slightly smaller size than modern specimens. The large relative size of many Pleistocene fossils is a commonly observed phenomenon, possibly a result of the cooler temperatures then prevailing. Larger animals have a lower ratio of skin area to body volume and a consequently lower rate of heat loss. As a result, larger animals are better able to survive at low temperatures (Edwards 1967). Dwarfing of animals on islands generally occurs in species that are relatively large, such as elephants and anoas, whereas gigantism on islands generally occurs in small species such as rats.
The Pleistocene sharks are all still found around Sulawesi and occasionally in rivers, but Hemipristis is now represented only by a single rare species in the Red Sea and Zanzibar (Hooijer 1958). The stingray is also common in Indo-Pacific seas and rivers.
The animal remains found to date are much poorer in numbers and species than those from the Trinil fauna deposits in East Java. Whereas the Trinil deposits may be used to some extent to illustrate the fauna present over some of Sundaland, it is not possible to use the Trinil finds as a basis for discussion of the Sulawesi palaeofauna. Two of the four elephants but none of the three pigs in the Trinil deposits are known from Sulawesi (Hooijer 1974), and the pygmy stegodont on Sulawesi is now also recognized as the same species as the fossil stegodont found on Timor and Flores. The Plio-Pleistocene fauna of Sulawesi is thus not as distinctive as was once thought and the only species not known from outside Sulawesi is the giant pig.
* extinct forms
After Hooijer 1948a, b, 1949, 1950, 1954, 1958, 1964, 1967, 1969, 1972; Clason 1976; 1982; Musser 1984
Figure 1.21. Relative sizes of the giant tortoise, 1.65 m man and a pygmy elephant.
The ancestors of the elephant and stegodonts could have reached Sulawesi via Taiwan and the Philippines (stegodont remains have been found on both islands). The 3,000-m deep sea between Java, Flores, Alor, Timor and Sulawesi might be thought to be too great for even water-loving elephants to cross by swimming (Hooijer 1967, 1970, 1972). If it is assumed that the presence of similar pygmy stegodonts on all these islands must be a result of migrations across land and only short water gaps, it is necessary to postulate massive downfaulting of at least 3,000 m in the intervening straits and seas. Before the downfaulting occurred there would have been routes from Flores to Sulawesi along the Kaloatoa ridge and then via Tanahjampea and Salayar, and from Java to Sulawesi via Madura, the Kangean Islands and the Doangdoangan Islands (Audley-Charles and Hooijer 1973) or perhaps more easily from the Kangean Islands via the Sabalana and Postilyon Islands. The routes from Java may in fact have been viable even without downward faulting during the glacial periods of the Pleistocene during which sea-level dropped 100-130 m below present levels (p. 16).
A recent examination of the swimming powers of elephants has suggested, however, that such massive downfaulting need not have occurred. Elephants have been recorded swimming across lakes and seas to islands in Africa and Asia and the distance record is held by an unfortunate elephant that swam ashore after being washed overboard from a boat 48 km off the South Carolina coast in 1856. Aspects of elephant morphology such as the spongy skull bones and absence of a pleural cavity, the need to bathe regularly and their ancestry (dugongs, or sea cows, are among their closest relatives), suggest that their ancestors may have lived in a semi-aquatic habitat. Some of the distances prehistoric elephants would have had to have swim are in the upper limits of known ability and so the crossing of elephants to and from Sulawesi and other islands does not necessarily require that the sea bed was much higher, and exposed land more extensive, in the recent past (Johnson 1980).
Giant tortoises are said to be able to float and survive long periods in seawater. Thus Geochelone atlas could have drifted to Sulawesi from another landmass (Hooijer 1982), and in fact remains of the same species are found in deposits from Java, Timor and India.
Present Fauna
The fauna of Sulawesi is one of the most distinctive in all Indonesia particularly among the mammals (table 1.8). Of the 127 indigenous mammal species,14 79 (62%) are endemic15 and the percentage rises to 98% if the bats are excluded (table 1.9). The mammal fauna is also characterized by its relatively primitive characters (Musser in press). New species of mammals continue to be found (e.g., Musser 1981, 1982; Bergmans and Rozen-daal 1982; Hill 1983; Boeadi pers. comm.) and revisions of old and new specimens combined with fieldwork help to clarify the exact number and identity of species (e.g., Musser 1971a, b, 1973, 1977; Musser et al. 1982; Groves 1980a, b; Takenaka 1982).
One of the earliest descriptions by a European of a Sulawesi animal was of the curly-tusked babirusa Babyrousa babyrussa by Piso in 1658. It was kept and even bred by early rulers, perhaps as a gift for visiting diplomats, and it is conceivable that some were taken to Bali by Buginese seafarers. It is possible that knowledge of the babirusa influenced the way in which 'Raksasa', a Balinese demonic man-beast, was first drawn, for it has a curved tusk piercing each cheek (Groves 1980a).
The babirusa are enigmatic animals as their name, which means 'pig-deer', implies. They are variable in hair covering (from nearly naked to densely hairy), hair colour (from off-white to brown), and eye colour (from whitish, through grey to brown) (Wemmer and Watling 1982) and none of these characteristics appear to be related to sex. There may be geographical trends in these characters, however, but too few specimens with detailed location data are available. The babirusa is generally grouped with pigs but they have had no common ancestor since the Oligocene (about 30 Ma), and they are barely more similar to pigs than they are to hippos, remains of primitive forms of which have been found in Java (C. Groves pers. comm.).
Two dwarf buffaloes, or anoa, Bubalus depressicornis and B. quarlesi are said to be found in the lowlands and mountains respectively (Groves 1969) but it has been suggested that the differences in horn-shape, a major means of distinguishing the species,16 may simply be a function of age (Wind and Amir 1978). Also the 'mountain' anoa is sometimes found at sea-level and the 'lowland' anoa is sometimes found on high mountains (Thornback 1983). The diurnal bear cuscus Ailurops ursinus is quite a large animal, with a head and body 45 cm long and a tail 55 cm long, and is quite frequently seen in lowland forests. It is very distinct morphologically from other cuscus and is in its own genus in recognition of its primitive characters (C. Groves pers. comm.). The smallest cuscus of all, Strigocuscus celebensis (head and body only 34 cm), is nocturnal, and consequently rarely seen.
The composition of the Sulawesi mammal fauna is very different from that of Borneo or Irian Jaya with many fewer families represented (fig. 1.22). The rats and bats are major constituents of the fauna as they are in the insular regions of the Moluccas and Lesser Sundas.
Based on Anon 1982b
Figure 1.22. The faunas of Borneo, Sulawesi, Flores, Maluku and Irian Jaya (Vogelkop) compared.
There are 332 species of birds known from Sulawesi of which 92 (27%) are endemic, and 81 (25%) are migratory (p. 145; table 1.10) (White 1974, 1976, 1977; White and Bruce 1986). New records of species not previously known from Sulawesi are still being made (Escott and Holmes 1980; Watling 1983). Among the resident birds, 17 genera are endemic to Sulawesi and its surrounding islands including a large number of spectacular endemic birds such as the dark-green bee-eater Meropogon forsteni, the large brightly-coloured hornbill Rhyticeros cassidix, the crowned myna Basilornis celebensis, and the finch-billed starling Scissirostrum dubium which nests in huge numbers in holes bored out of tall dead trees (fig. 1.23). Sulawesi's best-known bird is the maleo Macrocephalon
