Elsevier Masson SAS. Tous droits réservés.

Mots clés : Samotherium ; Bohlinia ; Perissodactyla ; Proboscidea ; Biome Pikermien ; Distribution Afro-Arabique ; Paléobiogéographie

Geobios 44 (2011) 189–197

§ Invited editor: Jordi Agustí.* Corresponding author.

E-mail address: amarra@unime.it (A.C. Marra).

0016-6995/$ – see front matter # 2011 Elsevier Masson SAS. All rights reserveddoi:10.1016/j.geobios.2010.11.005

1. Introduction

The late Miocene site of Cessaniti (Vibo Valentia, SouthernItaly; Fig. 1) yields a fossil mammal assemblage which can

.

[()TD$FIG]

Fig. 1. Geographical location of the sites quoted in the text.

A.C. Marra et al. / Geobios 44 (2011) 189–197190

contribute to the reconstruction of the complex history of thepalaeogeography of the central Mediterranean. In the lateMiocene of Italy, two distinct paleobioprovinces wererecognized, both affected by remarkable changes during theMessinian: the Tusco-Sardinian and the Abruzzi-Apulian(Rook et al., 2006). The two isolated paleobioprovinces werecharacterized by endemic mammals very different from thecoeval faunas of Eurasia and Africa. The recovery of amandible of the Afro-Arabian elephantid Stegotetrabelodonsyrticus at Cessaniti represented the first evidence of a landconnection between the Monte Poro – Capo Vaticano area(Calabria, southern Italy) and North Africa (Ferretti et al.,2003). The other terrestrial mammals recovered in the area ofMonte Poro – Capo Vaticano, mainly at Cessaniti and Zungri(Fig. 1), still under study, are all attributable to not endemicforms and include the giraffids Samotherium cf. boissieri andBohlinia cf. attica, a rhino, a boselaphine bovid, and anhexaprotodont hippopotamid. In addition, the Sirenian Metax-ytherium serresii is well represented in the area (Carone andDomning, 2007).

The occurrence in the assemblage of two giraffids has aparticular relevance in the discussion about the paleogeo-graphic significance of the assemblage. As a matter of fact, thespecies S. boissieri and B. attica have been so far considered

typical elements of the Pikermian biome, a peculiar associationwidespread in the Greco-Iranian bioprovince during theTurolian, and they have never been recorded out of this easterndomain (Kostopoulos, 2009a, 2009b; Fortelius, 2010).Although in North Africa the genera Bohlinia and Samotheriumare scantily represented, if not actually absent, the presence ofthe two species at Cessaniti confirms the existence of a landconnection and the diffusion of elements belonging to thePikermian biome in Africa.

2. The Cessaniti fossiliferous outcrops

The Neogene succession of the Monte Poro – Capo Vaticanoarea, where Cessaniti is located, rests on a Paleozoic crystallinesubstratum. The succession encompasses a transgressive(Tortonian) – regressive (Messinian) cycle. The units recentlydescribed by Gramigna et al. (2008) at Cessaniti – as well asthose previously described by Nicotera (1959) for the northernsector of Monte Poro – crop out over the whole Monte Poro –

Capo Vaticano area, with differences in thickness and lateralvariations. Gramigna et al. (2008) established four units for theoutcrop of Cessaniti (Fig. 2):

� unit 1: dark grey shales alternating with gray coarsesandstones, with abundant Cerithium ssp. and ostreids(marginal lagoon);� unit 2: gray sandstones lying above the previous unit through

a sharp, erosional contact. The sandstones contain a fullymarine fauna, with extremely abundant echinoids, mainlyClypeaster (unit 2 represents the onset of marine conditions;lower shoreface);� unit 3: yellowish poorly cemented sandstones, showing a

depositional style similar to unit 2 (frankly marinedeposition);� unit 4: thin-bedded blue hemipelagic marls, rich in

planktonic foraminifers, known as ‘‘Marne ad Orbulina’’(this latter unit is badly exposed at Cessaniti).

The unit 2, which can reach a thickness of 150 m, yielded themain part of the fossil vertebrates recovered so far in the MontePoro – Capo Vaticano area. Unfortunately, the site of Cessaniti isintensively interested by quarry activities and the recovering offossils, in order to not interfere with the ongoing quarry activities,often lacks the record of adequate stratigraphic and taphonomiccontext. According to the relatively shallow water depositionalenvironment (Gramigna et al., 2008), it is reasonable that more orless disarticulated terrestrial mammal remains should havereached the sea occasionally. Sometimes, abrasions and bio-encrustations occur on the bones, indicating they have beentransported and laid on the sea-bottom for a certain time beforeburial. Bones often occur isolated, and only in one case a manusof giraffid has been recovered almost complete, its bones beingassociated and slightly dispersed (Marra and Carone, 2008). Theoccurrence of Heterostegina papyracea in the level capping theunit 2 is an ante quam term, this species being present, althoughrare, in the Mediterranean until the late Tortonian (Adams, 1976;Papazzoni and Sirotti, 1999).

[()TD$FIG]

Fig. 2. Stratigraphical section of Cessaniti (modified after Gramigna et al.,2008).

A.C. Marra et al. / Geobios 44 (2011) 189–197 191

3. The Cessaniti terrestrial mammals

The Cessaniti deposits yielded mammalian faunas that aredated to the pre-Messinian latest Miocene. Despite scanty, theidentified mammalian taxa illustrate the Late Mioceneinhabitants of the emerged lands in this portion of the centralMediterranean region. The most representative mammal atCessaniti is the Sirenian Metaxytherium serresii, exhaustivelydescribed by Carone and Domning (2007). The terrestrial

mammal assemblage includes the elephantid Stegotetrabelodonsyrticus, the giraffids Samotherium cf. boissieri and Bohlinia cf.attica, a boselafine bovid, an hexaprotodont hippopotamid, anda rhino. The only continental vertebrate from Cessaniti so fardescribed in details is the elephantid S. syrticus, a species withan exclusive Afro-Arabian distribution (Ferretti et al., 2003;Rook et al., 2006).

3.1. Proboscideans

Gomphothere remains are quite common elements inEuropean Late Miocene deposits. Instead most surprisingwas the discovery, in the late 1990s, of an early elephantid withAfrican affinities from the Cessaniti succession. The first recordof a proboscidean from Cessaniti is the mandible (andassociated tusk fragment) recorded at ‘‘Brunia’’ quarry andattributed to S. syrticus (Ferretti et al., 2003). The elongated anddown-turned symphysis of the mandible, the number andstructure of the molar plates, and the oval cross section andlamellar inner structure of the tusk all agree with the charactersof the genus Stegotetrabelodon and can be attributed to thespecies S. syrticus (Ferretti et al., 2003). In recent years, furtherproboscidean specimens have been recovered (Marra andCarone, 2008). They consist of a molar fragment, a metapodialand a partial femur (recovered at ‘‘Gentile’’ quarry) alsoattributable to S. syrticus.

The specimens from Cessaniti have some plesiomorphiccharacters with respect to the sample referred to S. syrticusfrom As Sahabi, a late Miocene locality from Northern Libya,which released the type material (Ferretti et al., 2003). Atpresent, the finding of Cessaniti is the only record of this speciesout of the Afro-Arabian province and is interpreted as anevidence of a land connection of the Calabrian Arch to NorthAfrica (Ferretti et al., 2003; Ferretti, 2008).

3.2. Artiodactyls

Artiodactyls usually are among the most abundant largemammal elements in the Old World Late Miocene sites.Several fragmentary remains attest the occurrence ofartiodactyls in the Cessaniti assemblage. While bovidae arescantily represented (a single mandible and a few postcranialfragments, identified as a Boselaphine, although pending adeeper study; postcranial elements of a small sized bovid),and an hexaprotodontid hippopotamus is documented by afew teeth fragments, the best-represented Artiodactyls aremid-sized giraffids, known by a number of cranial andpostcranial elements recovered at Gentile’s quarry (Cessaniti,VV), and at Malopara, a locality near Zungri (Fig. 1). Theexact stratigraphical position is not available in the recordfrom Cessaniti, although all fossils come from the unit 2,which is intensively excavated in the Gentile’s quarry, whilethe fossils from Zungri derive from a level of sandstonestentatively correlated to the unit 2 (Marra and Carone, 2008;Marra et al., 2009).

A left upper tooth row (P3-M3, Fig. 3(1)), a distal radius,two astragalii (Fig. 3(2)), two cubo-naviculars, and a distal tibia

[()TD$FIG]

Fig. 3. Bohlinia cf. attica from Cessaniti. 1: Left upper tooth row (P4-M3; P3 is under restoration) in labial (a), occlusal (b), and lingual (c) views; 2: Left astragalusin dorsal (a), lateral (b), plantar (c), and medial (d) views. Scale bar: 5 cm for 1; 10 cm for 2.

A.C. Marra et al. / Geobios 44 (2011) 189–197192

recovered at Cessaniti fall within the variability of Bohliniaattica.

The astragalii show the most significant features of Bohlinia:the general aspect is slender; in dorsal view the neck is wide andthe collum talii is massive; in plantar view the articular facet isuniformly convex and the posterior notch of the lateral throcleais elevated. The cubonaviculars have metatarsal facetsseparated by the sulcus for the long-peronian muscle, as inBohlinia (Geraads, 1994). Molars and premolars are brachyo-dont and large. The mesostyle of the molars is strong. P4 is asbroad as M1. The cingulum is evident on the labial side, andweak on the lingual side of the upper teeth. The teeth ofBohlinia are poorly known (Geraads et al., 2005) and they are

very similar in dimensions and morphology to Palaeotragusones (i.e., P. coelophrys, Gentry, 2003), except for thepremolars P2 and P3 which are small and reduced inPalaeotragus, contrary to Bohlinia. In Bohlinia, upperpremolars have a large and bulky parastyle and upper molarshave a strong metacone rib (Gentry et al., 1999); same featuresare observable in the specimen form Cessaniti. In absence ofother skeletal elements attributable to Palaeotragus, and inconsiderations of the morphometric features, the upper toothrow from Cessaniti can be attributed to Bohlinia.

Remains attributable to Samotherium cf. boissieri arerecorded both at Cessaniti and at Zungri. Two metacarpals,two metatarsals, and a distal tibia come from Cessaniti (Fig. 4),

[()TD$FIG]

Fig. 4. Samotherium cf. boissieri from the Cessaniti-Zungri area. 1: Left metacarpal in dorsal (a), ventral (b) and proximal (c) views; 2: Right metatarsal in dorsal (a)and ventral (b) views; 3: Left metatarsal in proximal view. Scale bar: 10 cm.

A.C. Marra et al. / Geobios 44 (2011) 189–197 193

while at Zungri this species is represented by a complete leftmanus (including scaphoid, semilunar, pyramidal, magnum,uncinatum, metacarpal, great sesamoids, first, second and thirdphalanges), a partially preserved left ulna and a distal epiphysisof a left radius probably belonging to the same individual, astrongly damaged humerus fragments, and some bonesprobably belonging to the right fore-limb (semilunar, pyr-amidal, unicinatum, pisiform, metacarpal, and a distal rightradius) of the same individual. The bones of the left fore-limb

were not recovered in anatomical connection, but in closespatial association, while those of the right one were moredispersed. At Zungri, some bones were exposed and partiallyweathered. Some fossils, in particular the distal ends ofmetacarpals, have encrustations by marine organisms. As inS. boissieri (Solounias, 2007; Kostopoulos, 2009a), themetacarpals (Fig. 4(1)) have a semicircular proximal articularsurface, the crests on the posterior rims of the diaphysis are notelevated, and the sulcus is shallow. Their robusticity index

A.C. Marra et al. / Geobios 44 (2011) 189–197194

(transverse diameter of the shaft/length) is 11.7 and 11.8, withinthe values known for S. boissieri from Samos (from 10.4 to13.6, mean: 12.45, 12 specimens; Kostopoulos, 2009a). Thecarpal bones of the species are still not known in literature(Kostopoulos, 2009a), and the record of Zungri is particularlyvaluable in this respect. The metatarsals (Fig. 4(2, 3)) have aweak plantar crest in the shaft. The robusticity index (TD shaft/length) for the two specimens from Cessaniti is 10.6 and 10.4,and fall within the variability of S. boissieri from Samos, whereit ranges from 9.5 to 12.5 (mean = 10.7, 20 specimens,Kostopoulos, 2009a). Morphometric data of the specimensfrom Cessaniti and Zungri are also comparable to Samotheriumsp. from Kemiklitepe (Turkey; Geraads, 1994), whosemeasurements fall within the variability of S. boissieri fromSamos (Kostopoulos, 2009a).

Before their finding at Cessaniti, B. attica and S. boissierihave been considered typical species of the Pikermian biome, apeculiar mammal community spread in the Greco-Iranianbiogeographic province during the Late Miocene, and foundassociated only at Samos. Moreover, the occurrence of Bohliniaand Samotherium in Africa is poorly documented and stilldebated (cf. Harris et al., 2010).

3.3. Perissodactyls

A Rhinocerotid fragmentary skull, two fragmentary teeth, andfew postcranial bones (femur diaphisis, distal humerus, and ametapodial) show the presence of one species of Diceros (bestcomparable to Diceros primaevus or ‘‘Diceros’’ neumayri). Thespecies D. primaevus is based on the rhinoceros material from theearly Late Miocene locality of Bou Hanifia (Oued el Hammam;Geraads, 1986), and preserves an ancestral morphologypersisting, almost unmodified, in the extant black rhinoceros(Giaourtsakis et al., 2009). According to Giaourtsakis et al.(2009), populations similar to D. primaevus could have migratedoutside Africa, around the Middle-Late Miocene boundary, andevolved to ‘‘D.’’ neumayri, a Late Miocene species typical of theGreco-Iranian bioprovince. A second upper premolar of ‘‘D.’’neumayri found at As Sahabi would represent the first occurrenceof the species in Africa (Bernor et al., 1987).

4. Palaeogeographic significance of the Cessanitigiraffes

Faunal composition of fossil assemblages is a precious toolfor detecting similarities and differences among comparedfaunas and thus for paleogeographic reconstructions. Differentapproaches have been developed since the pioneering work bySimpson (1945, 1960), and most workers dealing with fossilfaunas still appreciate Simpson’s Faunal Resemblance Index(FRI) because of its greater consistency when dealing withfossil datasets (e.g., Bernor and Pavlakis, 1987; Bernor et al.,2001, 2008). Despite the fauna currently recovered fromCessaniti is too small for any attempt to perform a GFRI indexcomparison, the paleogeographic significance of single taxacan provide interesting paleogeographic evidences. Theimportance of the species S. syrticus has been already discussed

in Ferretti et al. (2003) and Ferretti (2008). We herein focus onthe palaeogeographic significance of B. cf. attica and S. cf.boissieri in the Monte Poro – Capo Vaticano area. The record ofCessaniti-Zungri represents the second co-occurrence of thesetwo species of the Pikermian biome; the other one is Samos.

The genus Bohlinia includes only one species, B. attica,recorded from the Late Vallesian to the Late Turolian of theGreco-Iranian bioprovince (Geraads et al., 2005). Inside thisbiogeographic province, B. attica is known in sites located in thewestern part, such as Greece and Bulgaria (see range chart byKostopoulos, 2009b; Fortelius, 2010). Despite the occurrence ofBohlinia in the Samos faunal list is not recorded in the NOWdatabase (Fortelius, 2010) and is reported as a dubitablepresence, probably restricted to the lower fossil-levels, byKostopoulos (2009a), a Bohlinia palate from Samos is present inthe collections of the Carnegie Museum in Pittsburgh (CM370, aspecimen sent around 1900 by M. Schlosser as part of anexchange from Munich; NS personal observation). The presenceof Bohlinia out of the Eastern domain was not attested so far, withthe only exception of a new species referred to Bohlinia in theLate Miocene of Chad (Likius et al., 2007), whose occurrence is,for some authors, disputable (Harris et al., 2010).

S. boissieri is an early Turolian species. It is extremely wellrepresented in the western part of the Greco-Iranian bio-province, such as Samos in Greece and Turkey (Kostopoulos,2009b). The species firstly occurs at Gülpinar (Turkey, MN10)and is relatively widely distributed in the MN11 and MN12biozone (Fortelius, 2010). Although the intermediated stagesare still unknown, the fossil record of Samos indicates that S.boissieri is the ancestor of the larger species S. major(Kostopoulos, 2009a). At Samos, S. major replaces S. boissieriat the MN11/MN12 transition (Kostopoulos et al., 2003;Kostopoulos, 2009a); at Kemiklitepe the species replacement isalso identified approximately at the same moment (Geraads,1994; Sen et al., 1994).

The Pikermian Biome is intended as a peculiar mammalcommunity of savannah-like type gradually emerged in the C3dominated and progressively open habitats of the Greco-Iranian bioprovince; this biome was widespread during theLate Miocene from Greece through Bulgaria, Ukraine,Turkey, Southern Russia, Iran, and eastward all the way toChina. Elements of it have also been found in southern France,Spain, and North Africa (Solounias et al., 1999; Kostopoulos,2009b). Solounias et al. (2010) suggest a new interpretation ofthe Pikermian biome as woodland with limited grassyhabitats. According to these authors, the feeding styles ofthe herbivores from Samos and Pikermi are adapted tosignificant browsing and mixed feeding. It is relevant to pointout that two classical sites representative of the Pikermianbiome, Samos and Pikermi, show ecological and/or chron-ological differences (Kostopoulos, 2009b; Solounias et al.,2010). At Pikermi, Samotherium is absent, probably forecological or chronological reasons, while B. attica isrecorded (Kostopoulos, 2009b; Solounias et al., 2010). S.boissieri was probably a grazer, and grazer herbivores werescarce at Pikermi, suggesting a woodland habitat more closedthan at Samos (Solounias et al., 2010).

A.C. Marra et al. / Geobios 44 (2011) 189–197 195

Also the physiography played a role. According toKostopoulos (2009b), the western part of the Greco-Iranianbioprovince (Southern Balkans, including Pikermi, Greece andBulgaria) has been a distinct biogeographic region, and onlyabout 7.3–7.1 My it has been connected to the sub-Paratethyanbioprovince, allowing the emergence and expansion of thePikermian Biome, which drastically declined after 7.0 My.According to Kostopoulos (2009b), the Pikermian fauna wasnot uniform and was controlled by climatic and physiographicfactors. S. boissieri never occurs in the western part of theAegean region, probably geographically isolated during theearly Turolian (Kostopoulos, 2009a, 2009b) and its co-occurrence with Bohlinia at Cessaniti – Zungri, is particularlyinteresting.

The palaeoecology of Southern Italy seems to be similar toSamos, where both Samotherium and Bohlina are present.Solounias et al. (2010) suggested that Samos and Pikermi havehad very similar habitats, with predominantly browsers andmixed feeders with relatively few grazers. However, Samosseems to have had a more open habitat as there are more grazinghippotheria found there (Solounias et al., 2010). The co-occurrence of the two giraffids at Cessaniti points towardspaleoecological conditions more similar to Samos. Thespreading of the two species out of the Greco-Iranianbioprovince can be related to the maximum expansion of thePikermian biome. However, the record in Africa is very poorand uncertain.

The occurrence of Samotherium in Late Miocene Africanassemblages is attested in several, although scant Late to latestMiocene sites. In North Africa, few remains attributed toSamotherium sp. have been found at Oued el Hammam(Algeria, Arambourg, 1959; Geraads, 1989), Bled el Douhara(Tunisia, Arambourg, 1959; Geraads, 1989), As Sahabi (Libya;Harris, 1987), and Wadi Natrun (Egypt; Jouleoud, 1937). Thefossils attributed by Churcher (1970) to an African species,Samotherium africanum, have been later placed to the mostcommon giraffid of the site, Giraffokeryx primaevus (Harriset al., 2010). In the opinion of Harris et al. (2010) the fossilsattributed to Samotherium in Africa need revaluation.

Although still debated (Harris et al., 2010), the first Africanoccurrence of Bohlinia in Chad could partially fill ageographical gap in the distribution of Giraffidae in centralAfrica. The new species B. adoumi comes from differentlocalities of the Toros-Menalla fossiliferous sector (northernChad), dated about 7.0 My (Vignaud et al., 2002; Likius et al.,2007). If confirmed, the presence of Bohlinia in Chad wouldrepresent an evidence of the paleogeographical relationshipbetween central Africa and Eurasia during the Late Miocene(Likius et al., 2007). The site of As Sahabi, and the coeval andgeographically close fossiliferous area of Toros Menalla inChad, is particularly representative of the faunal exchangesbetween Africa and Eurasia. Faunal analyses of the largemammal assemblage allowed considering As Sahabi as a LateMiocene biogeographic crossroads fauna between westernEurasia and sub-Saharan Africa Late Miocene faunas (Bernorand Pavlakis, 1987; Bernor and Rook, 2008). This site, datedabout 6.7 My, released a mammal fauna which reveals

consistent biogeographic connections between contempora-neous Arabian and Kenyan localities and older Late MiocenePikermian localities, about 8.2–7.5 My.

Finally, the record of Cessaniti appears comparable to thenorth African site, S. syrticus being undoubtedly present at AsSahabi while the occurrence of Samotherium sp. is morespeculative. Lastly, the occurrence of Bohlinia at Cessanitiwould make the identification of this genus in Chad morerobust.

5. Concluding remarks

The existence of a land connection between the Monte Poro– Capo Vaticano area and North Africa has been firstlysuggested by the occurrence of S. syrticus at Cessaniti (Ferrettiet al., 2003). The first faunal list provided by Ferretti et al.(2003) indicated the not-endemic character of the fauna,excluding an isolation of the area in the Late Miocene. Thepresence of the two giraffids Bohlinia cf. attica andSamotherium cf. boissieri at Cessaniti-Zungri represents thefirst record of these species out of the Greco-Iranianbioprovince and, in our view, is coherent with the land-connection of the Monte Poro – Capo Vaticano area with NorthAfrica. Moreover, it would confirm the presence of Bohliniaand Samotherium in North Africa, where their findings arescanty and dubious. The rarity of Samotherium and the possibleabsence of Bohlinia ( fide Harris et al., 2010) in North Africacould be due to a short persistence or to ecological reasons, sotheir presence at Cessaniti can be interpreted also in thisrespect.

The spreading of Bohlinia and Samotherium out of theeastern domain is one of the evidences of the expansion inAfrica of the so-called ‘‘Pikermian biome’’, whose paleogeo-graphical range and paleoecological character are going to berevised (Kostopoulos, 2009b; Solounias et al., 2010). Theassociation of Bohlinia and Samotherium in the same site is apeculiar character of the area of Cessaniti – Zungri, whichassimilate it to Samos (Greece). The record of Cessaniti-Zungrican represent one of the latest occurrences of the species S.boissieri, which in the Greco-Iranian bioprovince is replaced bythe more derived species Samotherium major at the MN11/MN12 transition (Geraads, 1994; Kostopoulos, 2009a).

The vicinity in space and comparable chronology betweenCessaniti and As Sahabi, give to Cessaniti an intriguingpotential. Like As Sahabi, positioned as a Late Miocenebiogeographic crossroad, the fossiliferous succession atCessaniti could provide new evidence for the faunal evolutionof this portion of the circum Mediterranean area in very lateMiocene times, a crucial time period in the evolution of Africanvertebrate assemblages. The record of Cessaniti-Zungri inviteto reconsider times and ways of expansion of the Pikermianbiome, in particular its diffusion in North Africa.

Acknowledgments

We thank the Superintendent Dr Simonetta Bonomi and theOfficial Dr Maria Teresa Iannelli (Soprintendenza ai Beni

A.C. Marra et al. / Geobios 44 (2011) 189–197196

Archeologici della Calabria) for allowing the study of thefossils under their responsibility. Background work for thispaper has been supported by the University of Messina (PRA2006-2007, to ACM) and NGS (8788-10, to LR).

References

Adams, C.G., 1976. Larger Foraminifera and the late Cenozoic history of theMediterranean region. Palaeogeography, Palaeoclimatology, Palaeoecology20, 47–66.

Arambourg, C., 1959. Vertébrés continentaux du Miocène supérieur de l’Afri-que du Nord. Publications du Service de la Carte Géologique d’Algérie,Nouvelle Serie, Paléontologie, Mémoire 1, 1–159.

Bernor, R.L., Fortelius, M., Rook, L., 2001. Evolutionary biogeography andpaleoecology of the Oreopithecus bambolii Faunal Zone (Late Miocene,Tusco-Sardinian Province). Bolletino della Società Paleontologica Italiana40 (2), 139–148.

Bernor, R.L., Heissig, K., Tobien, H., 1987. Early Pliocene Perissodactylafrom Sahabi, Libya. In: Boaz, N.T., El-Arnauti, A., Gaziry, A., deHeinzelin, J., Boaz, D.D. (Eds.), Neogene Paleontology and Geologyof Sahabi. Alan R. Liss, New York, pp. 233–254.

Bernor, R.L., Pavlakis, P., 1987. Zoogeographic relationships of the Sahabilarge mammal fauna (Early Pliocene, Libya). In: Boaz, N.T., El-Arnauti,A., Gaziry, A.W., de Heinzelin, J., Boaz, D.D. (Eds.), Neogene Paleon-tology and Geology of Sahabi. Alan R. Liss, New York, pp. 337–

348.Bernor, R.L., Rook, L., 2008. A current view of Sahabi large mammal

biogeographic relationships. In: Boaz, N.T., El-Arnauti, A., Pavlakis,P., Salem, M. (Eds.), Circum-Mediterranean Neogene Geology and BioticEvolution: the Perspective from Libya. Garyounis Scientific Bulletin.Special Issue 5. pp. 285–292.

Bernor, R.L., Rook, L., Haile-Selassie, Y., 2008. Paleobiogeography. In: Haile-Selassie, Y., Woldegabriel, G. (Eds.), Ardipithecus kadabba, Late MioceneEvidence from the Middle Awash, Ethiopia, 2. University of CaliforniaPress, Berkeley, pp. 549–564.

Carone, G., Domning, D., 2007. Metaxytherium serresii (Mammalia: Sirenia):new pre-Pliocene records, and implications for Mediterranean paleoecologybefore and after the Messinian Salinity Crisis. Bollettino della SocietàPaleontologica Italiana 46, 55–92.

Churcher, C.S., 1970. Two new upper Miocene giraffids from Fort Ternan,Kenya, East Africa: Palaeotragus primaevus n. sp. and Samotheriumafricanum n. sp. In: Leakey, L.S.B., Savage, R.J.G. (Eds.), Fossilvertebrates of Africa 2. Academic Press, London & New York, pp. 1–

109.Ferretti, M.P., 2008. Miocene proboscideans from Italy. African elements and

paleobiogeographic implications. In: Salem, M.J., El-Arnauti, A., El SogherSaleh, A. (Eds.), The Geology of East Libya 3, Third Symposium on theSedimentary Basins of Lybia. Binghazi, November 11-13 2004. EarthScience Society of Lybia, Tripoli, pp. 325–334.

Ferretti, M.P., Rook, L., Torre, D., 2003. Stegotetrabelodon cf. syrticus(Proboscidea, Elephantidae) from the Upper Miocene of Cessaniti (Calab-ria, southern Italy) and its bearing on Late Miocene paleogeographyof central Mediterranean. Journal of Vertebrate Paleontology 23,659–666.

Fortelius, M, 2010. Neogene of the Old World Database of Fossil Mammals(NOW). University of Helsinki. http://www.helsinki.fi/science/now/

Gentry, A.W., 2003. Ruminantia (Artiodactyla). In: Fortelius, M., Kappleman,J., Sen, S., Bernor, R.L. (Eds.), Geology and Paleontology of the MioceneSinap Formation, Turkey. Columbia University Press, New York, pp.332–379.

Gentry, A.W., Rössner, G.E., Heizmann, E.P.J., 1999. Ruminantia. In: Rössner,G.E., Heissig, K. (Eds.), The Miocene Land Mammals of Europe. VerlagFriedrich Pfeil, Munich, pp. 225–258.

Giaourtsakis, I.X., Pehlevan, C., Haile-Selassie, Y., 2009. Rhinocerotidae. In:Haile-Selassie, Y., Woldegabriel, G. (Eds.), Ardipithecus kadabba, LateMiocene Evidence from the Middle Awash, Ethiopia, 2. University ofCalifornia Press, Berkeley, pp. 429–468.

Geraads, D., 1986. Sur les relations phylétiques de Dicerorhinus primaevusArambourg, 1959, rhinocéros du Vallésien d’Algérie. Comptes Rendu del’Académie des Sciences de Paris (2) 302, 835–837.

Geraads, D., 1989. Vertébrés fossiles du Miocène supérieur du Djebel Krechemel Artsouma (Tunisie Centrale). Comparaisons biostratigraphiques. Geobios22, 777–801.

Geraads, D., 1994. Les gisements de mammifères du Miocène supérieur deKemiklitepe, Turquie : 8. Giraffidae. Bulletin du Muséum Nationaled’Histoire Naturelle 16, 169–173.

Geraads, D., Spassov, N., Kovachev, D., 2005. Giraffidae (Artiodactyla, mam-malian) from the Late Miocene of Kalimantsi and Hadjidimovo, South-western Bulgaria. Geologica Balcanica 35, 11–18.

Gramigna, P., Guido, A., Mastandrea, A., Russo, F., 2008. The paleontologi-cal site of Cessaniti: a window on a coastal marine environment of sevenmillion years ago (Southern Calabria, Italy). Geologica Romana 41,25–34.

Harris, J.M., 1987. Fossil Giraffidae from Sahabi, Libya. In: Boaz, N.T., El-Arnauti, A., Gaziry, A.W., de Heinzelin, J., Boaz, D.D. (Eds.), NeogenePaleontology and Geology of Sahabi. Alan R. Liss, New York, pp. 317–322.

Harris, J.M., Solounias, N., Geraads, D., 2010. Giraffoidea. In: Werdelin, L.,Sanders, W.J. (Eds.), Cenozoic Mammals of Africa. University of CaliforniaPress, Berkeley, pp. 805–819.

Jouleoud, L., 1937. Remarques sur les giraffidés fossiles d’Afrique. Mammalia1, 85–96.

Kostopoulos, D., 2009a. 13. Giraffidae. In: Koufos, G.D., Nagel, D. (Eds.), TheLate Miocene Mammal Faunas of the Mytilinii Basin, Samos Island. NewCollections, Beiträge zur Paläontologie, Greece, 31, pp. 299–343.

Kostopoulos, D.S., 2009b. The Pikermian event: temporal and spatial resolutionof the Turolian large mammal fauna in SE Europe. Palaeogeography,Palaeoclimatology, Palaeoecology 274, 82–95.

Kostopoulos, D.S., Sen, S., Koufos, G.D., 2003. Magnetostratigraphy andrevised chronology of the Late Miocene mammal localities of Samos,Greece. International Journal of Earth Sciences 92, 779–794.

Likius, A., Vignaud, P., Brunet, M., 2007. Une nouvelle espèce du genreBohlinia (Mammalia, Giraffidae) du Miocène supérieur de Toros-Menalla,Tchad. Comptes Rendus Palevol 6, 211–220.

Marra, A.C., Carone, G., 2008. Complete manus of a medium-sized giraffidfrom the Late Miocene of Zungri (Vibo Valentia, Italy). Giornate diPaleontologia 85–86 Abstract book.

Marra, A.C., Rook, L., Carone, G., Gramigna, P., 2009. Taxonomic assessmentand palaeogeographic significance of the giraffid remains from Cessaniti(Late Miocene, Southern Italy). 13th Congress RCMNS (Regional Com-mittee on Mediterranean Neogene Stratigraphy). Acta Naturalia de ‘‘L’Ate-neo Parmense’’ 45 (1/4), 299–301.

Nicotera, P., 1959. Rilevamento geologico del versante settentrionale del montePoro (Calabria). Memorie e Note dell’Istituto di Geologia Applicata diNapoli 7, 1–92.

Papazzoni, C.A., Sirotti, A., 1999. Heterostegina papyracea Seguenza, 1880from the upper Miocene of Cessaniti (Vibo Valentia, southern Italy).Bollettino della Società Paleontologica Italiana 38, 15–21.

Rook, L., Gallai, G., Torre, D., 2006. Lands and endemic mammals in the LateMiocene of Italy: constrains for paleogeographic outlines of Tyrrhenianarea. Palaeogeography, Palaeoclimatology, Palaeoecology 238, 263–

269.Sen, S., de Bonis, L., Dalfes, N., Geraads, D., Koufos, G.D., 1994. Les

gisements de mammifères du Miocène supérieur de Kemiklitepe,Turquie : 1. Stratigraphie et magnétostratigraphie. Bulletin du MuséumNational d’Histoire Naturelle 16 (C (1)), 5–17.

Simpson, G.G., 1945. The principles of classification and a classification ofthe mammals. Bulletin of the American Museum of Natural History 85,1–350.

Simpson, G.G., 1960. Notes on the measurement of faunal resemblance.American Journal of Science 258-A, 300–311.

Solounias, N., 2007. Family Giraffidae. In: Prothero, D.R.,Foss, S.E. (Eds.), TheEvolution of Artiodactyls. The Johns Hopkins University Press, Baltimore,pp. 257–277.

Solounias, N., Plavcan, J.M., Quade, J., Witmer, L., 1999. The paleoecology ofthe Pikermian biome and the savanna myth. In: Agustì, J., Rook, L., Andrews,

A.C. Marra et al. / Geobios 44 (2011) 189–197 197

P. (Eds.), Hominoid Evolution and Climatic Change in Europe, The Evolutionof Neogene Terrestrial Ecosystems in Europe, 1. Cambridge University Press,Cambridge, pp. 436–453.

Solounias, N., Rivals, F., Semprebon, G.M., 2010. Dietary interpretation andpaleoecology of herbivores from Pikermi and Samos (Late Miocene ofGreece). Palaeobiology 36, 113–136.

Vignaud, P., Duringer, P., MacKaye, H.T., Likius, A., Blondel, C., Boisserie, J.R.,de Bonis, L., Eisenmann, V., Etienne, M.E., Geraads, D., Guy, F., Lehmann,T., Lihoreau, F., Lopez-Martinez, N., Mourer-Chauviré, C., Otero, O., Rage,J.C., Schuster, M., Viriot, L., Zazzo, A., Brunet, M., 2002. Geology andPalaeontology of the Upper Miocene Toros-Menalla ‘hominid locality,Djurab Desert, northern Chad. Nature 418, 152–155.