Chapter 10 Patterns in Insular Evolution of Mammals: A Key to Island Palaeogeography JOHN DE VOS, LARS W. VAN DEN HOEK OSTENDE, AND GERT D. VAN DEN BERGH Nationaal Natuurhistorisch Museum Naturalis, P.O. Box 9517, 2300 RA Leiden, The Netherlands, hoek@naturalis.nl, vos@naturalis.nl 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Gargano, Island Faunas on the Present Mainland. . . . . . . . . . . . . . . . . . . . . . . . . . 3. The Greek Isles, a Developing Archipelago . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Tertiary Faunas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Pleistocene Faunas of the Aegean Islands . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Southeast Asia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Sunda Shelf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Wallacea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3. The Philippines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Observations and Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316 318 320 320 323 326 326 328 334 334 337 Abstract The clearest examples of dwarfism and gigantism on islands are found in the fossil record. They form part of unbalanced faunas, which attest that only a few non-volant mammals were able to reach the island. Pygmy elephants and giant rats evolved in the isolation of these insular environments. Thus, the telltale signs of an insular fauna can be used to deduce the island’s palaeogeography. The faunas from the Gargano (Italy), a region presently forming part of the mainland, contain various giant rodents and a giant insectivore indicating that the Gargano was an island during the Mio–Pliocene. On the other hand, the Miocene and Pliocene faunas from the present-day Greek islands are balanced, indicating that they were connected with the continent at that time. The Pleistocene faunas from the same islands, however, are unbalanced, showing they lived in an isolated insular environment, and thus the faunas bear witness of the timing of tectonic processes. The same patterns as in the Mediterranean can also be found on the Indonesian islands. The islands of the Sunda shelf, which during glacials are connected to the mainland, have balanced faunas. In Wallacea we find the familiar pattern of dwarfed large mammals 315 W. Renema (ed.), Biogeography, Time, and Place: Distributions, Barriers, and Islands, 315–345 © 2007 Springer. 316 vos, hoek ostende, and bergh and giant murids in unbalanced faunas. The island of Flores even yielded remains of Homo floresiensis, a small hominin that shows that Man as well could be subject to the pattern of insular evolution. 1. Introduction Islands have been of vital importance in shaping the ideas of evolution formulated by Darwin and Wallace. On his voyage around the world on board of the “Beagle”, Darwin visited the Galapagos Islands. Here he discovered a fauna that, although showing signs of resemblance with the fauna of Patagonia, yet had peculiarities of its own. Moreover, he observed that species differed from island to island, even if less than 100 km apart. The remark of Mr. Lawson, vice-governor of the Archipelago, that he could tell immediately from which island a particular tortoise came, made him realize that every island had its own evolutionary history. Similarly, Wallace, while travelling among the islands of the East Indies, developed his ideas of evolution, independently of Darwin. Darwin and Wallace were by no means the only naturalists, or even the first, to recognize pecularities in island faunas. The first descriptions of the fauna from the Pacific islands date back to the 18th century, when naturalists Johann Forster and Joseph Banks accompanied James Cook on his exploration voyages. Ever since Darwin and Wallace, islands have played a dominant role in our thinking on evolution, biogeography and ecology. This is understandable, as was pointed out by Lomolino et al. (2006: 469): “Islands and other insular habitats, such as mountaintops, springs, lakes, and caves, are ideal subjects for natural experiments. They are well defined, relatively simple, isolated, and numerous – often occurring in archipelagos of tens or hundreds of islands.” The study of island faunas and floras got a major boost in the 1960s, when MacArthur and Wilson (1963, 1967) published their equilibrium theory of island biogeography. They considered the species richness of islands, which as a rule is lower than on the mainland, as a result of a balance between ongoing colonization and extinctions. Thus, they introduced a dynamic view on island communities, where earlier workers had assumed a more static history. Until then, low species richness was considered the result of limited resources, and after initial settlement by early colonizers the niches were filled, leaving no room for change. This is sometimes referred to as the static theory of islands (Dexter, 1978). The opposing view of MacArthur and Wilson inspired a wide range of articles on island biogeography (see Lomolino et al., 2006 and references therein). Also in the 1960s, Foster (1964) introduced the concept of dwarfism and gigantism for insular large and small mammals, respectively. The general pattern he noted became later known as the “Island Rule” (Van Valen, 1973). Foster first recognized this pattern in extant mammals and its general applicability was stressed by Lomolino (2005). Meiri et al. (2006), however, refuted the conclusions of Lomolino. Whereas neontologists debate the general character of the island rule, it patterns in insular evolution of mammals 317 has become widely accepted among palaeontologists. This is not surprising, since the clearest example of dwarfism (pygmy elephants and hippo’s) and gigantism (giant rats and dormice) can be found in the fossil record. The presence of these typical island mammals on the islands of the Mediterranean was already known through, for instance, the work of Dorothy Bate, who travelled the various islands in search of fossils (Bate, 1904, 1905, 1907, 1909). It is no coincidence that the best examples of insular evolution are within the realm of palaeontology. Most of these animals became extinct during the Holocene as their ecosystems were disturbed. In the Mediterranean these extinctions seem to have occurred when Neolithic Man introduced alien species, including domesticated cattle and rats (e.g., Reese, 1996). In other parts of the world such extinctions took place in historical times, such as for the moas on New Zealand and the dodo on Mauritius. Other possible causes for extinctions are the climate change at the end of the Pleistocene, or, in specific cases, volcanic eruptions, as Morwood et al. (2004) suggested for Flores. The patterns governing these fossil island faunas were described by Paul Sondaar, who first worked on the Mediterranean insular faunas (Sondaar and Boekschoten, 1967; Sondaar, 1976; Sondaar et al., 1986) and later continued studying the island faunas of the Indonesian Archipelago (Sondaar et al., 1994, van den Bergh et al., 2001). Sondaar (e.g., in Dermitzakis and Sondaar, 1978) explored to what extend the classical dispersal methods for mammals listed by Simpson (1940) could apply to the insular faunas of the Mediterranean. These possible dispersal methods are: (1) A corridor in which faunal interchange from one region to another is possible; (2) Filter dispersal: spread is probable for some organisms, but definitely improbable for others; (3) The pendel route: a route that is easily crossed vice versa between regions by some mammals, but an insurmountable barrier for others; and (4) A sweepstake: spread is impossible for most and very improbable for some organisms, but does occur accidentally. The patterns found on islands that displayed clear dwarfism or gigantism suggested that chance dispersal, the sweepstake route, was the most important way for colonizing these islands (Sondaar, 1976). In particular, the highly unbalanced nature of the faunas suggested the fauna was nearly isolated. The same groups of mammals appear over and over again as colonizers of islands. Proboscideans, which are known to be prolific swimmers, hippopotami, which have a semiaquatic life style and float more readily because of their fat, and artiodactyls (mostly deer), which gain buoyancy because of the gasses in their gut. Small mammals may have reached islands floating on driftwood. Mammal groups with more limited swimming capabilities, such as carnivores and perissodactyls, are invariably missing. This suggests that islands showing such selective pattern lay at the fringes of the reach of non-volant mammals. The isolation also gave room for speciation as evidenced by the various dwarf and giant species, and in some cases to adaptive radiations, such as found for the deer of Crete (de Vos, 2000). Even though the fossil record has yielded the clearest examples for dwarfism and gigantism, their role in theories on island biogeography seems limited. There appears to be a gap between the work done by palaeontologists and ecologists, and the first attempts to bridge that gap are only now being made (e.g., Raia and Meiri, 2006). 318 vos, hoek ostende, and bergh A possible explanation of the apparent differences lies in the isolation that allowed the evolution of dwarf elephants or giant rats on these islands. Since the islands could be reached by chance dispersal only, the dynamic equilibrium of ongoing colonizations and extinctions that plays a pivotal role in MacArthur and Wilson’s theory is lacking. It is doubtful whether the faunas of islands which are nearer to the mainland and do have a certain degree of exchange would be recognizable as insular environments in the fossil record. But some oceanic islands yielded because of their isolated position typical faunas, which are totally different from those found on the nearest mainland. Reversing the argument we can state that the presence of an unbalanced mammal fauna in the fossil record, particularly when some of its members display gigantism or dwarfism, points to an isolated insular environment. If, on the other hand, on a present-day island a balanced fauna is found, it indicates that in the past the island was connected to, or at least in the vicinity of, the mainland. In this paper we discuss several examples of the way in which fossil mammal faunas can be used to explore the palaeogeography of islands, focussing on the regions we know from personal experience. 2. Gargano, Island Faunas on the Present Mainland The Mediterranean is an area of intense tectonic activity, leading to dramatic changes in the palaeogeography throughout the Cainozoic. A large submergence zone, as a result of the northward movement of Africa sets the stage. One of the more active orogenetic zones is found in present-day Italy. Mammal faunas from that region show that during the Tertiary, islands emerged and submerged again. One such island fauna is found on Sardinia, where an unbalanced small mammal fauna with endemic species was recovered at Oschiri Road Cut (de Bruijn and Rümke, 1974). The most important Italian island faunas, however, were discovered in the 1970s on the mainland, when M. Freudenthal studied fossils from fissure fillings on the Gargano peninsula. Basically the Gargano consists of a block of uplifted Mesozoic limestone, which is exploited as “marble”. This block is connected to the mainland by a low area, the so-called Foggia Graben or Tavoliere plain. The limestone of Gargano shows extensive superficial karst development, and the various quarries in the western area contain numerous fissure fillings, many of which are fossiliferous. Soon it became apparent that the assemblages found represent an island environment. The faunas are unbalanced, lacking perissodactyles, proboscideans, and carnivores, with the exception of the otter Paralutra garganensis (Willemsen, 1983). The artiodactyls are represented only by one aberrant form, the deer-like Hoplitomeryx (Leinders, 1984). Another indication that we are dealing with island faunas is the presence of various giant forms among the micro mammals, for example the murid Mikrotia (Freudenthal, 1976, 2006), the dormouse Stertomys (Daams and Freudenthal, 1985), and the erinaceid Deinogalerix (Freudenthal, 1972; Butler, 1980). patterns in insular evolution of mammals 319 The importance of the Gargano lies in the occurrence of insular faunas from different ages from a single palaeo-island. Usually only one or a few sites are known from a certain island, but in Gargano c. 75 fissure fills were sampled, which seem to represent sequential time slices. Unfortunately, we can only reconstruct the time sequence based on the fossil contents of the different fissure fillings. A biostratigraphy of the Gargano faunas was made by Freudenthal (1976), who based his sequence primarily on the stage of evolution in the various Mikrotia lineages. Thus, the development of the fauna and its components can be reconstructed through time. As to the age of the fauna, an upper limit is set by the marine sedimentary rock covering, at least in some places, the top of the quarries and fissures. Unfortunately, opinions about the age of these marine sediments vary. D’Allessandro et al. (1979) considered them to be Tortonian, whereas Valleri (1984) described a marine sequence of Lower and Upper Pliocene, in an area where d’Allesandro et al. (1979) had recognized Upper Miocene sedimentary rocks. The faunas set the lower limit themselves, but here too, there is little agreement. Freudenthal (1985) saw in the presence of Cricetulodon in Biancone, one of the oldest fissure fillings, an indication that the fauna must have entered the area in the Turolian. The presence of an Apodemus-like murid in the older faunas he considered in line with such an age. de Giuli et al. (1985; 1986a, b; 1987a, b) preferred a Messinian or even postMessinian age for the colonization of the island. However, colonization after the Miocene does not seem very plausible, given the absence of arvicolids. Thus, the correlation of the insular faunas from Gargano to the continental faunas of that period remains uncertain. Of course, one of the problems with correlating island faunas to the mainland is that the insular forms have changed so much that it is difficult to pinpoint the mainland ancestor. In the case of Gargano the typical examples of island evolution such as Hoplitomeryx and Deinogalerix have been studied most extensively. Answers to the origin of the fauna may, however, lie with the relatively unaltered micro mammals, some of which are currently under study. In Gargano a second major phase of superficial karst development occurred during the Early Pleistocene, affecting the Neogene sedimentary rocks covering the Mesozoic block and the fissure fillings it contains. Cave deposits intercalated in these Neogene sediments have yielded a Villafranchian fauna with, a.o., Mammuthus meridionalis, Equus, Stephanorhinus cf. etruscus, Homotherium latidens, Panthera gombagzoegensis, Pachycrocuta brevirostris, Apodemus, and Allophaiomys (Freudenthal, 1971; Abbazzi et al., 1996). This is the typical Early Pleistocene mammal fauna found elsewhere on the continent. Therefore, we can conclude that at that time Gargano was already connected to the mainland. The island faunas from Gargano are not the only Tertiary insular faunas from the Mediterranean. The islands of Mallorca and Menorca (Balears, Spain) have also yielded insular faunas, the oldest of which dates back to the Pliocene. There are some remarkable similarities to Gargano. For instance, the faunas are also characterized by the presence of an aberrant artiodactyle, here the caprine Myotragus (Bover and Alcover, 2003). Like in Gargano, there are fissure fillings of various ages, allowing us to follow the changes in the fauna over a respectable period of time. Unlike 320 vos, hoek ostende, and bergh Gargano, this faunal evolution continued right up to the beginning of the Holocene. The island faunas disappeared as Man entered the scene (Bover and Alcover, 2003), just like they did on other Mediterranean Islands, e.g., in the Greek archipelago. 3. The Greek Isles, a Developing Archipelago The Greek islands make up only a minor part of the total surface of Greece. Nevertheless, many of the mammal faunas found in that country have been excavated on the islands, and together they give an almost continuous record of 20 My of faunal evolution. 3.1. Tertiary Faunas The oldest small mammal fauna from the Greek islands (Fig. 1) was retrieved from a lignite mine near Aliveri on the island of Evia. The site yielded an MN (=Mammal Neogene Zone) 4 smaller mammal fauna (de Bruijn et al., 1980; van der Meulen and de Bruijn, 1982; Klein Hofmeijer and de Bruijn, 1985; Doukas, 1986; Lopez-Martinez, 1986; Alvarez-Sierra et al., 1987), and a felid (SchmidtKittler, 1983). Palaeogeographic reconstructions of the Early Miocene of the area suggest that Aliveri was situated on the same landmass as present-day Anatolia (de Bruijn and Saraç, 1991, Fig. 1). However, Anatolia and Europe formed two clearly separated bioprovinces at the time (van den Hoek Ostende, 2001a). The presence of Eomyidae and, e.g., the insectivore Heterosorex shows Aliveri has a fauna similar to the ones found in Central and Western Europe, and unlike the ones found in Anatolia. The oldest larger mammals from the Greek archipelago are found on the island of Psara, Lesbos and Chios. From Psara an upper right M1 is known from Prodeinotherium bavaricum (Besenecker and Symeonidis, 1974). The same species is also known from Lesbos from St. Gavathas (Koufos et al., 2003). Based on K/Ar dating of the overlying volcanic rocks the authors placed the Gavathas specimen in the upper part of MN 3, suggesting that it is the oldest deinothere in Europe. In 2004 small bones from an artiodactyle were found at Gavathas (de Vos, unpublished data). P. bavaricum is also known from Chios. This species is found south of Cape St. Helena, between Thymniana and Keramia, together with an Aragonian fauna (MN 5 till MN 7/8, maybe MN 9; Kondopulou et al., 1993; Lehmann and Tobien, 1995) consisting of the rodent Megapedetes aegaeus, the bovids Hypsodontus cf. gaopense and Tethytragus cf. koehlerae, the tragulid Dorcatherium sp., Lophocyon paraskevaidisi, Sanitherium slagintweiti, Listriodon sp., aff. Euprox furcatus, Georgiomeryx georgalasi, Choerolophodon chioticus (Paraskevaidis, 1940; Thenius, 1956; Melentis and Tobien, 1968; Tobien, 1969, 1980; Sen, 1977; Rothausen, 1977; Lehman and Tobien, 1995; Koufos et al., 1995; de Bonis et al., 1997a, b, 1998). 9 3 TYPE OF FAUNA UNBALANCED ISLAND FAUNA Pleistocene loc. Damatria Apolakkia Maritsa Kalithies Pleistocene loc. Ag. loannes Pleistocene loc. Vryses Petras Kastellios Plakia Malembes Thymina Vatera Gavathas Alveri CONTINENTAL STAGES 6 5 Aragonian 7/8 4 KARPARHODES THOS CRETE forest open woodland BALANCED MAINLAND FAUNA 10 Burdigalian MIOCENE Early Middle Late 11 CHIOS steppe biotope 13 12 Vallesian 14 Mes- Zan- Piacen- Gelasinian clean zian sian RusciTurolian Villanyian nian 15 Tortonian 16 LESBOS Biharian MQ2 MQ1 17 EV- PSAIA RA 321 Toringian EPOCHS MN. ZONES PLIOCENE Early Late PLEISTOCENE Early Middle Late HOLOCENE patterns in insular evolution of mammals Fig. 1 A survey of the mammal localities of Evia, Psara, Lesbos, Chios, Crete, Karpathos, and Rhodes, and their supposed ages (altered after Sondaar et al., 1986). Most Miocene faunas from the Greek isles have been found on Crete. At Malembes a faunule was found with cf. Prohyrax hendeyi, Dorcatherium naui and a bovid. The fauna was placed in MN 6, Middle Aragonian (van der Made, 1996). At Plakia a smaller mammal fauna was found consisting of Spermophilinus cf. bredai, Blackia? sp., Forsythia? sp., Democricetodon aff. cretensis, Cotimys sp., and Glirudinus sp. (de Bruijn and Meulenkamp, 1972). van der Made (1996) added the pig cf. Propotamochoerus palaeochoerus to the faunal list and assumed a Late Aragonian age for the fauna (MN 7/8). Somewhat younger are the sites at Kastellios Hill, which are placed in the Vallesian (MN 9-10, van der Made, 1996). The fauna consists of Cricetulodon cf. sabadellensis, Progonomys woelferi, P. cathalai, Spermophilinus bredai, Muscardinus cf. crusafonti, Schizogalerix sp., Hipparion sp., Taucanamo?/ Yunnanochoerus sp., cf. Pliocervus pentelici, and unidentified remains of a bovid and a carnivore (van der Made, 1996). Apart from these faunas there are some isolated finds, like Microstonyx cf. major at Petras (van der Made, 1996), and a mastodon at Vryses (Benda et al., 1968). 322 vos, hoek ostende, and bergh The latter site also yielded the ochotonid Prolagus sp. (de Bruijn in: van der Made, 1996) and is considered to be of Late Miocene age (Benda et al., 1968). The sparse finds from Vryses are the youngest of the Miocene mammals on Crete. During the Late Miocene tectonic changes led to the fragmentation of the Cretan area, transforming it into an archipelago during the Late Tortonian/Middle Pliocene (Benda et al., 1974; Drooger and Meulenkamp, 1973). The whole island of Crete became subject to submergence from the Early Tortonian onwards and became a shallow sea with islands and shoals (Meulenkamp, 1971). Crete was divided into at least four islands during the Pliocene (Sondaar and Boekschoten, 1967). The absence of terrestrial mammals from this period may indicate that the area was too small to support stable populations (Sondaar et al., 1986) or that the fossilization potential was so small that they just not have been found. Crete re-emerged as an island during the Late Pliocene (Benda et al., 1974). Although, there are no mammal faunas known from the Neogene of Crete, there are Late Miocene–Pliocene faunas from Rhodes. The oldest fauna has been found in Kalithies (Sondaar et al., 1986), and consists of the typical elements from Early or Middle Turolian (MN11/12) from the Greek mainland, the so-called Pikermi fauna (Sen et al., 1978), including the three-toed horse Hipparion sp., the hyaenid Ictitherium orbingyi and the sabre-toothed cat Machairodus aphanistus. A similar fauna was found on Samos (Bernor et al., 1996). At Maritsa on Rhodes a large assemblage of small mammals was collected, which included sixteen rodents species, two insectivores and one lagomorph (de Bruijn et al., 1970). This fauna is considered to be of an Early Pliocene age (MN 14). From Apolakkia an MN 15 fauna was described by van de Weerd et al. (1982). The Apolakkia fauna contains among others Castor fiber, Hipparion aff. crassum, Cervus rhenanus, and Mimomys occitanus. Theodorou et al. (2000) described Anancus arvernensis from this site. Additionally, Karpathos has yielded a Pliocene faunule. At Ag. Ioannis a faunal association with Muscardinus and Kowalskia was found. This faunule is placed in MN 14/first part MN 15 (van de Weerd et al., 1982). Doukas and Athanassiou (2003), in their review of the Plio/Pleistocene Proboscidea from Greece, mention the Late Pliocene/Early Pleistocene continental proboscideans A. arvernensis and M. meridionalis from Kos. The latter species was also found on Evia and Kythera. The best-known faunas from the Greek isles from that period are found in different sites near Vatera on Lesbos (de Vos et al., 2002). Here the typical Villafranchian fauna, also known from Balkan sites such as Volax, Sesklo, Dafnero, Gerakarou, and Pyrgos in Greece (van der Meulen and Van Kolfschoten, 1986; Koufos et al., 1991; Athanassiou, 1996; Kostopoulos, 1996; Koufos and Kostopoulos, 1997), Slivnitsa in Bulgaria (Spassov, 1998) and Valea Gràunceanului and Fîntîna lui Mitilan in Roumania (Radulescu and Samson, 1991), is found. The presence of both Anancus and Mammuthus suggests a Late Pliocene age for the Vatera sites. Since Equus is already present at the sites, they must be younger than the first occurrence of this genus in Eurasia (2.5–2.7 My). None of the Tertiary faunas from the Greek islands show any signs of endemism. The faunal elements from the various sites are the same as found in coeval patterns in insular evolution of mammals 323 localities on the mainland. Therefore, Sondaar et al. (1986) concluded that the various islands studied at the time were still connected to the mainland during the Tertiary. This certainly also holds true for Samos, which was studied later. The close resemblance of the Vatera assemblages to the mainland faunas, and the presence of carnivores such as the canid Nyctereutes and a sabre-toothed cat, clearly shows that Samos was still connected to the mainland during the Late Pliocene. 3.2. Pleistocene Faunas of the Aegean Islands Pleistocene Crete is a classical example of an island that was colonized by sweepstake dispersal. From the time of colonization until the Holocene it had an unbalanced endemic island fauna. Crete got its present configuration in the late Early Pleistocene (Sondaar et al., 1986). The Pleistocene island fauna only contains cervids (de Vos, 1996a), elephants (Mol et al., 1996), hippos (Spaan, 1996), murids (Mayhew, 1996), shrews (Reumer and Payne, 1986; Reumer, 1996), birds (Weesie 1988), and reptiles (Brinkerink, 1996) (Fig. 2). Perissodactyles and carnivores, with the exception of an otter (Willemsen, 1996), are lacking. Mayhew (1996) distinguished five biozones based on the endemic Pleistocene murid species of Crete. These species belong to two genera, Kritimys and Mus. The endemic murid Kritimys was larger than the brown rat, Rattus norvegicus (Mayhew, 1996), while the species of Mus are of small size. If we look at the ungulates, there is one faunal turnover. The faunal assemblages can be summarized as follows (Mayhew, 1996): 3.2.1. Mammuthus creticus – Hippopotamus creutzburgi Fauna, or the Kritimys Zone The oldest Pleistocene land vertebrates are from the locality Siteia 1. Besides Kritimys aff. kiridus (Mayhew, 1996), a rib of Hippopotamus creutzburgi was found (Spaan, 1996). At Cape Maleka 1, the dwarf elephant M. creticus is associated with Kritimys (Mayhew, 1996). The evolutionary stage of Kritimys indicates that this site is somewhat younger in age than Siteia 1, but it is still placed in the Early Pleistocene. The probable ancestors of Kritimys and the shrew Crocidura zimmermanni, found from the K. catreus subzone onwards, are of a Late Pliocene/ Early Pleistocene mainland stock (Mayhew, 1996; Reumer, 1996). H. antiquus is considered to be the ancestor of the dwarf H. creutzburgi, which had a more unguligrade stance than the mainland species (Spaan, 1996). M. creticus probably descended from M. meridionalis (Mol et al., 1996). Apparently the Early Pleistocene H. antiquus and M. meridionalis dispersed by sweepstake route to Crete and adapted to the island environment by becoming dwarfed. 3.2.2. Elephas creutzburgi – Candiacervus Fauna, or the Mus Zone The earliest occurrence of Mus on Crete is at the Stavros Micro site. The earliest find of the cervid Candiacervus is in Charoumbes 2, the first co-occurrence of Elephas and Candiacervus in Charoumbes 3. The younger site (Gerani 5) has an Fig. 2 Elephas creticus Charoumbes A Xeros Milatos 1 Bali 2 Cape Meleka 1 Cape Meleka 3 Sitia 1 ? Biostratigraphy and faunal turnovers on Crete during the Pleistocene (after de Vos, 1996). K.a . kiridus Kritimys kiridus Kritimys catraus Milatos 3 lower Mavro Mouri 4c Zourida Rethymnon ssure Kalo ChoraÞ Simonelli Cave Charoumbes 3 Charoumbes 2 Milatos 2 and 4 Milatos 3 upper Stavros Cave inside Stravos micro Hippopotamus Stavros Cave outside Kato Zakros creutzburgi parvis Hippopotamus Katharo creutzburgi creutzburgi Elephas antiquus Elephas creutzburgi Gerani 22 Gerani 5 Gerani 6 Gerani 23 Gerani 4 Gerani 2 Bate Cave4 Liko Localities Candiacervus sp.VI Candiacervus sp.V Candiacervus rethymnensis Candiacervus cretensis Candiacervus spp. II Candiacervus ropalophorus Candiacervus sp. indet. Mus bateae Mus minotaurus Range-zones Holocene Kritimys Mus Zones Sub-zones Deer species 324 vos, hoek ostende, and bergh Pleistocene patterns in insular evolution of mammals 325 absolute Electro Spin Resonance (ESR) age of 127,000 years ± 20% (Reese, 1996). The youngest locality is Gerani 2, which is AAR (Amino Acid Racemization) dated at 47,000 years ± 20% (Reese, 1996). Early Neolithic deposits cover the Pleistocene deposits in this cave. de Vos (1984a) recognized six sizes in the endemic genus Candiacervus, which can be explained as an adaptive radiation of the ancestral stock (de Vos, 2000). Since there is still considerable overlap, particularly in the postcranial elements, no true species can be recognized. Nevertheless, species names are sometime attributed to extremes of the size spectrum (e.g., Capasso Barbato and Petronio, 1986). The adaptive radiation probably resulted from sympatric speciation (de Vos, 1996a). The large Elephas from Crete is a little smaller than the continental Elephas antiquus. Its taxonomic status is not clear and continues to be under discussion. Dermitzakis and Sondaar (1978) classified their material as Elephas cf. antiquus. Other authors have considered it to be a species on its own like E. creutzburgi by Kuss (1965) or E. chaniensis by Symeonides et al. (2001). Poulakakis et al. (2002) took an intermediate position by considering the Cretan elephant as a subspecies of the continental form (Elephas antiquus creutzburgi). The otter Lutrogale cretensis is the only mammalian carnivore in the Cretan fauna (Willemsen, 1996). It shows an adaptation to terrestrial life and was feeding on fish, crustaceans and small land vertebrates (Willemsen, 1996). The genus Mus probably arrived in the early Middle Pleistocene (Mayhew, 1996). The first occurrence of E. antiquus in Europe is also in the early Middle Pleistocene, c. 700 ka. So, the faunal turnover from the M. creticus – H. creutzburgi fauna to the E. creutzburgi – Candiacervus fauna was no earlier than the transition from Early to Middle Pleistocene. Due to the many localities of different ages, Crete can be considered as a case history for colonization, island adaptation of ungulates and rodents, and extinction of island endemics during the Holocene. It was, however, certainly not the only Greek island with a Pleistocene insular fauna. Doukas and Athanassiou (2003) give an overview from the islands of the Aegean with unbalanced endemic island faunas, mostly consisting of solely (dwarf) proboscideans. Apart from Crete these are Rhodes, Tilos, Dilos, Astypalaea, Seriphos, Milos, Naxos, Paros, and Kytnos. On Karpathos and Kassos, probably one single island at the time, also endemic cervids are found (Sondaar et al., 1996). Examples of insular evolution, besides the Greek archipelago, can also be found on other Mediterranean islands. The Balears were mentioned above, with their Myotragus faunas that survived up to the end of the Pleistocene. Sicily is famed for having the smallest of all dwarf elephants, E. falconeri. Other than that, the giant dormouse Leithia was found here. On Cyprus, remains have been found of a dwarfed hippopotamus and elephant. The study of the Mediterranean islands, which for a large part was the work of the late Paul Sondaar at Utrecht University, thus showed a consistent pattern occurring over and over again. Sondaar (1977) proposed a model for these islands. Colonization by sweepstake dispersal resulted in unbalanced endemic island faunas. At the time, dwarfism among large insular mammals was considered a token 326 vos, hoek ostende, and bergh of degeneration. Sondaar (1977) showed that dwarfism, as well as gigantism among small mammals, was part of an adaptive pattern. Although the consistency of the pattern could be demonstrated by comparing different Mediterranean islands, a weakness was that the island faunas of a single region were compared, often based on the same ancestral species and under comparable paleo-climatic conditions. The need was felt to compare the Mediterranean islands with another archipelago. The Indonesian islands offered a promising perspective, particularly since collections were already available in Naturalis. 4. Southeast Asia Naturalis holds the Dubois collection, with fossils from sites on Sumatra and Java, extensively described in taxonomical studies by Hooijer (e.g., 1947, 1955). In the 1980s they formed the basis of a new stratigraphic and palaeoecological interpretation of the hominid bearing deposits of Java (de Vos et al., 1982). The other research areas were the islands of Wallacea. Here too, a lot of taxonomical work had already been published by Hooijer (e.g., 1949, 1953a, b, c, 1957a, 1964, 1969, 1972a, b) on material stored in Naturalis. These collections formed the basis for research of the islands of the Sunda Shelf and the islands of Wallacea. 4.1. Sunda Shelf 4.1.1. Early/Middle Pleistocene of Java Based on work of de Vos et al. (1982), two faunal turnovers can be recognized on Java. Comparison of them with mainland faunas can be used to deduce dispersal routes. The Satir Fauna The oldest proboscidean on Java, the mastodon Sinomastodon bumiajuensis, originates from the so-called Satir Fauna from the late Early Pleistocene. The age, considered to be 1.5 Ma, is based on the faunal similarity between the Satir fauna from the type locality and the lower part of the Sangiran Formation, which has been fission track dated in the Sangiran area (Suzuki et al., 1985). The only other species in the fauna are a hippo (Hexaprotodon simplex), cervids and a giant tortoise Geochelone. In this fauna no fossil hominids have been found. The pollen spectrum suggests that the environment had been swampy. Sinomastodon and Hexaprotodon are insufficiently well studied to know to what degree they are endemic. The unbalanced character of the fauna, with the same families as found on the isles of the Mediterranean and the presence of a giant tortoise, points to island conditions. As Hexaprotodon is an element of the Siwalik and Burma fauna, a Siva-Malayan origin and migration route is plausible. In conclusion, there are indications that Java was an island during the late Early Pleistocene, and was colonized by hippos, cervids and proboscideans by sweepstake dispersal. patterns in insular evolution of mammals 327 Stegodon-Homo erectus fauna The sites Trinil, Kedung Brubus and Ngandong are attributed to the Stegodon-H. erectus fauna association. This association clearly shows affinities with the faunal association from the Indian Subcontinent (the Siwaliks) and Burma. However, species diversity is much lower. For example, horses and camels never reached Java. Five species and one genus occur both in the Siwaliks faunas and at Java, viz., Hexaprotodon sivalensis, P. brevirostris, Caprolagus cf. sivalensis, H. ultimum [? = H. latidens; Galobart et al., 2003], Nestoritherium cf. sivalense, and Megantereon megantereon. Three species from this Javan fauna are closely related to Siwalik species, viz., Stegodon trigonocephalus with Stegodon ganesa, E. husudrindicus with E. hysudricus, and Duboisia santeng with the Boselaphini. de Vos (1996b) postulated that the Stegodon-H. erectus fauna association originated from the Siwaliks and reached Java via the so-called Siva-Malayan Route. The absence of camels and horses, but the presence of three carnivore species (the Pachycrocuta, Homotherium, and Megantereon) indicates that colonization took place via filter dispersal. 4.1.2. Late Pleistocene of Java At the end of the Middle Pleistocene the species of the Stegodon-H. erectus fauna association became extinct. A faunal turnover took place and a new fauna migrated into the Indonesian Archipelago, the Pongo-H. sapiens fauna to which the Punung site is attributed (de Vos, 1996b). In this fauna we find the Indian elephant (E. maximus), orang-utan (Pongo pygmaeus), gibbon (Hylobates syndactylus), pig-tailed macaque (Macaca nemestrina) and Malayan bear (Ursus malayanus), all species which are still extant on the continent or in other places of the Indonesian Archipelago, but are no longer found on Java (Badoux, 1959). The large number of orang-utan and the presence of other primates indicate a humid tropical rainforest environment. Recently, a Punung-like fauna was discovered at Gunung Dawung (Storm et al., 2005). The faunal association from the Gunung Dawung site has been OSL (Optical Stimulated Luminescence) dated at 128 ka (Westaway et al., in preparation), corresponding with the onset of the last interglacial. A younger faunule from the terminal phase of the last interglacial (radiocarbon dated at 35 ka) originates from the site Cipeundeuy, containing E. maximus, Rhinoceros sondaicus, Muntiacus sp., cervids, and Bubalus sp. (van den Bergh, 1999). Late Pleistocene of Sumatra Before searching in Java for the “missing link”, Dubois (1889) excavated in the Padang Highlands, central Sumatra, during 1889–1890. The bulk of the material originated from three caves, viz., the Lida Ajer Cave near Pajakombo, the Sibrambang Cave and the Djambu Cave near Tapisello, all containing more or less the same fauna. Hooijer (1947, 1948, 1955) described the fossils of these caves and de Vos (1983) gave their faunal lists. The Sumatran cave faunas are characterized by an abundance of Pongo, suids, Hylobates, Acanthion and Rhinoceros teeth. Further, there are E. maximus molars and teeth of P. tigris. According to de Vos (1983), the Sumatran cave fauna suggests the presence of a tropical rainforest. 328 vos, hoek ostende, and bergh A right upper central incisor, which is semi-shovel shaped, and a left upper molar, both from Lida Ajer cave, were identified as H. sapiens by Hooijer (1948). The Cave material of Djambu is dated at 60–70ka, while the Lida Ajer material gives a date of 80 ka, all by AAR. 4.1.3. Late Pleistocene of Borneo Lydekker (1885) described a tooth of a Mastodon (Stegolophodon latidens) from Borneo. Whether or not this small-sized proboscidean is an island form is not yet clear. Medway (1979) provided an overview of the fauna in the Niah Cave, which is dated as 50 ka, to Recent. The list is composed of both Pleistocene and Holocene species found in Niah. Next to domesticated mammals (like Canis familiaris in the upper 15–30 cm level and Sus scrofa dom. in the 0–60 cm level), Pongo pygmaeus in the levels of 260–270 cm and Hylobates in the levels of 150–180 cm are also present. The last two species indicate a tropical rainforest. The fauna is balanced. Similar faunas as in the Sumatran caves, Borneo, and Java (Punung) with orang-utan are also found in fossil sites on the Asian continent, like Vietnam (Lang Trang Cave), Cambodia, (Phnom Loang) and China (de Vos, 1984b; Vu the Long et al., 1996) Since the faunas of Punung, Borneo, and Sumatra are balanced and represent a tropical rainforest, which cannot cross a water barrier, we may assume a land connection with the mainland. A lowering of the sea level between 126 and 81 ka (Storm, 2001), but slightly earlier according to the new OSL dating evidence from Gunung Dawung (Westaway et al., in preparation), connected Sumatra, Java, and Borneo (the Sunda shelf) to the continent (Fig. 3), allowing corridor dispersal of the mainland fauna unto the Sunda Shelf (de Vos et al., 1999; de Vos and Vu the Long, 2001). Among the species that entered the region at this stage was H. sapiens (Storm et al., 2005). 4.2. Wallacea The many islands and seas between the Sunda and Sahul shelves have been given the name Wallacea. In contrast to the islands of the Sunda Shelf, these islands were not connected to the mainland at times of low sea levels. Here we find unbalanced endemic island faunas. Every island had its own evolutionary history, although like in the Mediterranean a general pattern can be recognised. 4.2.1. Sulawesi The first dwarf proboscidean remains from Southwest Sulawesi (Fig. 4) were described by Hooijer (1949) as Archidiskodon celebensis. Van Heekeren had recovered the fossil material near the village of Sompoh west of the Walanae River in the Sengkang district. Later collecting at several localities in the same area yielded abundant fossil material that permitted Hooijer (1953a, 1954a, 1955, 1972a) to assign the particular characteristics to this dwarf proboscidean, which was later reclassified as E. celebensis. patterns in insular evolution of mammals 329 BORNEO 40º Sangiran Trinil Modjokerto 500 km JAVA Satir Punung Ngandong Kedung Brubus CHINA 30º INDIA SIVA MALAYAN ROUTE 20º SINO MALAYAN ROUTE L A O S Langtrang BURMA THAILAND CAMBODIA PHILIPPINES VIETNAM 10º Niah SANGIHE BORNEO Equator Caves SUMATRA SULAWESI Sangiran Trinil Modjokerto SUMBA FLORES JAVA 10º Satir Ngandong Fig. 3 Punung Kedung Brubus TIMOR Geographic position of the sites in southeast Asia mentioned in the text. 330 vos, hoek ostende, and bergh Stratigraphy Lithostratigraphic Unit Age Faunal unit Cave deposits Toalian sites Holocene Subrecent to Recent Colluvium Tanrung Formation Late Pleistocene or Holocene Late or Middle Pleistocene Tanrung Geochelone atlas Crocodile species Tryonychidae gen.et sp. indet Celebochoerus heekereni Celebochoerus, shortlegged species “Elephas”celebensis Stegodon sompoensis Meium-large sized Stegodon Stegodon sp. B Highcrowned Elephas Anoa sp. Sus celebensis Species ?? ? Beru member Walanae Formation ? Subunit Early B Pleistocene Walanae Subunit Late Pliocene A 2,5 Ma Samaoling Late Pliocene member Fig. 4 Biostratigraphy and faunal turnovers on Sulawesi during the Late Pliocene– Pleistocene (after van den Bergh et al., 2001). In addition to E. celebensis, Hooijer (1953a) announced the presence of Stegodon. Initially he doubted whether the fragmentary material should be attributed to a pygmy or a normal-sized Stegodon. In 1964, after additional fossil Stegodon material had been collected, he concluded that all the Stegodon material described so far belonged to a dwarf species, which he named S. sompoensis. Hooijer (1972b) described some Stegodon molar fragments collected in the previous years, which he attributed to S. cf. trigonocephalus, based on their supposedly large size. He concluded that there must have been both a large- and small-sized Stegodon in the fossil fauna. Hooijer (1948b, 1954) also described remains of an endemic suid from Southwest Sulawesi, Celebochoerus heekereni, and a giant tortoise (Hooijer 1948c, 1954b). Stratigraphic data of the material described by Hooijer are lacking. Fieldwork during 1990–1994 clarified the stratigraphic sequence (van den Bergh et al., 2001; van den Bergh, 1999). In the Late Pliocene sedimentary rocks of c. 2.5 Ma a pigmy Stegodon (Stegodon sompoensis) and a pigmy Elephas (“Elephas” celebensis) are present in association with Celebochoerus heekereni and giant tortoise. During the Early Pleistocene, a large-sized Stegodon, represented by the few large-sized molar patterns in insular evolution of mammals 331 fragments and postcranial fragments, migrated into south Sulawesi. This might either be Stegodon trigonocephalus from Java or another large-sized Stegodon from the Philippines or the Asian mainland. By the middle or late Pleistocene, both pygmy proboscideans had become extinct, while the large-sized Stegodon continued or, alternatively, a new immigration of a large Stegodon took place besides a new immigration of a highly advanced Elephas species. 4.2.2. Flores The discovery of stone artefacts in association with remains of a large Stegodon, S. trigonocephalus florensis, and large murids, Hooijeromys nusatenggara at the localities Mata Menge and Boa Leza in west Central Flores (Fig. 5) was reported by Maringer and Verhoeven (1970). The same fauna had also been recovered from the locality Ola Bula, though from the latter locality no in situ artefacts were reported. Stratigraphy Lithostratigraphic Unit Holocene Subrecent Fauna Cave depostits Late Pleistocene 18 ky Member B Cave depostits 0.8-0.7 Ma Fauna B Member A Faunal Unit Ola Bula Formation Age 0.9 Ma Fauna A Varanus komodoensis Crocodile sp. Geochelone sp. Stegodon sondaari Homo sapiens Homo erectus / Homo floresiensis Stegodon florensis Hooijeromys nusatenggara Macaca Deer Pigs Rattus spp. Species Ola Kile Formation Fig. 5 Biostratigraphy and faunal turnovers on Flores during the Pleistocene (after van den Bergh et al., 2001). 332 vos, hoek ostende, and bergh Based on the association with S. trigonocephalus florensis, the artefacts, described as a number of pebble tools and retouched flakes mostly made of volcanic rock, were inferred to be middle or late Pleistocene and it was speculated that H. erectus? might have reached the Lesser Sunda islands. The large to medium-sized Stegodon from the localities Ola Bula, Boa Leza, Mata Menge and Dhozo Dhalu is slightly more advanced than Stegodon trigonocephalus from Java (Hooijer 1953b, 1972b). The taxonomic status is, like the Cretan large elephant, under discussion. Hooijer (1957a) considered it to be a subspecies of the Javanese species (Stegodon trigonocephalus florensis), while it was considered to be a species on its own (Stegodon florensis) by van den Bergh (1999) and van den Bergh et al. (2001). Additionally, fossil remains of the giant rat Hooijeromys nusatenggara have been found at Mata Menge, Ola Bula, Dhozo Dhalu, and Boa Leza. Maringer and Verhoeven (1970) had discovered this murid earlier in the same area (Musser, 1981). Further, at Mata Menge a few teeth of a small crocodile were found, while at Dhozo Dhalu teeth of Varanus komodoensis occurred in association with the younger fauna. This still extant species seems to have been the only element from the older vertebrate fauna that did not become extinct. In 1982 a new locality was discovered 2.5 km east of Mata Menge and 250 m southeast of Ola Bula, yielding fossils of a giant tortoise and a pygmy Stegodon (Sondaar, 1987). This locality, known as Tangi Talo, contained a distinct fauna that is stratigraphically below the Ola Bula excavation of Verhoeven (Sondaar et al., 1994; van den Bergh, 1999; van den Bergh et al., 2001). Based on faunal correlations, the Tangi Talo fauna was also inferred to be older than the artefact-bearing layer at Mata Menge. The fossil locality near Tangi Talo is the only one that yielded the giant tortoise-pygmy Stegodon fauna. The pygmy Stegodon from Tangi Talo represents a distinct species, Stegodon sondaari, differing from the pygmy stegodonts known from Java, Sulawesi and Timor, and the tortoise is on average much smaller than the fossil giant tortoises known from the other islands (van den Bergh, 1999; van den Bergh et al., 2001). According to Erick Setiabudi (personal communication), who currently studies the fossil giant tortoises from the various Indonesian islands, every island has its own species of Colossochelys, a genus also known from the Upper Siwaliks. At Tangi Talo remains of Varanus komodoensis were also recovered from the same layer. No artefacts have been found in the fossiliferous layer at Tangi Talo or in the tuffaceous interval in which this layer occurs. The colonization of the island by humans coincides with a faunal turnover on the island. An endemic island fauna with giant tortoise and a pygmy Stegodon is replaced by an endemic island fauna with Stegodon florensis and Hooijeromys nusatenggara. This younger fauna is associated with artefacts at the localities Mata Menge and Boa Leza, indicating the co-occurrence with humans. Palaeomagnetic dating results suggest a late early Pleistocene age for the Tangi Talo fauna and early Middle Pleistocene for the Mata Menge fauna (Sondaar et al., 1994), indicating an age of c. 0.7 Ma for the artefact-bearing layer. Fission track ages of the same layer by Morwood et al. (1998) range between 0.88 ± 0.07 and 0.80 ± 0.07 Ma. More recent excavations carried out in 2004–2005 have yielded a much larger sample of patterns in insular evolution of mammals 333 stone tools from Mata Menge, which shows resemblance with the Late Pleistocene artefact assemblage from Liang Bua, which have been attributed to H. floresiensis recovered from the same deposits (Brumm et al., in press). Brown et al. (2004) and Morwood et al. (2005) announced a new small-bodied hominin (H. floresiensis) from the cave Liang Bua. It was 1 m tall and had an endocranial volume of 380 cm3. According to the authors the most likely explanation for its existance is long-term isolation, with subsequent dwarfing, of an ancestral H. erectus. Based on three alternative dating methods, the age was considered to be between >38 ka and c. 13 ka (Morwood et al., 2004). In the Liang Bua deposits H. floresiensis co-occurs with Stegodon and Varanus komodoensis. The ancestor of the Liang Bua Stegodon is S. florensis, and it can be classified as a distinct chrono-subspecies, S. florensis insularis, which is characterized by a 30% linear size reduction and more advanced molar ridge formula compared to its ancestor S. florensis florensis from the Middle Pleistocene Soa Basin localities (van den Bergh et al., in press). The excavations in Liang Bua also yielded a large series of micro vertebrate fossils from the Late Pleistocene and Holocene, mostly murids, but also shrews (van den Hoek Ostende et al., 2006) and bats. Whereas currently only one species of giant rat, Papagomys armandvillei, survives on Flores, it was already apparent from excavations at another cave, Liang Toge, that several large murid species were present during the the Holocene. Hooijer (1957b) had described from the cave Spelaeomys florensis and Papagomys verhoeveni, as well as a subspecies of the recent Papagomys, P. a. besar. Musser (1981) subscribed Hooijer’s conclusion that a second, smaller species of Papagomys was present at Liang Toge, but demonstrated that the holotype of P. verhoeveni was referable to P. armandvillei. A new type was designated and the smaller species was named P. theodorverhoeveni. The three giant murids from Liang Toge have all been found at Liang Bua. In addition, a third, larger species of Papagomys appears to be present in the Liang Bua fauna. The Flores murid fauna also clearly shows that smaller species remain present next to the giants. Musser (1981) described middle-sized material from Liang Toge as Komodomys rintjanus and Floresomys naso, later changing the latter into Paulamys naso, as Floresomys proved to be occupied (Musser et al., 1986). The discovery of a recent (new or extant?) specimen of this species showed it to be in fact referable to the genus Bunomys (Kitchener et al., 1991a). Both Komodomys rintjanus and Bunomys naso are present in the extensive material from Liang Bua. Apart from these fossils a small-sized murid has been found, which had hitherto not been noted in literature. Part of this material is probably referable to R. hainaldi, an endemic rat from Flores with was discovered at the end of the last century (Kitchener et al., 1991b). It seems plausible that more than three taxa of small and middle-sized murids are present in the Liang Bua material, which is still awaiting taxonomical study. 4.2.3. Timor, Sumba, and Sangihe On Timor Verhoeven (1964) discovered the first remains of a dwarf Stegodon at Weaiwe. The posterior part of an M3 of this pygmy Stegodon timorensis was described and figured by Sartono (1969). More molar material from this Timor 334 vos, hoek ostende, and bergh species was described and figured by Hooijer (1969), who also reported a largesized Stegodon from the island, represented by a slightly worn DM3 from Sadilaun. Hooijer classified this fossil as S. timorensis subspecies D. More material of both the dwarf and the large-sized Stegodon was collected during a 1970 expedition in which Hooijer participated. Hooijer (1972b) described molar remains as well as postcranial elements obtained during this expedition. Sartono and Marino (1978) described additional material of the dwarf species from Timor. Sartono (1979) announced the discovery of a pygmy Stegodon from Sumba. Unfortunately, the stratigraphic context is not clear. The small island of Sangihe is located between the northern tip of Sulawesi and the Mindanao (the Philippines). In 1989, Dr. F. Aziz and Dr. Shibazaki collected some stegodont material from the island, including an upper tusk and some molar remains. All material originates from the Pintareng Formation on the southeast of Sangihe Island. The age of this formation is thought to be Pleistocene. The material was briefly described and figured by Aziz (1990) and attributed to S. sp. B cf. trigonocephalus. 4.3. The Philippines From the islands of the Philippines only a few mammal fossils are found and described (de Vos and Bautista, 2003). Based on the size and morphology of the molars there is only one species of Stegodon, namely Stegodon luzonensis and a large species Elephas. Stegodon luzonensis is a little smaller than the continental form. Postcranial elements of the proboscideans show that there is a small and large proboscidean. Other than that, there is a relatively small rhino, Rhinoceros philippinensis and material of a giant tortoise (de Vos and Bautista, 2003). This fauna composition indicates an endemic island fauna. The stratigraphic position of the various fossils is not well known. Archaeological data (Fox and Paralta, 1974) suggest that the faunal remains are of Mid-Pleistocene age (following Von Koenigswald in Durkee and Pederson, 1961, p. 160) and that at least some of the tools found in the Philippines are coeval with the fossils (in: Wasson and Cochrane, 1979). Possibly this is the same succession as in Flores, that there first was a pygmy proboscidean and a giant tortoise, followed by a large proboscidean and artefacts. 5. Observations and Remarks This overview of the results of the research on palaeo-isles in the latter half of the 20th century and beyond, allows us to make the following observations and remarks. Island evolution follows distinct patterns. A characteristic of island faunas is their unbalanced nature. They consist mainly of herbivores, like elephants, hippos, and cervids. Carnivores are absent, although both on Gargano and on Crete patterns in insular evolution of mammals 335 an otter has been found; presumably their semi-aquatic lifestyle enabled them to reach the islands, whereas other mammal predators could not. Perissodactyles, too, seem to be absent as a rule. Horses are absent in the Pleistocene faunas from Java, showing that even filter dispersal provides a barrier. Nevertheless, on Luzon a rhinoceros, was found, and fossil rhino’s are also known from Japan. Possibly the gases in their gut provided enough floating capacity to reach the island via sweepstake dispersal. The most common elements on the Quaternary islands are proboscideans, hippopotamuses, and cervids. Whether such a selectivity for certain families exists among smaller mammals is not clear. The island forms we know mainly belong to glirids (Leithia, Stertomys) and murids (Mikrotia, Kritimys, Hooijeromys), but it can easily be argued that these were common elements on the mainland at the time of colonization and that their arrival on islands, by, for example, rafting on driftwood, was simply a matter of chance. This also holds true for Deinogalerix, which is probably derived from Parasorex, a very common insectivore in the Late Miocene of Europe (van den Hoek Ostende, 2001b). Apart from a preference for certain taxa, the telltale signs of an island fauna are dwarfism for the larger and gigantism among the smaller mammals. However, these phenomena are not necessarily clear-cut. Extreme dwarfism occurs, such as in the Sicilian E. falconeri. On the other hand, the Elephas found in the Middle and Late Pleistocene of Crete is only a little smaller than the mainland E. antiquus. In Wallacea and the Philippines, too, the Middle and Late Pleistocene proboscideans are usually somewhat smaller than the mainland forms, with the exception of the Late Pleistocene S. florensis insularis, which had an estimated body weight of between 350 and 900 kg. Other true pygmies occur on the Early Pleistocene islands of the region. The smallest pygmy Stegodon so far recorded, the Early Pleistocene S. sondaari, had an estimated body weight of between 200 and 500 kg (van den Bergh, 1999). That dwarfism among proboscideans is not restricted to the Mediterranean or Wallacea, is clear by the finds of pygmy mammoths on the Channel Islands of California (Agenbroad, 2003 and references therein). Not all large mammals will reduce in size, nor will all small mammals obtain large proportions. The niches for small rodents are still available, as is clear from the large number of relatively unchanged rodents, insectivores and lagomorphs on Gargano. Also, on Flores, in Late Pleistocene sedimentary rocks in the Liang Bua cave, we find, apart from the giant murids, Papagomys armandvillei, P. theodorverhoeveni, Papagomys sp. and Spelaeomys florensis (Musser, 1981) at least three small to middle-sized rats. Another example of the absence of gigantism in smaller mammals is the Early Miocene fauna from Oschiri Road Cut on Sardinia (de Bruijn and Rümke, 1974). The largest glirid, Glis major, is of respectable dimensions, but certainly not a giant as Leithia or Stertomys. The Oschiri assemblage was identified as an insular fauna on the basis of its unbalanced composition and endemic forms as the mole Nuragha, but not on the presence of giants. The specific characteristics of island faunas help us to determine whether or not an area was an island in the past. This is, for instance, clear from the Tertiary mammals of the Greek archipelago. Since they have a similar composition as the 336 vos, hoek ostende, and bergh contemporaneous continental faunas, it is clear these islands were at the time still connected to the mainland. However, one has to bear in mind that if an island is sufficiently large, or if its time of isolation is limited, no island fauna will develop. This holds true, for example for the large islands of the Sunda Shelf. The only exception is the Early Pleistocene Satir fauna on Java, but probably large parts of present-day Java were still submerged when this island fauna developed (Bergh et al., 1996a). Whether or not an island fauna developed on Borneo, as indicated by Stegolophodon latidens, is not yet clear. The islands of the Mediterranean and Wallacea do not only show striking similarities in the pattern of evolution, but also in the timing of certain events. Both on Crete and on, for example, Flores, a faunal turnover took place near the Early to Middle Pleistocene transition. This is also the period of a major faunal change on the continent in Eurasia, as the Villafranchian faunas gave way to the mammals of the Galerian. Since the turnover in insular faunas occurred on different sides of the globe, the explanation must be sought in a global cause. Probably this is related to the more extreme climatic fluctuations of the Middle Pleistocene. Glacial periods did not only cause major changes in the ecosystem on the continent, but also a considerable drop in sea level. Thus islands became more readily accessible and new faunal elements could reach the islands, either actively replacing existing faunas or simply filling up the empty space left after extinction. Increased competition between large-sized island herbivore species like proboscideans and intermediate-sized herbivores could be held responsible for the reduced degree of dwarfing seen in some of the island faunas (Raia and Meiri, 2006). With the niches of intermediate-sized herbivores already occupied, megaherbivores like proboscideans seem to dwarf to a lesser degree than when these intermediate-sized species would be missing. The total lack of predators may further enhance dwarfing of large herbivores to a minimal size with an optimal energetic balance (Palombo, 2007). In the case of Flores this lead to the colonization of the island by H. erectus (Bergh et al., 1996b; Sondaar et al., 1996; Bergh, 1999), ultimately leading to the development of H. floresiensis (Brown et al., 2004), showing that Man too could be affected by the patterns of insular evolution. The presence of a large predator (H. floresiensis) could be the reason why Stegodon florensis did not became as small as Stegodon sondaari. Acknowledgements The completion of this paper has been a long and bumpy road. Different versions of the manuscript were read and commented upon by various colleagues. We thank Andy Currant, Steve Donovan, Don MacFarlane, Fiona Fearnhead, Rienk de Jong, and Leo Kriegsman for their textual and conceptual comments. We are grateful to the editor Willem Renema for his patience and help in seeing this project through. patterns in insular evolution of mammals 337 References Abbazzi, L., Benvenuti, M., Boschian, G., Dominci, S., Masini, F., Mezzabotta, C., Piccini, L., Rook, L., Valleri, G. and Torre, D., 1996, Revision of the Neogene and Pleistocene of the Gargano region (Apulia, Italy), the marine and continental successions and the mammal fauna assemblages in an area between Apricena and Poggio Imperiale (Foggia), Memorie della Societa Geologica Italiana 51 (1): 383–402. Agenbroad, L.D., 2003, New absolute dates and comparisons for California’s Mammuthus exilis, in: Reumer, J.W.F., de Vos, J., and Mol, D. (eds), Advances in Mammoth Research. Proceedings of the Second International Mammoth Conference, Rotterdam, May 16–20 1999, Deinsea 9: 1–16. d’Allesandro, A., Laviano, A., Ricchetti, G., and Sardella, A., 1979, Il Neogene del monte Gargano, Bolletino della Società Paleontologia Italiana 18: 9–116. Alvarez-Sierra, M.A., Daams, R. and van der Meulen, A.J., 1987, The mammals from the lower Miocene of Aliveri (Island of Evia, Greece); VII, The Eomyidae, Proceedings Koninklijke Nederlandse Akademie van Wetenschappen, Series B 90 (1): 47–56. Athanassiou, A., 1996, Contribution to the study of the fossil mammals of Thessaly, PhD thesis, University of Athens (in Greek). Aziz, F., 1990, Pleistocene mammal faunas of Sulawesi and their bearings to paleozoogeography, PhD thesis, Kyoto University. Badoux, D.M., 1959, Fossil mammals from two deposits at Punung (Java), PhD. thesis, Utrecht University. Bate, D.M.A., 1904, Further note on the remains of Elephas Cypriotes from a cave-deposit in Cyprus. Philosophical Transactions of the Royal Society London B 197: 347–360. Bate, D.M.A., 1905, Four and a half months in Crete in search of Pleistocene mammal remains. Geological Magazine Series 5 2: 193–202. Bate, D.M.A., 1907, On elephant remains from Crete, with description of Elephas creticus, sp. n. Proceedings of the Zoological Society London 1907: 238–250. Bate, D.M.A., 1909, Preliminary note on a new artiodactyle from Majorca, Myotragus balearicus, gen. et sp. nov. Geological Magazine Series 5 6: 385–388. Benda, L., Hiltermans, H., Kuss, S.E., and Symeonides, N.K., 1968, Der erste Mastodon Fund der insel Kreta, Annales Géologiques des Pays Helléniques 12: 167–177. Benda, L., Meulenkamp, J.E. and Zachariasse, W.J., 1974, Biostratigraphic correlations in the eastern Mediterranean Neogene: 1. Correlation between planktonic foraminiferal, uvigerinid, sporomorphal and mammal zonations of the Cretan and Italian Neogene, Newsletter of Stratigraphy 3: 205–217. Bernor, R., Solounias, N., Swisher, C., III, and Van Couvering, J.A., 1996, The correlation of three classical ‘Pikermian’ mammal faunas – Marageh, Samos, and Pikermi – with the European MN Unit System, in: Bernor, R., Fahlbusch, V., and Mittmann, H. (eds), The Evolution of Western Eurasian Neogene Mammal Faunas, Columbia University Press, New York, pp. 137–156. Besenecker, H. and Symeonidis, N.K., 1974, Der erste Säugetierfund aus dem Neogen der Griechischen Insel Psara (Ostägäis), Annales de Géologie des Pays Helleniques 46: 647–652. Bover, P. and Alcover, J.A., 2003, Understanding Late Quaternary extinctions: the case of Myotragus balearicus (Bate, 1909), Journal of Biogeography 30 (5): 771–781. Brown, P., Sutikna, T., Morwood, M.J., Soejono, R.P., Jatmiko, Wayhu Saptomo, E., and Rokus Awe Due, 2004, A new small-bodied hominin from the Late Pleistocene of Flores, Indonesia, Nature 431: 1055–1061. 338 vos, hoek ostende, and bergh Brinkerink, H.P., 1996, Tortoises and turtles from the Pleistocene of Crete, in: Reese, D.E. (ed.), Pleistocene and Holocene Fauna of Crete and Its First Settlers, Monographs in World Archaeology 28: 207–209. Brumm, A., Aziz, F., van den Bergh, G.D., Morwood, M.J., Moore, M.W., Kurniawan, I., Hobbs, D.R., and Fullagar, R., 2006, Early stone technology on Flores and its implications for Homo floresiensis, Nature, 441: 624–628. Butler, M., 1980, The giant erinaceid insectivore, Deinogalerix Freudenthal, from the upper Miocene of Gargano, Italy, Scripta Geologica 57: 1–72. Capasso Barbato, L.C. and Petronio, C., 1986, Cervus major n. sp. of Bate Cave (Rethymnon, Crete), Atti della Accademia Nazionale dei Lincei, Memorie Lincee, Scienze Fisische e Naturali, Serie VIII 18 (2): 59–100. de Bonis, L., Koufos, G.D., and Sen, S., 1997a, A giraffid from the Middle Miocene of the island Chios, Greece, Palaeontology 40 (1): 121 –123. de Bonis, L., Koufos, G.D., and Sen, S., 1997b, The Sanitheres (Mammalia, Suoidea) from the middle Miocene the island Chios, Greece, Revue de Paléobiologie 16 (1): 259–270. de Bonis, L., Koufos, G.D., and Sen, S., 1998. Ruminants (Bovidae and Tragulidae) from the Middle Miocene (MN 5) of the island Chios, Aegean sea (Greece), Neues Jahrbuch für Geologie und Paläontologie, Abhandlungen 210 (3): 399–420. de Bruijn, H., Dawson, M.R., and Mein, P., 1970, Upper Pliocene Rodentia, Lagomorpha and Insectivora (Mammalia) from the island of Rhodes (Greece), I, II and III, Proceedings Koninklijke Nederlandse Akademie van Wetenschappen B 73 (5): 535–584. de Bruijn, H. and Meulenkamp, J.E., 1972, Late Miocene rodents from the Pandanassa Formation (Prov. Rethymnon), Crete, Greece, Proceedings Koninklijke Nederlandse Akademie van Wetenschappen B 75 (1): 54–60. de Bruijn, H. and Rümke, C.G., 1974, On a peculiar mammalian association from the Miocene of Oschiri (Sardinia). I and II, Proceedings Koninklijke Nederlandse Akademie van Wetenschappen B 77 (1): 44–79. de Bruijn, H., Meulen, van der A.J., and Katsikatsos, G.C., 1980, The mammals from the lower Miocene of Aliveri (Island of Evia, Greece); Part 1, The Sciuridae, Proceedings Koninklijke Nederlandse Akademie van Wetenschappen, B 83 (3): 241–261. de Bruijn, H. and Saraç, G., 1991, Early Miocene rodent faunas from the eastern Mediterranean area Part I. The genus Eumyarion. Proceedings Koninklijke Nederlandse Akademie van Wetenschappen 94 (1): 1–36. de Giuli, C., Masini, F., and Torre, D., 1985, Effetto arcipelago: un esempio nelle fauna fossili del Gargano, Bolletino della Società Paleontologia Italiana 24: 191–193. de Giuli, C., Masini, F., Torre, D., and Boddi, V., 1986a, Endemism and biochronological reconstructions: the Gargano case history, Bolletino della Società Paleontologia Italiana 25: 267–276. de Giuli, C., Masini, F., Torre, D., and Valleri, G., 1986b, Mammal migration events in emerged areas of the Apulian platform during the Neogene, Giornale di Geologia, Serie 3 48 (1/2): 145–162. de Giuli, C., Masini, F., Torre, D., and Valleri, G., 1987a, Paleogeography and mammal faunas in the Apulo-Dalmatic area, Annals of the Hungarian Geological Institute 70: 471–476. de Giuli, C., Masini, F., and Valleri, G., 1987b, Paleogeographic evolution of the Adriatic area since Oligocene to Pleistocene, Rivista Italiana di Paleontologia e Stratigrafia 93: 109–126. Daams, R. and Freudenthal, M., 1985, Stertomys laticrestatus, a new glirid (dormice, Rodentia) from the insular fauna of Gargano (Prov. of Foggia, Italy), Scripta Geologica 77: 21–27. patterns in insular evolution of mammals 339 Dermitzakis, M.D. and Sondaar, P.Y., 1978, The importance of fossil mammals in reconstruction paleogeography with special reference to the Pleistocene Aegean Archipelago, Annales Géologiques des Pays Helléniques 46: 808–840. Dexter, R.W., 1978, Some historical notes on Louis Agassiz’s lectures on zoogeography. Journal of Biogeography 5: 207–209. Doukas, C.S., 1986, The mammals from the lower Miocene of Aliveri (Island of Evia, Greece); V, Insectivores, Proceedings Koninklijke Nederlandse Akademie van Wetenschappen B 89 (1): 15–38. Doukas, C.S. and Athanassiou, A., 2003, Review of the Pliocene and Pleistocene Proboscidea (Mammalia) from Greece, in: Reumer, J.W.F., de Vos, J., and Mol, D. (eds), Advances in Mammoth Research (Proceedings of the Second International Mammoth Conference, Rotterdam, May 16–20 1999), Deinsea 9: 97–110. Drooger, C.W. and Meulenkamp, J.E., 1973, Stratigraphic contribution to geodynamics in the Mediterranean area; Crete as a case history, Bulletin of the Geological Society of Greece 10: 193–200. Dubois, E., 1889, Over de wenschelijkheid van een onderzoek naar diluviale fauna van Ned. Indië, in het bijzonder van Sumatra, Natuurkundig Tijdschrift voor Nederlandsch-Indië 48: 148–165. Durkee, E.F. and Pederson, S.L., 1961, Geology of Northern Luzon, Philippines, American Association of Petrologists Geological Bulletin 45: 137–168. Foster, J.B., 1964, Evolution of mammals on islands. Nature 202: 234–235. Fox, R.B. and Peralta, J.T., 1974, Preliminary report on the Palaeolithic archaeology of Cagayan Valley, Philippines and the Cabalwanian Industry, in: Proceedings of the first Regional Seminar on South East Asian Prehistory and archseology, National Museum of the Philippines, Manila, pp. 100–147. Freudenthal, M., 1971, Neogene vertebrates from the Gargano Peninsula, Italy, Scripta Geologica 3: 1–10. Freudenthal, M., 1972, Deinogalerix koenigswaldi nov. gen., nov. spec., a giant insectivore from the Neogene of Italy, Scripta Geologica 14: 1–19. Freudenthal, M., 1976, Rodent stratigraphy of some Miocene fissure fillings in Gargano (prov. Foggia, Italy), Scripta Geologica 37: 1–23. Freudenthal, M., 1985, Cricetidae (Rodentia) from the Neogene of Gargano (Prov. of Foggia, Italy), Scripta Geologica 77: 29–76. Freudenthal, M., 2006, Mikrotia nomen novum for Microtia Freudenthal, 1976 (Mammalia, Rodentia), Journal of Vertebrate Paleontology 26 (3): 748. Galobart, A., Pons-Moyà, J., Anton, M., and Maroto, J., 2003, Descripción del material de Homotherium latidens (Owen) de los yacimientos del Pleistocene inferior de Incarcal (Girona, NE de la Península Ibérica), Paleontologia i Evolucío 34: 99–141. van den Bergh, G.D., de Vos, J., Sondaar, P.Y., and Aziz, F., 1996a, Pleistocene zoogeographic evolution of Java (Indonesia) and glacio-eustatic sea-level fluctuations: a background for the presence of Homo. Indo-pacific Prehistory Association Bulletin (Chiang Mai Papers), 14: 7–21. van den Bergh, G.D., Mubroto, B., Sondaar, P.Y., and de Vos, J., 1996b, Did Homo erectus reach the island of Flores? Indo-pacific Prehistory Association Bulletin (Chiang Mai Papers), 14: 27–36. van den Bergh, G.D., 1999, The Late Neogene elephantoid-bearing faunas of Indonesia and their palaeozoogeographic implications; A study of the terrestrial faunal succession of Sulawesi, Flores and Java, including evidence for early hominid dispersal east of Wallace’s line, Scripta Geologica 117: 1–419. 340 vos, hoek ostende, and bergh van den Bergh, G.D., de Vos, J., and Sondaar P.Y., 2001, The Late Quaternary palaeogeography of mammal evolution in the Indonesian Archipelago, Palaeogeography, Palaeoclimatology, Palaeoecology 171: 385–408. van den Bergh, G.D., Rokus, A.D., Morwood, M.J., Sutikna, T., Jatmiko, P., and Wahyu Saptomo, E., in press, The youngest Stegodon remains in south-east Asia from the Late Pleistocene archaeological site Liang Bua, Flores, Indonesia. Quaternary International. van den Hoek Ostende, L.W., 2001a, Insectivore faunas from the Lower Miocene of Anatolia – Part 8: stratigraphy, palaeoecology, palaeobiogeography, Scripta Geologica 122: 101–122. van den Hoek Ostende, L.W., 2001b, A revised generic classification of the Galericini (Insectivora, Mammalia) with some remarks on their palaeobiogeography and phylogeny, Geobios 34 (6): 681–695. van den Hoek Ostende, L.W., van den Bergh, G., and Rokus Awe Due, 2006, First fossil insectivores from Flores, Annales Geologiques de Pays Helleniques 41: 67–72. Hooijer, D.A., 1947, On fossil and prehistoric remains of Tapirus from Java, Sumatra and China, Zoologische Mededelingen 27: 253–299. Hooijer, D.A., 1948a, Evolutie van zoogdieren in het Quartair van Z.O.-Azie, Vakblad voor Biologen 28 (7): 117–122. Hooijer, D.A., 1948b, Pleistocene Vertebrates from Celebes. I. Celebochoerus heekereni nov. gen. nov. spec. Proceedings of the Koninklijke Nederlandse Akademie van Wetenschappen 51: 1024–1032. Hooijer, D.A., 1948c, Pleistocene vertebrates from Celebes. II. Testudo margae nov. spec. Proceedings of the Koninklijke Nederlandse Akademie van Wetenschappen 51: 1169–1182. Hooijer, D.A., 1949, Pleistocene vertebrates from Celebes IV. Archidiskodon celebensis nov. spec., Zoologische Mededelingen 30: 205–226. Hooijer, D.A., 1953a, Pleistocene vertebrates from Celebes V – lower molars of Archidiskodon celebensis Hooijer, Zoologische Mededelingen 31: 311–318. Hooijer, D.A., 1953b, Pleistocene vertebrates from Celebes VI – Stegodon sp., Zoologische Mededelingen 32: 107–112. Hooijer, D.A., 1953c, Pleistocene vertebrates from Celebes VII – milk molars and premolars of Archidiskodon celebensis Hooijer, Zoologische Mededelingen 32: 221–230. Hooijer, D.A., 1954a, Pleistocene vertebrates from Celebes XI – molars and a tusked mandible of Archidiskodon celebensis Hooijer, Zoologische Mededelingen 33: 103–120. Hooijer, D.A., 1954b, Pleistocene vertebrates from Celebes. X. Testudinata. Proceedings of the Koninklijke Nederlandse Akademie van Wetenschappen 57: 486–489. Hooijer, D.A., 1955, Fossil Proboscidea from the Malay Archipelago and India, Zoologische Verhandelingen 28: 1–146. Hooijer, D.A., 1957a, A Stegodon from Flores, Treubia 24: 119–129. Hooijer, D.A. 1957b, Three new giant prehistoric rats from Flores, Lesser Sunda Islands, Zoologische Mededelingen 35: 299–314. Hooijer, D.A., 1954, Pleistocene vertebrates from Celebes. VIII. Dentition and skeletton of Celebochoerus heekereni. Zoologische Verhandelingen, 24: 3–46. Hooijer, D.A., 1964, Pleistocene vertebrates from Celebes XII – Notes on pygmy Stegodonts, Zoologische Mededelingen 40: 37–44. Hooijer, D.A., 1969, The Stegodon from Timor, Proceedings Koninklijke Nederlandse Akademie van Wetenschappen, Amsterdam, Series B 72 (3): 201–210. Hooijer, D.A., 1972a, Pleistocene vertebrates from Celebes. XIV. Additions to the Archidiskodon – Celebochoerus fauna, Zoologische Mededelingen 46: 1–16. patterns in insular evolution of mammals 341 Hooijer, D.A., 1972b, Stegodon trigonocephalus florensis Hooijer and Stegodon timorensis Sartono from the Pleistocene of Flores and Timor, Proceedings Koninklijke Nederlandse Akademie van Wetenschappen, Series B 75 (1): 12–33. Kitchener, D.J., How, R.A., and Maharadatunkamsi, R.A., 1991a, Paulamys sp. cf. P. naso (Musser, 1981) (Rodentia: Muridae) from Flores Island, Nusa Tenggara, Indonesia – description from a modern specimen and a consideration of its phylogenetic affinities, Records of the Western Australian Museum 15 (1): 171–189. Kitchener, D.J., How, R.A., and Maharadatunkamsi 1991b, A new species of Rattus from the mountains of West Flores, Indonesia, Records of the Western Australian Museum 15 (3): 611–626. Klein Hofmeijer, G. and Bruijn, H. de, 1985, The mammals from the lower Miocene of Aliveri (Island of Evia, Greece); IV, The Spalacidae and Anomalomyidae, Proceedings Koninklijke Nederlandse Akademie van Wetenschappen, Series B 88 (2): 185–198. Kondopulou, D., Bonis, L. de, Koufos, G.D., and Sen, S., 1993, Palaeomagnetic data and biostratigraphy of the middle Miocene vertebrate locality of Thymiana (Chios island, Greece), Proceedings 2nd Congress Hellenic Geophysical Union, Florina, Greece 2–3: 626–635. Kostopoulos, S.D., 1996, The Plio-Pleistocene artiodactyls of Macedonia (Greece); systematics, palaeoecology, biochronology, biostratigraphy, PhD Thesis, University of Thessaloniki (in Greek). Koufos, G.D. and Kostopoulos, D.S., 1997, Biochronology and succession of the Plio-Pleistocene macromammalian localities of Greece, Mémoires et Travaux EPHE, Inst. Montpellier 21: 619–634. Koufos, G.D., de Bonis, L., and Sen, S., 1995, Lophocyon paraskevaidisi, a new viverrid (Carnivora, Mammalia) from the middle Miocene of Chios island (Aegean sea, Greece), Géobios 28: 511–523. Koufos, G.D., Kostopoulos, S.D., and Kaliadimou, K.K., 1991, Un nouveau gisement de mammifères dans le Villafranchien de Macédoine occidentale (Grèze), Comptes Rendus de l’Académie des Sciences, Paris Série II 313: 831–836. Koufos, G.D., Zouros, N., and Mourouzidou, O., 2003, Prodeinotherium bavaricum (Proboscidea, mammalia) from Lesvos island, Greece; the appearance of deinotheres in the Eastern Mediterranean, Geobios 36: 305–315. Kuss, S.E., 1965, Eine pleistozäne Säugetierfauna der Insel Kreta, Berichte der Naturforschenden Gesellschaft zu Freiburg im Breisgau 55 (2): 271–348. Lehmann, U. and Tobien, H., 1995, Artiodaktyla-Fossilien (Mammalia) aus dem Miozän von Thymiana, C., Annales de Géologie des Pays Helleniques 60: 403–414. Leinders, J.J.M., 1984, Hoplitomerycidae fam. nov. (Ruminantia, Mammalia) from Neogene fissure fillings in Gargano (Italy), Scripta Geologica 70: 1–68. Lomolino, M.V., 2005, Body size evolution in insular vertebrates: generality of the island rule. Journal of Biogeography 32: 1683–1699. Lomolino, M.V., Riddle, B.R., and Brown, J.H., 2006, Biogeography, 3rd edition. Sindauer Associates, Sunderland, MA. Lopez-Martinez, N., 1986, The mammals from the lower Miocene of Aliveri (Island of Evia, Greece); VI. The ochotonid lagomorph Albertona balkanica n.gen. n.sp. and its relationships, Proceedings Koninklijke Nederlandse Akademie van Wetenschappen, Series B 89 (2): 177–194. Lydekker, R., 1885, Description of a tooth of Mastodon latidens Clift, from Borneo, Proceedings Zoological Society, London, pp. 777–779. 342 vos, hoek ostende, and bergh MacArthur, R.H. and Wilson, E.O., 1963, An equilibrium theory of insular zoogeography. Evolution 17: 373–387. MacArthur, R.H. and Wilson, E.O., 1967, The Theory of Island Biogeography, Monographs in Population Biology, Vol. 1, Princeton University Press, Princeton, NJ. van der Made, J., 1996, Pre-Pleistocene land mammals from Crete, in: Reese, D.S. (ed.), Pleistocene and Holocene Fauna of Crete and Its First Settlers, Monographs in World Archaeology 28: 69–79. Maringer, J. and Verhoeven, T.H., 1970, Die Steinartefacte aus der Stegodon-Fossil-Schicht von Mengeruda auf Flores, Indonesien, Anthropos 65: 229–247. Mayhew, D.F., 1996, The extinct murids of Crete, in: Reese, D.S. (ed.), Pleistocene and Holocene Fauna of Crete and Its First Settlers, Monographs in World Archaeology 28: 167–171. Medway, L., 1979, The Niah excavations and an assessment of the impact of early man on ammals in Borneo, Asian Perspectives 20: 51–69. Meiri, S., Dayan, T., and Sinberloff, D., 2006, The generality of the island rule re-examined. Biogeography 33: 1571–1577. Melentis, J.K. and Tobien, H., 1968, Paläontologische Ausgrabungen auf der Insel Chios (Eine vorläufige Mitteilung), Annales de Géologie des Pays Helleniques 19: 647–652. van der Meulen, A.J. and de Bruijn, H., 1982, The mammals from the lower Miocene of Aliveri (Island of Evia, Greece); II, The Gliridae, Proceedings Koninklijke Nederlandse Akademie van Wetenschappen, Series B 85 (4): 485–524. van der Meulen, A.J. and van Kolfschoten, T., 1986, Review of the Late Turolian to Early Biharian mammal faunas from Greece and Turkey, Memoria Società Geologia Italiana 31: 201–211. Meulenkamp, J.E., 1971, The Neogene in the Southern Aegean Area, in: Strid, A. (ed.), Evolution in the Aegean, Opera Botanica 30: 5–12. Mol, D., de Vos, J., van den Bergh, G.D., and Sondaar, P.Y., 1996, The taxonomy and ancestry of the fossil elephants of Crete: faunal turnover and a comparison with Proboscidean faunas of Indonesian islands, in: Reese, D.S. (ed.), Pleistocene and Holocene Fauna of Crete and Its First Settlers, Monographs in World Archaeology 28: 81–98. Morwood, M.J., Brown, P., Jatmiko, S.T., Wahyu, S.E., Westaway, K.E., Rokus, A.D., Roberts, R.G., Maeda, T., Wasisito, S., and Djubiantono, T., 2005, Further evidence for small-bodied hominins from the Late Pleistocene of Flores, Indonesia, Nature 437: 1012–1017. Morwood, M.J., O’Sullivan, P.B., Aziz, F., and Raza, A., 1998, Fission-track ages of stone tools and fossils on the east Indonesian island of Flores, Nature 392: 173–176. Morwood, M.J., Soejono, R.P., Roberts, R.G., Sutikna, T., Turney, C.S.M., Westaway, K.E., Rink, W.J., Zhao, J.-X, van den bergh, G.D., Rokus, A.D., Hobbs, D.R., Moore, M.W., Bird, M.I., and Fifield, L.K., 2004, Archaeology and age of a new hominin from Flores eastern Indonesia, Nature 431: 1087–1091. Musser, G.G., 1981, The giant rats of Flores and its relatives east of Borneo and Bali, Bulletin of the American Museum of Natural History 169 (2): 69–175. Musser, G.G., van der Weerd, A., and Strasser, E., 1986, Paulamys, a replacement name for Floresomys Musser, 1981 (Muridae), and new material of that taxon from Flores, Indonesia, American Museum Novitates, 2850: 1–10. Palombo, M.R., in press, How can endemic proboscideans help us understand the “Island Rule”? A case study of Mediterranean islands. Quaternary International. Paraskevaidis, I., 1940, Eine obermiocäne fauna von Chios, Neues Jahrbuch für Mineralogie, Geologie und Paläontologie, B 83: 363–442. patterns in insular evolution of mammals 343 Poulakakis, N., Mylonas, M., Lymberakis, P., and Fassoulas, C., 2002, Origin and taxonomy of the fossil elephants of the island of Crete (Greece): problems and perspectives, Palaeogeography, Palaeoclimatology, Palaeoecology 186 (1–2): 163–183. Radulesco, C. and Samson, P., 1991, Traces d’activité humaine a la limite Pliocène/ Pléistocène dans le Bassin Dacique (Roumainie), in: Bonifay, E. and Vandermeersch, V. (eds), Les Premiers Européens, Paris, Éditions du C.T.H.S.: 203–207. Raia, P. and Meiri, S., 2006, The island rule in large mammals: paleontology meets ecology. Evolution 60 (8): 1731–1742. Reese, D.S. (ed.), 1996, Pleistocene and Holocene fauna of Crete and its first Settlers, Monographs in World Archaeology 28: 207–210. Reumer, J.W.F. and Payne, S., 1986, Notes on the Soricidae (Insectivora, Mammalia) from Crete, II, The shrew remains from Minoan and Classical Kommos, Bonner zoologische Beiträge 37 (3): 173–182. Reumer, J.W.F., 1996, Shrews (Soricidae) on islands, with special reference to Crocidura zimmermanni from Crete, in: Reese, D.S. (ed.), Pleistocene and Holocene Fauna of Crete and Its First Settlers, Monographs in World Archaeology 28: 173–179. Rothausen, K., 1977, Die mittelmiozänen Wirbeltierfundstellen südlich Thymiana (Insel Chios, Ägäis, Griechenland), 2 Teil: Geologie: Die Fundstellen und ihre Abfolge, Annales de Géologie des Pays Helleniques 28: 495–515. Sartono, S., 1969, Stegodon timorensis: a pygmy species from Timor (Indonesia), Proceedings Koninklijke Nederlandse Akademie van Wetenschappen, Series B 72: 192–202. Sartono, S. and Marino R., 1978, A mandibula and a maxilla of Stegodon timorensis. Modern Quaternary Research in SE Asia, 4, 41–50. Sartono, S., 1979, The discovery of a pygmy Stegodon from Sumba, East Indonesia: an announcement, Modern Quaternary Research in SE Asia 5: 57–63. Schmidt-Kittler, N., 1983, The mammals from the lower Miocene of Aliveri (Island of Evia, Greece); III, On a new species of Sivanasua Pilgrim, 1931 (Feliformia, Carnivora) and the phylogenetic position of this genus, Proceedings Koninklijke Nederlandse Akademie van Wetenschappen, Series B 86 (3): 301–318. Sen, S., 1977, Megapedetes aegaeus n. sp. (Pedetidae) et à propos d’autres “rongeurs africains” dans le Miocene d’Anatolie, Geobios 10 (6): 983–986. Sen, S., Sondaar, P.Y., and Staesche, U., 1978, The biostratigraphical applications of the genus Hipparion with special references to the Turkish representatives, Proceedings Koninklijke Nederlandse Akademie van Wetenschappen, Series B 81: 370–385. Simpson, G.G., 1940, Mammals and land bridges. Journal of the Washington Acadamy of Science 30: 137–163. Sondaar, P.Y., 1976, The island sweepstakes. Why did pygmy elephants, dwarf deer and large mice populate the Mediterranean? Natural History 95 (9): 50–57. Sondaar, P.Y., 1977, Insularity and its effect on mammal evolution, in: Hecht, M.N., Goody, P.C., and Hecht, B.M. (eds), Major Patterns in Vertebrate Evolution, Plenum, New York, pp. 671–707. Sondaar, P.Y., 1987, Pleistocene man and extinctions of island endemics. Memoirs du Societe Geologique de France, N.S. 150: 159–65. Sondaar, P.Y. and Boekschoten, G.J., 1967, Quaternary mammals in the south Aegean island arc with notes on other mammals from the coastal regions of the Mediterranean, I/II, Proceedings Koninklijke Nederlandse Akademie van Wetenschappen, Series B 7: 556–576. Sondaar, P.Y., de Vos, J., and Dermitzakis, M.D., 1986, Late Cenozoic faunal evolution and Palaeogeography of the South Aegean Island arc, Modern Geology 10: 249–259. 344 vos, hoek ostende, and bergh Sondaar, P.Y., Dermitzakis, M.D., and Vos, J. de, 1996, The Paleogeography and faunal Evolution of the Land Mammals of Crete, in: Reese, D.S. (ed.), Pleistocene and Holocene Fauna of Crete and Its First Settlers, Monographs in World Archaeology 28: 61–67. Sondaar, P.Y., van den Bergh, G.D., Mubroto, B., Aziz, F., de Vos J., and Batu Unkap L., 1994, Middle Pleistocene faunal turnover and colonization of Flores (Indonesia) by Homo erectus, Comptes rendus de l’Academie des Sciences, Série II 320: 1255–1262. Spaan, A., 1996, Hippopotamus creutzburgi: the case of the Cretan Hippopotamus, in: Reese, D.S. (ed.), Pleistocene and Holocene Fauna of Crete and Its First Settlers, Monographs in World Archaeology 28: 99–110. Spassov, N., 1998, A new late Villafranchian locality of vertebrate fauna, Slivnitsa (Bulgaria) and the carnivore dispersal events in Europe on the Pliocene/Pleistocene boundary, Historia Naturalis Bulgarica 9: 101–113. Storm, P., 2001, The evolution of humans in Australia from an environmental perspective, Palaeogeography, Palaeoclimatology, Palaeoecology 171: 363–383. Storm, P., Aziz, F., de Vos, J., Kosasih, D., Baskoro, S., Ngaliman and van den Hoek Ostende, L.W., 2005, Late Pleistocene Homo sapiens in a tropical rainforest fauna in East Java, Journal of Human Evolution 49: 536–545. Suzuki, M., Wikarno, B., Saefudin, I., and Itihara, M., 1985, Fission track ages of Pumice tuff, tuff layers and Javites of Hominid fossil bearing formations in Sangiran area, Central Java, in: Watanabe, N. and Kadar, D. (eds), Quaternary Geology of the Hominid Fossil Bearing Formations in Java, Geological Research and Development Centre, Special Publication 4: 309–331. Symeonides, N.K., Theodorou, G.E., and Giannopoulos, V.I., 2001, New data on Elephas chaniensis (Vamos cave, Chania, Crete), Proceedings of the 1st International Congress “The World of Elephants” Rome, 2001, pp. 510–513. Thenius, E., 1956, Die Suiden und Tayassuiden des seirischen Tertiärs, Sitzungsberichten Österreichische Akademie der Wissenschaften, Mathematisch-naturwissenschaflicht Klasse, Abteilung I 165 (4): 337–382. Theodorou, G., Spjeldnaes, N., Hanken, N.M., Lauritzen, S.E., Velitzelos, E., Athanassiou, A., and Roussiakis, S., 2000, Description and taphonomic investigations of Neogene Proboscidea from Rhodos, Greece, Annales Géologiques des Pays Helléniques 38 (C): 133–156. Tobien, H., 1969, Wirbeltiergrabungen im Miozän der Insel Chios (Ägäis), Praktika Akademie Athen 43: 151–157. Tobien, H., 1980, A note on the skull and mandible of a new Choerolophodont Mastodont (Proboscidea, Mammalia from the Middle Miocene of Chios (Aegean Sea, Greece), in: Jacobs, L.L. (ed.), Aspects of Vertebrate History, Essays in Honor of E.H. Colbert, pp. 299–307. Valleri, G., 1984, New data on planktonic foraminifera biostratigraphy from the Neogene of the Gargano Peninsula (Foggia, southern Italy), Rivista Italiana di Paleontologia e Stratigrafia 90 (3): 375–406. Van Valen, L., 1973, Body size and the number of plants and animals. Evolution 27: 27–35. Verhoeven, Th., 1964, Stegodon-Fossilien auf der Insel Timor, Anthropos 59: 634. de Vos, J., 1983, The Pongo faunas from Java and Sumatra and their significance for biostratigraphical and paleo-ecological interpretations, Proceedings Koninklijke Nederlandse Akademie van Wetenschappen, Amsterdam, Series B 86 (4): 417–425. de Vos, J., 1984a, The endemic Pleistocene deer of Crete, Verhandeling der Koninklijke Nederlandse Akademie van Wetenschappen, Afdeling Natuurkunde, Eerste reeks 31: 1–100. patterns in insular evolution of mammals 345 de Vos, J., 1984b, Reconsideration of Pleistocene cave faunas from South China and their relation to the faunas from Java, Courier Forschungsinstitut Senckenberg 69: 259–266. de Vos, J., 1996a, Taxonomy, Ancestry and speciation of the endemic Pleistocene deer of Crete compared with the taxonomy, ancestry and speciation of Darwin’s Finches, in: Reese, D.S. (ed.), Pleistocene and Holocene Fauna of Crete and Its First Settlers, Monographs in World Archaeology 28: 111–124. de Vos, J., 1996b, Faunal turnovers in Java in relation to faunas of the continent, Odontologie, Association for comparative Odontology 1: 32–36. de Vos, J., 2000, Pleistocene deer fauna in Crete: its adaptive radiation and extinction, Tropics (special issue) 10 (1): 125–134. de Vos, J., Aziz, F., Sondaar, P.Y., and van den Bergh, G.D., 1999, Homo erectus in S.E. Asia: time space and migration routes; a global model III. Migration routes and evolution, in: Gibert, J. Sánchez, F., Gibert, L., and Ribot, F. (eds), The Hominids and Their Environment During the Lower and Middle Pleistocene of Eurasia, Proceedings of the International Conference of Human Paleontology, Orce 1995, pp. 369–383. de Vos, J. and Bautista, A., 2003, Preliminary notes on the vertebrate fossils from the Philippines, Semantic and Systematics: Philippine Archaeology, Proceedings of the Society of Philippine Archaeologists 1: 42–62. de Vos, J., Sartono, S., Hardja-Sasmita, H.S., and Sondaar, P.Y., 1982, The fauna from Trinil, type locality of Homo erectus; a reinterpretation, Geologie en Mijnbouw, 61: 207–211. de Vos, J. and Vu the Long, 2001, First settlements: relations between continental and Insular Southeast Asia, in: Sémah, F., Falguères, C., Grimaud-Hervé, D., and Sémah, A.-S. (eds), Origine des peuplements et chronologie des Cultures paléolithiques dan le Sud-est Asiatique, pp. 225–249. de Vos, J., van der Made, J., Athanassiou, A., Lyras, G., Sondaar, P.Y., and Dermitzakis, M.D., 2002, Preliminary note on the Late Pliocene fauna from Vatera (Lesvos, Greece), Proceedings of the 1st International Workshop “On Late Plio/Pleistocene extinction and evolution in the Paleartic. The Vatera site”. Annales Géologiques des Pays Helléniques, première série 29 A: 37–70. Vu the Long, de Vos, J., and Ciochon, R., 1996, The fossil mammal fauna of the Lang Trang Caves, Vietnam, compared with southeast Asian fossil mammal faunas: the geographical implications, Indo-Pacific Prehistory Association Bulletin 14: 101–109. Wasson, R.J. and Cochrane, R.M., 1979, Geological and geomorphological perspectives on archaeological sites in the Cagayan valley, Northern Luzon, The Philippines, Modern Quaternary Research SE Asia 5: 1–26. van de Weerd, A., Reumer, J.W.F., and Vos, J. de, 1982, Pliocene mammals from the Apolakkia Formation (Rhodes, Greece), Proceedings Koninklijke Nederlandse Akademie van Wetenschappen, B 85 (1): 89–112. Weesie, P.D.M., 1988, The Quaternary avifauna of Greece, Palaeovertebrata 18 (1): 1–94. Westaway, K.E., Morwood, M.J., Roberts, R.G., Rokus, A.D., Zhao, J-X., Storm, P., Aziz., F., van den Bergh, G.D., Hadi, P., Jatmiko and de Vos, J., 2007, Age and biostratigraphic significance of the Punung rainforest fauna, East Java, Indonesia. Proceedings of the National Academy of Science (under review). Willemsen, G.F., 1983, Paralutra garganensis sp. nov. (Mustelidae, Lutrinae), a new otter from the Miocene of Gargano, Italy, Scripta Geologica 72: 1–9. Willemsen, G.F., 1996, The Cretan Otter Lutrogale cretensis, in: Reese, D.S. (ed.), Pleistocene and Holocene fauna of Crete and its first Settlers, Monographs in World Archaeology 28: 153–157. View publication stats