With only about 1% of the world’s land mass, Island Southeast Asia is nevertheless one of the most species-rich regions on Earth, with, for example, c. 13% of all mammal species. The region’s dynamic geological past (Holloway & Hall, 1998) and relatively stable climatic conditions probably result in many speciation opportunities and relatively few extinction events.
Biogeographical theory for the region dates back to the 19th century, when scientists such as Salomon Müller and Philip Lutley Sclater noted the faunal breaks between east and west Indonesia. These breaks were thought to be associated with deep-water channels that separated the different parts of the Malay Archipelago, resulting in well-known biogeographical boundaries, such as Wallace’s Line between Sundaland in the west and Sulawesi and other islands in the east. Another faunal break, separating the Indochinese and Sundaic subregions along the Thai/Malay Peninsula, is less well known, but is at least as significant, being an important floristic boundary (van Steenis, 1950) that also separates significantly different avifaunal communities (Hughes et al., 2003). There are no obvious geographic breaks in this area, and the mechanisms that separated the faunas of Indochina and Sundaland remain unclear.
Woodruff & Turner’s (2009) recent contribution sheds new light on the biogeography of this region. Along a transect of 20 latitudinal degrees from Singapore in the south to the northernmost point of Thailand, they identify distinct areas of high and low mammal diversity among 325 volant and non-volant species.
To explain these patterns, Woodruff & Turner explore various alternative, potentially overlapping, mechanisms. There is a general trend of increasing mammal diversity towards the equator. This fits observations made in other tropical regions (Stevens, 1989), but does not explain why there should be areas along the transect with unexpectedly low (1°N, 10°N, 16°N and 20°N) or high (4°N and 14°N) species diversity.
As the authors point out, the low species diversity just above the equator could be explained by a peninsula effect. Such an effect at the end of a peninsula has been demonstrated in other places in the world, and is possibly caused by reduced dispersal and population replenishment opportunities at the far end of a narrowing land mass, as typified by southern Malaya and Singapore.
Another possible explanation for the distinct peaks and troughs in the species-richness curve along the transect is that these may represent a simple area effect. This was pointed out by the authors, but will be investigated only in a future study. This possibility perhaps deserved more attention in the Woodruff & Turner paper, as areas of unexpectedly high species diversity around 4°N and 14°N coincide with those parts of the transect where the peninsula is at its widest, whereas the areas of lowest diversity coincide with the narrowest part, or with the faunistically impoverished most northern parts.
Woodruff & Turner (2009) instead focus on an alternative hypothesis for explaining the mammalian diversity patterns. They assume that Neogene changes in sea level led to the divergence between Indochinese and Sundaic faunas. In earlier works on the region’s avifauna, it was thought that early Pliocene sea-level highs (some 4–5 Ma) fragmented the peninsula into several land areas separated by wide sea straits (Hughes et al., 2003; Woodruff, 2003). A new global sea-level curve (Miller et al., 2005), however, suggests that such global highstands were probably less extreme than previously thought and could not fully account for the observed patterns. High sea levels that could have fragmented the Thai/Malay Peninsula into an island chain did occur several times between 43 and 63 Ma. These, however, appear to pre-date many of the species- and genus-level divergence events between the Indochinese and Sundaic faunas (Meijaard, 2004).
Rather than any particular sea-level high fragmenting the peninsula, Woodruff & Turner hypothesize that the significant sea-level lows of the Pliocene and Pleistocene, followed by more than 50 episodes of rapid sea-level rise, compressed populations of species into small parts of their ranges, leading to extinctions. These extinctions would have been most severe in the narrowest parts of the peninsula, explaining the mammal species diversity patterns found along the transect.
Woodruff & Turner’s paper provides useful new data and insights into the potential mechanisms that underlie the divergence between the Sundaic and Indochinese faunas. Two important issues require further study and should be incorporated into future models. First, Woodruff & Turner assume that the Thai/Malay Peninsula was similar in its basic extent and physiography to how it is at present. This is not supported by the existence of a number of depositional sequences in the Thai and Malay Basins east of the peninsula, extending eastwards by a distance of at least the width of the peninsula. Meijaard (2004) provides an overview of the Oligocene and Miocene environments in these basins, which suggests that they were above sea level for most of that time, with successive fluvio-lacustrine, littoral and deltaic or swamp deposits suggesting changes in coastlines but mostly emergent conditions. Not until the start of the Pliocene, some 5 Ma, did this vast region become inundated, giving the peninsular region its approximate present geography. The filling of these basins also makes it difficult to super-impose past sea-level changes onto local sea-bed topographies, because we do not know the extent to which sediment deposition and subsidence have changed the actual depth of the sea bed. Further data are needed to understand how sea-level changes have changed the region’s geography and to what extent this could have led to the compression of populations and possible extinction.
A second issue requiring further study is how the Sundaic and Indochinese faunas evolved. Woodruff and Turner assume that an allopatric speciation model underlies the divergence between the faunas of Sundaland and the Asian mainland, and that the speciation events occurred somewhere on the Thai/Malay Peninsula. This may not necessarily be the case. Speciation in Sundaland seems to have often occurred on outer parts of the island arc, for example on Java, the Mentawai Islands, Borneo, and to some extent also on Sumatra (e.g. Ziegler et al., 2007). Sundaic endemics are often isolated on these islands or within specific habitats such as mountainous forests or evergreen rain forests. There appears to be a trend in many species groups for new, ecologically flexible species groups dispersing to Sundaland at times when it was connected to the Asian mainland (Meijaard, 2004; Meijaard et al., 2008). These species appear to have displaced Sundaic species, restricting them to isolated parts of the region or specialized habitats. Some of those Sundaic specialists might have migrated back to the Thai/Malay Peninsula during times of low sea level and generally drier and cooler conditions, but they would probably have been out-competed in the drier, semi-evergreen or deciduous forest types of the northern Thai/Malay Peninsula because they are ecologically specialized for living in permanently wet tropical habitats (MacKinnon, 1997). The distinction between the Sundaic and Indochinese faunas may thus be maintained by ecology rather than by localized extinction patterns.
The hypotheses proposed by Woodruff and Turner as well as other possible mechanisms for the divergence of the Indochinese and Sundaic faunas require further study. The paper by Woodruff & Turner encourages the collection of necessary data, including better palaeogeographical and palaeoenvironmental information, as well as DNA sequences to establish times since divergence between taxa, and basic ecological information on the region’s mammal species. This research group deserves credit for almost single-handedly addressing biogeographical research needs in one of the most enigmatic faunal transition zones in the world.