New Caledonia is on the margin of the Australian and Pacific plates, 1200 km east of Queensland and 1700 km north of New Zealand in the south-west Pacific. It is renowned for a high level of endemism and ancient taxa, including: among the flora, one of the earliest angiosperm lineages (Amborella), a clade of five species of Nothofagus and 14 genera of conifers; and among the fauna, the flightless bird kagu (Rhynochetos jubatus), 70 terrestrial species of reptiles, inland fishes (e.g. Galaxias neocaledonicus), about 4000 species of insects and 200 endemic species of land snails (references in Ladiges & Cantrill, 2007). Biogeographers have variously interpreted elements of the biota as old, ‘holding the master key to the past’ (Corner, 1963, p.238; Raven & Axelrod, 1972), or young, derived by long-distance, over-water dispersal (e.g. Pole, 1994).
Michael Heads (2008) has provided a comprehensive account of the biota – relationships, biology and ecology – and of the geology of this key region. His is a unique analysis for a number of reasons. First, he draws together a wealth of biological evidence to identify congruent biogeographical patterns (tracks) on Grand Terre and the adjacent Loyalties. This contrasts with the current trend in historical biogeography (particularly in molecular phylogenetic studies) to report on single taxa. Second, Heads deals masterfully with the complexity of the geology. Third, as one anonymous reviewer of Heads’ manuscript wrote, it is a ‘holistic synthesis of the geological and biological evidence’, with lessons from the biota informing the geology and vice versa. The geologist Rawling (1998) noted that models of the geological history of New Caledonia can be informed by, and should include, biogeography.
Heads starts with the Jurassic/Early Cretaceous history of the basement terranes of New Caledonia, formed from island-arc material accumulated in the pre-Pacific Ocean, later accreted onto eastern Gondwana. He concludes that the Eocene collision of basement terranes with an island arc to the north-east is likely to have brought together different biotas following separation from Gondwana. The Loyalties are situated today only a short distance from Grande Terre, but the two were brought together after 30 million years of separation, which is reflected in the fact that the taxa of the Loyalties have a greater affinity with Pacific groups than they do with those on Grande Terre and Australia. The complexity of this history also defies summary as a simple geological/area cladogram. New Caledonia cannot be represented simply as a single area related to New Zealand (e.g. Sanmartín & Ronquist, 2004, fig. 1, p. 217, their geological area cladogram for land masses in the Southern Hemisphere). Historical area relationships based on taxa should show links to Australia, New Guinea, New Zealand and the Pacific. Croizat (1964) summarized these relationships as a number of generalized tracks intersecting New Caledonia (his fig. 39, p. 154; see Ladiges, 1998, fig. 2).
Heads argues that taxa are not necessarily the same age as the rocks on which they now perch, which has implications for those who use such ages as calibrations for molecular dating. Islands such as the present-day Loyalties (currently 2 million years old) may derive their flora and fauna from previous islands that have come and gone. Similarly for Grande Terre, the diverse sclerophyllous vegetation (maquis) on the ultramafic soils (ophiolite material) may derive from adjacent subaerial land. Heads refers to ‘a dynamic metapopulation surviving in situ’, similar to the notion of Ladiges & Cantrill (2007), who emphasized ‘a biota of greater New Caledonia…not just simply the biota of the present-day island that could have been submerged for part of the Cenozoic’. Others will argue that the distinctive maquis vegetation is unlikely to pre-date the age of the ophiolite, and that Heads’ hypothesis faces the difficulty that the adjacent land, which he claims supported the source biota, has now vanished or been overlain. This highlights the point that reconstructing the geological history is especially difficult, but, as Dr G. Chaproniere has reported (personal communication cited in Ladiges & Cantrill, 2007), the presence of Middle Eocene shallow-water limestones in the south-west Pacific implies that land was very widespread in the region, including in that of New Caledonia, during this period.
Recently, various authors have referred to the ‘balance’ of long-distance dispersal and vicariance (including geo-dispersal) in historical biogeography, and that the pendulum has swung in favour of the former. Heads’ paper provides a note of caution for those who dismiss vicariance and infer dispersal on the basis of low molecular differences among taxa and divergence dates that are too young, assuming precision of calibration and knowledge of evolutionary rates that are unwarranted.