Intercontinental relationship Europe - Africa and the Indian Subcontinent 45 Jan van der Made* A great number of Miocene genera, and even Palaeogeography, global climate some species, are cited or described from both Europe and Africa and/or the Indian Subconti- nent. In other cases, an ancestor-descendant re- After MN 3, Europe formed one continent with lationship has been demonstrated. For most of Asia. This land mass extended from Europe, the Miocene, there seem to have been intensive through north Asia to China and SE Asia and is faunal relationships between Europe, Africa and here referred to as Eurasia. This term does not the Indian Subcontinent. This situation may seem include here SE Europe. At this time, the Brea normal to uso It is, however, noto north of Crete was land and SE Europe and During much of the Tertiary, Africa and India Anatolia formed a continuous landmass. The Para- were isolated continents. There were some peri- tethys was large and extended from the valley of ods when faunal exchange with the northern the Rhone to the Black Sea, Caspian Sea and continents occurred, but these periods seem to further to the east. The Tethys was connected have been widely spaced in time. During a larga with the Indian Ocean and large part of the Middle part of the Oligocene and during the earliest East was a shallow sea. During the earliest Mio- Miocene, Africa and India had been isolated. En- cene, Africa and Arabia formed one continent that demic faunas evolved on these continents. Fam- had been separated from Eurasia and India for a ilies that went extinct in the northern continents considerable time. Similarly the India n Continent during the Paleogene, tlorished in these southern had been isolated for a very long period (RÓGL & continents and other familias originated. Also in STEINIGER1983). Due to this long isolation, the the northern continents evolution went on. The southern continents had endemic faunas. result was that the northern and southern conti- During the Miocene, the African and Indian nents had totally different faunas. plates moved closer to Eurasia causing the fur- Then happened what MADDEN & VAN COUVER- ther elevation of the Alps, Himalayas and other ING (1976) named the "Proboscidean Datum mountain belts. The Paratethys and Tethys be- Eventn. They apparently believed that it was a carne smaller and extensive Breas in the Middle single synchronous faunal exchange event be- East beca me shallow seas and finally beca me tween Africa and Eurasia, involving many taxa, land. The area of SE Europe beca me connected including proboscideans. THOMAS (1985) recog- to Eurasia, while the collapse of the ares north of nized two majar phases of faunal exchange. T ASSY Crete separated it from Anatolia. The Red Sea (1990) showed that the dispersal of the probosci- formed, and Arabia became disconnected from deans was diachronous. The UDatum Eventn turns Africa. During much of the Miocene, India may out to be a complex of events (VAN DERMADE have had a contact with Asia at the place of the 1996, 1997b). These events took place during the present Himalayas, but there is little evidence of Early and Middle Miocene. Faunal exchange be- faunal exchange through this ares. The Parateth- tween Eurasia and Africa during the Late Miocene ys was another barrier. Faunal exchange between is not well documented; the African record is very northern Asia and the southern continents may incomplete. The Indian record is better for the have occurred through the ares between Himala- Late Miocene, but nevertheless, relationships be- yas and Paratethys. tween the faunas are not very clear. The events While plate tectonics gradually changed the around the latest Miocene Messinian Salinity Cri- geography, sea level fluctuations had more direct sis are again well studied. effects. Sea level changes were, geologically J speaking, instantaneous, and had a magnitude of up to a hundred metres and occasionally succes- sive events had accumulative effects (HAQ et al. 1987). Between 15 and 10.5 Ma ago, sea level dropped in three successive events ayer two * Dr. Jan van der Made, Museo Nacional de Ciencias hundred metres (HAQ et al. 1987). Sea level chang- Naturales, José Gutiérrez Abascal, 2, E-2BOO6Ma- e8 were relaled 10 global climale; Ice accumula- drid, Spain tion on Antartica was responsible for sea level The Mk)cen8 LAnd Memmals ~ E¡wope. - Pp. 4&7-472. 457 drops (MILLER et al. 1996). While extensive land Stratigraphy masses were slowly being formed out of a number of smaller fragments, sea level changes connect- MN units or Neogene Mammal Units are widely ed and disconnected these Breas and allowed for, usad for Europe and northern Africa (DE BRUIJNet or inhibitted faunal exchange. The effects of the al. 1992). The MN units do not have a clear defi- sea level changes were strongest during the Early nition of the boundaries. It appears that global Miocene, whereas from the Middle Miocene on- events, causing majar dispersal or extinction ward, the land-sea distribution seems to have events, tend to be within MN units, whereas the been less affected and less important in control- transition of one unit to the other is diffuse. Here ling the geographical distribution of mammals. frequently informal reference will be made to Changes in global climate had a direct impact MN 4e (early MN 4), MN 41 (late MN 4) etc. to on the distribution of species: there are distinct indicate the part of the MN unit before or after an dispersal and extinction events. When, during the important dispersal evento In the same way MN 8 Early Miocene, global climate lowered the sea may be used instead of MN 7+8. In addition level, and allowed for faunal exchange between reference is made to continental stages and their Africa and Europe, global climate siso permitted biozones: Ramblian (zones Z-A), Aragonian (zones Asian animals to disperse into Europe. Late Mio- B-G), Vallesian (H-I) and Turolian (J-M) (DAAMS& cene global climate, may have had effects on sea FREUDENTHAl1988, VAN DAM 1997). Though these level, but its effect on the distribution of animals zones are defined with local (Spanish) fauna, the was directo fact that the local faunal events used tend to Faunal exchange between Europe and Africa coincide with global events, makes a more or less and the Indian Subcontinent occurred through SE precise correlation with a much larger Brea possi- Europe and Anatolia. Miocene climates do not ble. A number of long profiles, frequently in or necessarily have recent equivalents. Neverthe- correlated to type sections, were sampled for less, certain characters may have been the same: palaeomagnetism in arder to date biozones and higher temperatures towards the equator, the MN units (KRIJGSMANet al. 1996). These dates are climate of northern Eurasia may always have used here, but differ considerably in the MN 4-6 been more seasonal with respect to temperature range as used by for instance STEININGERet al. than the central African climate, etc. As we will (1996). see in the following sections, the distribution of For Africa a series of "faunal sets" were de- animals that may have been dependent on Borne fined (PICKFORD1981). Like the Spanish zones, degree of humidity suggests, that the distribution fuese sets tend to capture global events. Many of more and less humid climates during a large African localities are radiometrically datad (PICK- part of the Miocene may have shown only a FORD 1986b). limited variation. In general, the conditions in SE The best record for the Indian Subcontinent is Europe and Anatolia, may have acted as a filler to in Pakistan. A series of "faunas" were described faunal exchange. There is a well studied palien from Pakistan, later the names of these "faunas" record in Anatolia, that is correlated to the were applied to the formations from which they MN scale (BENDA& MEULENKAMP1990); there were were collected and still later, the names have important changes in vegetation. Both reasons been applied as stages: Bugti, Murree, Kamlial, may explain why certain laxa, did not take part in Chinji, Nagri, Dhok Pathan (SHAH1984), most of one faunal exchange event, but were very prom- which have been palaeomagnetically sampled inent in a later evento and which yielded detailed range charts (BARRY& A peculiar event took place at the end of the FLYNN1990). Collections of older fossils from the Miocene. The connection between the Mediterra- Bugti Brea, include fossils from several levels nean and the Atlantic Ocean was lost. Evapora- (PICKFORD1987), but the bulk of the material is tion in the Mediterranean is greater than the input probably from more or less the same age anterior through precipitation and rivers. The sea level fell Murree. by up to several kilometres, and Europe and Asia A framework of faunal correlations and the beca me connected, allowing for an exchange of study of intercontinental dispersal events were terrestrial animals. Since this event caused the matched to continental-marine tie-points, palaeo- formation of extensive salt deposits, it is known magnetic data and radiometric dates, and finally as the Messinian Salinity Crisis (HSü et al. 1977). to the eustatic sealevel curves of HAOet al. (1987) (VAN DER MADE 1996, 1997b).Table 45.1 repre- sents stratigraphic data and intercontinental dis- persals of mammals. 458 In this paper, reference will be made to bio- the latest Oligoceneand earliestMiocene fauna is stratigraphic units and ages, even if an original not very well known. Similarly,the record of fossil description of a taxon does not give these data. mammals prior to MN 6 is very restricted in SE The references in this section will usually provide Europe and Anatolia, and prior to MN 5 virtually the link.