82 VEGETATION OF TUVALU Colin D. Woodroffe Department of Geography University of Wollongong, Wollongong, NSW 2500 Australia ABSTRACT The nine islands of Tuvalu exhibit a complex mosaic of different vegetation units; a major part is covered with coconut woodland, but there are remnants of natural broadleaf woodland, and a series of coastal and inland scrub types. Fony-nine vegetation units were discriminated and mapped as a part of a Land Resources Survey of the islands. The most widespread of these are described, with particular emphasis on detailed mapping and field description from NuL Collection and sight records of pla,nts from the Tuvalu atolls and reef-top islands are listed. Although the islands have only a limited flora, largely of pantropical and easily dispersed species, principally as a result of the probable recolonisation of the islands by terrestrial biota during the Holocene, there are some notable anomalies in the distribution of several plants which indicate the complexity of the dispersal and establishment processes operating in such remote archipelagoes. INTRODUCTION There are a number of different island types in the Pacific basin. Figure I (modified from Scott and ~otondo, 1983) illustrates some of the main types. The simplest island is a basaltic volcanic island (i.e. Hawaii). Such an island may be surrounded by a fringing reef as on recent lava flows on Hawaii, or by a barrier reef with a lagoon separating the reef from land (Le. Tahiti). An almost atoll, where only a small remnant of the volcanic mass remains above S. Pa.c. 1. Nat. Sci., 11:82-128 al Volcanic island bl Volcanic island c) Volcanic island 83 with no fringing with fringing reef with emerzent reef and fringing reefs &J ~/":: ~:. :~~:~~:.:~ 1) Reef-top island gl Emergent hlEmergent atoU j) Makatea island limestone island [BJ Volcanic rocks § Limestone 0 Holocene reef and island sediments Figure 1 Main types of islands on the Pacific plate (modified from Scott and Rotondo, 1983). Initial Eroded Atoll Limestone Modem Form Island Atoll (Karst) S.L.-.......-.,..-- SL·li·· ···· -a!,=~e)Reef growth SL. '" -M~" .' Present (\,\ (\ ,- I o Sea Level I \ I \ '\ I I \ I .... I I : v J 50 J J I Depth J "J I " I \ I 100 (metres) " , , I \ J '..J '...1 150 160 120 80 40 0 Time (thousands of years before present) Figure 2 Late Quaternary sea-level fluctuations and their effect on coral atolls. The lower diagram shows actual reconstruction of sea level and ocean volume variations for the last 140,000 years, based on raised coral reefs in New Guinea and oxygen isotope analysis of deep-sea cores (after Chappell and Shackleton, 1986). The atoll is gradually subsiding. 84 sea level and the reef is dominated by low, carbonate reef islands like those found on an atoll, represents a funher stage in the evolutionary sequence of coral atoll development envisaged by Darwin (i.e. Aitutaki, Cook Islands). Still funher subsidence results in the total disappearance of all of the volcanic mass beneath the sea, but the continued upward growth of the surrounding reef. Islands of carbonate sands and rubble form on the surrounding reef and the island is known as an atoll (i.e. Bikini and Eniwetak, Marshall Islands).Within groups of atolls there are sometimes smaller reef platforms which do not contain an interior lagoon, but which have an island over most of the area of the reef platform; these are termed reef-top islands in this account (i.e. Niutao, Tuvalu). The sequence of islands, young volcanic island, island with fringing reef, island with barrier reef, almost atoll, atoll, is found in several linear island chains that occur in the Pacific Ocean, on which potassium-argon dating of basalts has shown an increasing age towards the northwest. These chains are a spatial analogue of the temporal stages through which anyone island may go. Beyond the northwestern-most island in the chain there are often submerged seamounts and guyots that represent further stages in the subsidence sequence (Grigg and Epp, 1989). Three other types of emergent island can be identified. An emergent atoll is an island that maintains an atoll form but upon which there are Pleistocene limestones above present sea level (i.e. Aldabra, western Indian Ocean). An emergent limestone island is an island which has lost any atoll form that it may have had. A makatea island is one on which there is a highly eroded and degraded volcanic interior, surrounded by emergent limestones (i.e. Makatea in French Polynesia). Emergent limestone islands, while common at plate margins, can also be found on oceanic plates situated up-plate of a subduction zone. It appears that the islands may be located on a flexed bulge of the plate prior to subduction (i.e. Niue). In biogeographical accounts of Pacific islands it has been usual to divide islands into only two categories, high islands and low islands. The high islands comprise all of the above types except coral atolls and reef-top islands which are low. Accounts of the biota of the Pacific tend to concentrate on the high islands for it is on these that there are high incidences of endemism of flora and evidence for recent speciation of fauna. The atolls and reef- top islands by contrast tend to be depauperate in species and characterised by pantropical, drift-dispersed flora (Fosberg, 1984). The reason for this disparity is almost certainly that the atolls and reef-top islands are composed entirely of sediments of Holocene age (formed in the last 10,000 years), and that during that period they have been submerged beneath the sea entirely as sea level has risen since the last ice age, such that they have had to be completely recolonised by plants in the last few thousand years. High islands on the other hand have retained their biota over this 85 period. The pattern of late Quaternary sea-level fluctuation, reconstructed from oxygen isotope analysis of deep sea cores and radiometric dating of reef terraces on an uplifting coast of New Guinea (Chappell and Shackleton, 1986) is shown in Figure 2. The last time at which the sea was around its present level was during the last interglacial, about 120,000 years ago. Reefal limestones of that age are a common feature of islands in the Pacific, and their relative elevation gives an indication of the vertical movement of individual islands since that time (Fig. 1). Atolls, as shown by drilling on Eniwetok, Bikini, Mururoa,Tarawa and the Cocos (Keeling) Islands, have subsided at rates of around 0.1 mrn/yr. When sea level was high an atoll perhaps somewhat similar to that presently found, would have existed where there are atolls today. When sea level was low, that atoll was exposed up to 100 m above the sea as an emergent limestone island, and it underwent solution giving rise to a highly eroded karst surface (Figure 2). During the 100,000 years or so that it has taken for the sea to return to its present level, the atoll has gradually subsided so that the former (last interglacial) surface is about 10-20 m below present sea level beneath the present atoll. This surface has been identified on a number of rtolls (7-14 m on Eniwetok, Szabo et al., 1985; 8-17 m on Tarawa, Marshall and Jacobsen, 1985; 8-12 m on Cocos (Keeling) Islands, Indian Ocean, Woodroffe et al., in press). The upper 10 or 20 m of reefal limestones and unconsolidated sediments built up in the Holocene (the last 10,000 years), with coral establishment over the karstic last interglacial limestones recommencing about 8000 years ago (Marshall and Jacobsen, 1985). At that time the older limestone basement was flooded and reef growth lagged behind sea level. Only since sea level has stabilised during the last few thousand years has reef growth caught up with sea level, reef flats typical of Pacific islands developed, and reef islands formed on the reef flat. In many cases reef island formation has been made easier by a fall of sea level of about I m during the last 3000-4000 years (Pirazzoli and Montaggioni, 1986). On the majority of atolls the entire terrestrial biota would have been eradicated 8000 years ago (and perhaps episodically prior to that time) when the sea submerged the last interglacial limestone substrate of the atoll, and, as no evidence of reef islands existing during the subsequent phase of catch-up reef growth has been found, probably could not have recolonised until around 3000 years ago when the present reef islands began to form. The flora and vegetation of atoll reef islands therefore represent the results of relatively recent transoceanic dispersal and colonisation. The reef islands ihemselves appear superficially rather simple and vary little from archipelago to archipelago. Comparison of flora and vegetation of different reef islands provides an opportunity to decipher the detail of ecological processes, such as dispersal and establishment of propagules. This approach received an enormous stimulus from the pioneering theories of island biogeography proposed by 86 MacArthur and Wilson (1967), who postulated a dynamic equilibrium between species diversity and the size and remoteness of islands. The relatively recent development of reef islands on atolls, has also coincided with human colonisation of the Pacific. Clearly atolls could not be colonised before reef islands had developed, but the archaeological record from mid-Pacific high islands aePears to indicate that colonisation of those islands also occurred around 3000 years ago (Bellwood, 1978; Gibbons and Clunie, 1986). People have had a profound influence on the vegetation of atolls, most notably through the widespread preferential planting and maintaining of coconut woodlands.