Ecology of Sulawesi. Tony Whitten

      A similar distribution is found among the spiny eels (Mastacembel-idae) in which four species are found in Borneo, three in Halmahera, but none in Sulawesi. Fish are in fact a useful group to examine with respect to different opinions concerning the narrowing or closing of the Makassar Straits during the Pliocene. For this it must be understood that freshwater fish can be divided into three ecological groups:

      • those confined to freshwater with no tolerance to brackish or salt water;

      • those generally encountered in freshwater but showing some tolerance to salt; and

      • those with considerable salt tolerance which either migrate between freshwater and marine habitats or are of marine origin and have colonized freshwater habitats (Darlington 1957). This grouping is particularly useful since there are strong correlations between the ecological and systematic groups (Myers 1949; Cranbrook 1981).

      The fish fauna of Sundaland has many species from the first two groups whereas Sulawesi has no strictly freshwater fish, only a few with a little tolerance to salt, and the majority are from the third group. The Makassar Straits or Wallace's Line therefore seems to separate two distinct fish faunas and it is therefore unlikely that Sulawesi and Borneo were ever a single landmass.

      Freshwater fish may not have been able to reach Sulawesi but all major terrestrial animal groups have. Amphibians, for example, have no indigenous species in common with Borneo but of the endemic species four are related to Sundaland species and two have Papuasian species (p. 47). Two of the three frogs that are found east of Sulawesi may in fact have evolved on Sulawesi and subsequently dispersed (Cranbrook 1981). Twenty-three lizard species found on Sulawesi are also found west of Wallace's Line (C. McCarthy pers. comm.). Similarly, of Sulawesi's 63 snake species, 38 are found both sides of Wallace's Line but only two are found in New Guinea and not in Sundaland (den Bosch 1985).

      The tortoise Indotestudo forsteni was thought to be of particular interest because it was the only land tortoise known from both sides of Wallace's Line having been first found in Minahasa and Halmahera. It now appears that it is the same species as the later-described I. travancoria of eastern India, Thailand and Burma, and was brought from there by traders as a food animal (Hoogmoed and Crumly 1984).

      The birds of Sulawesi are predominantly western with 67% of the species having affinities with the Sunda region (Mayr 1944). One, the Sulawesi Roller Coracias temmincki, in fact has no relatives in Sundaland and all the other members of its genus are found in Europe, mainland Asia and Africa (Klapste 1982).

      Only two of the non-flying mammals on the mainland, the cuscuses, are clearly of Australian/New Guinea affinity;²⁷ the remainder, including the endemics, have their origin in Asia. Wallace's Line delimits the eastern boundary of the distinctive Sundaland fauna comprising moles, flying lemur, tree shrews, lorises, gibbons, pangolins, porcupines, dogs, bears, otters, weasels, cats, elephants, tapir, rhinoceroses and mouse-deer. Thus the fauna of Sulawesi has closer affinities with the Sunda than with the Philippines, Lesser Sundas or Sahul region, but cannot really be said to be part of it.

      In general, Wallace's Line is not as clear a demarcation line for invertebrates as it is for vertebrates but there are many fewer genera in Sulawesi than in islands to the west (Gressitt 1961). For example, there are about 1,200 species of butterflies in the Malay Peninsula, 850 in Borneo and New Guinea, but only 450 in Sulawesi (R. Vane-Wright pers. comm.). As with the palms discussed above, the relatively low number of species may be the result of a previously unfavourable climate. Cicadas are poor dispersers and so their distributions show more detectable patterns than do many other groups. The cicadas of Sulawesi are largely of western origin but there are relatively large numbers of endemic species and some endemic genera (Duffels 1983). The same general pattern (major affinity with Borneo, relatively depauperate, but high level of species endemism) can be observed in other groups such as ground beetles, tiger beetles (N. Stork pers. comm.), and pond skaters/water striders (D. Polhemus pers. comm.), but the affinities of dung beetles Scarabaeidae, for example, are by no means clear (J. Krikken and H. Huijbregts pers. comm.). There are now about 200 species of ground beetle known from Bogani Nani Wartabone National Park and although less than half the total known from Borneo, this is still a massive total for such a relatively small area (N. Stork pers. comm.). This needs to be investigated further.

      A detailed analysis of butterfly and moth relationships has shown that Sulawesi species are most strongly associated with the fauna, not in Borneo, but in the Philippines, reflecting the ancient connection through the Sangihe Islands (Holloway 1987). As shown above, the plants of Sulawesi also show a greater affinity with the north, east and south than with Borneo, and the difference in affinity of butterflies when compared with other invertebrates, may be a result of the caterpillars being dependent on specific living plants rather than on detritus. The moth families examined in most detail showed virtually no elements from the direction of Australia possibly because this moth fauna had become established by the time the connection through the Sula Islands occurred (Hollaway in press). All these analyses are hampered to some extent, however, by a generally poor knowledge of the Moluccan fauna.

      One recent addition to the butterfly fauna of Sulawesi was not through human agency but through natural dispersal. The range of the Monarch butterfly Danaus plexippus used to be restricted to eastern and western North America where it performs remarkable annual migrations between the north and south of its range covering up to 3,000 km in a year travelling up to 125 km per day. This remarkable flying ability and the relatively long life of the adult has meant that when blown off course by strong winds it has reached and successfully colonized new areas. It first reached Hawaii in 1845, Australia in 1870, and Manado in 1873. It has not spread widely within Sulawesi, however, and was not collected by entomologists in or around Bogani Nani Wartabone National Park only 130 km from Manado (R. Vane-Wright pers. comm.).

      As might be expected there is no obvious difference in the composition of coastal marine faunas either side of Wallace's Line. Corals, for example, show a great similarity between the species inhabiting reefs throughout the Indo-Pacific region and a general absence of endemic species even in remote areas. Most genera have, in fact, been present for 20-40 million years and some species are probably this old too. This is probably a result of the high-frequency fluctuations in sea-level in the Quaternary (p. 16) which alternately exposed and covered coastal regions. There was probably simply not enough time for populations of long-lived corals to complete many generations before their descendants colonized new habitats, and this may have maximised variation within a species rather than resulted in the appearance of new species. This contrasts with the marked Quaternary speciation observed for more mobile organisms such as fishes and crustaceans (Potts 1983, 1984).

      Biogeographical Differences within Sulawesi

      The fauna and flora of Sulawesi are far from being homogeneous and evenly distributed, but although the geological history might be expected to have influenced the distribution of species, clear patterns cannot now be seen. The flora of the east peninsula might be expected to have close affinities with the Moluccas, but it is in fact closest to the central and west regions of Sulawesi and to Borneo than to New Guinea. Indeed, all areas of Sulawesi except one are closer to each other in floral composition than to regions outside Sulawesi. The exception is the southwest peninsula whose lowland (<500 m) flora is most closely related to the Lesser Sunda Islands (van Steenis 1972; van Balgooy in press). This could be a result of dispersal from Flores via Kalaotoa, Tanahjampea and Salayar or from Sumbawa via the Sabalana and Postilyon Islands. The tip of the southwest peninsula has a dry climate similar to the Lesser Sunda Islands, bounded to the north by wetter, less seasonal climates (p. 24).

      More recent geological history is reflected in the distribution of species within a particular genus. The most striking example