Testing biodiversity hypotheses in New Caledonia using phylogenetics

New Caledonia represents an extraordinary ‘laboratory’ for evolutionary biologists. It harbours a high level of species richness and endemism as well as numerous relict groups. Here, I briefly review the various hypotheses that have been proposed to explain New Caledonian biodiversity and discuss the use of the phylogenetic method to test these hypotheses. New Caledonia has long been depicted as a museum of biodiversity, where the ‘long isolation of the territory’ and its ‘climatic stability’ can explain the presence of supposedly old endemic relicts, as well as the high species richness and endemism. This classical view has been accepted by many biologists and journalists for decades, despite contradicting geological evidence that New Caledonia underwent complete submersion. For example, in the latest geological review on New Caledonia (written by a geologist working on-site), the conclusion of submersion is unambiguous: after the Cretaceous the Norfolk Ridge/New Caledonia Ridge was below sea level up to the Late Eocene (or possibly up to the Middle Eocene) [...] one may infer that the Gondwanian fauna and flora were introduced in New Caledonia during or after the Middle to Late Eocene. Other recent geological papers also refer to such submersion until c. 35 Ma (Schellart, 2007, figure 4 and Collot et al., 2008, figure 15). In addition, all of the regional boreholes (DSDP 206 – New Caledonian basin; DSDP 207 and 208 – Lord Howe Rise) indicate that the region east of the Norfolk Ridge was below sea level until the Late Palaeocene (Collot et al., 2008). In this context, the Late Cretaceous to Early Eocene, mostly non-marine, siliciclastic rocks of the Kenn Plateau highlighted by Ladiges & Cantrill (2007) are of no particular interest since this structure is located East of Lord Howe Rise. Interestingly, despite the dominance of the museum hypothesis in the literature, biologists have long acknowledged a general submergence of the territory after its break-up from the eastern margin of Gondwana. This has led to alternative hypotheses, the first of which is here called the mountain refugia hypothesis. Based solely on the presence of relict groups in the territory, it rejects the notion of a complete submersion of the island despite the lack of geological evidence: ‘In spite of geological arguments, these submersions can never have been complete, since floral distributions indicate that considerable surface must have remained above water and served as refuges’ (Morat et al., 1984, p. 84). In a similar paper, Morat et al. (1986) propose an alternative hypothesis, here called the island refugia hypothesis. This hypothesis states that close ephemeral islands may have existed and served as refugia for the old Gondwanan biota: ‘Land areas of varying size that may be supposed to have existed near New Caledonia even during periods of general submergence would provide refuges and stepping-stones for the vegetation’ (Morat et al., 1986, p. 645). This hypothesis has never been clearly attested by geological evidence, which would require the finding of organic debris in sediments of the same age as the submersion (B. Pelletier, personal communication). However, it is possible: If New Caledonia was not a refuge for the Gondwanian biodiversity, did other land – except Australia – exist in the region before the Middle–Late Eocene? [...] This provides the possibility for the building of an alignment of islands likely restricted in size east of the Norfolk–New Caledonia Ridge [...] Finally, the long-distance dispersal hypothesis suggests that the present New Caledonian biota reached the island from neighbouring regions (Pole, 1994). This hypothesis implies that for endemic relict groups, the source population went extinct in the territory of origination after the colonization of New Caledonia. In this context, phylogenetic methodology and molecular dating provide the necessary framework with which to study the origin of biodiversity in the region. It is worthwhile to remind ourselves of the difference between the origin and diversification of a group. The most recent common ancestor of a given group and its sister-group represents the origin, and the most recent common ancestor of all the representatives of a given group represents the diversification. If New Caledonian biodiversity follows the museum hypothesis (Fig. 1, A), we would expect both the origin and the diversification of groups to be old, dating back to the fragmentation of Gondwana around 80 Ma. If the diversification of groups is younger than their origin (Fig. 1, B), as for some invertebrates (e.g. Boyer et al., 2007), vertebrates (e.g. Bauer et al., 2006) and plants (e.g. Setoguchi et al., 1998), it is practically impossible to distinguish between the three alternative hypotheses (mountain refugia, island refugia and long-distance dispersal). Some authors have favoured one or other of the alternative hypotheses. Pole (1994) favoured a post-Eocene dispersal hypothesis; Jolivet & Verma (2008) favoured an explanation implying incomplete submersion; and Heads (2008) favoured an island refugia hypothesis. None of those authors based their argumentation on a phylogenetic methodology. Indeed, as shown in Fig. 1, B, the same phylogenetic pattern (old origin and recent diversification) can be explained by the three hypotheses. Favouring one hypothesis over another is an expression of personal opinion based on geological arguments. In the case of a demonstrated recent origin (more recent that 80 Ma), the only valid interpretation that remains is long-distance dispersal (Fig. 1, C). Alternative phylogenetic patterns over time and their possible interpretation. OG, outgroup; NC, New Caledonian. NC endemics are assumed to represent endemic monophyletic groups with their sister-group (OG) distributed outside New Caledonia. There is an increasing body of evidence showing that much of the New Caledonian diversifications do not date back to the fragmentation of Gondwana (e.g. Murienne et al., 2005). This new paradigm is in agreement with the geological history of the territory indicating an emergence around the Middle–Late Eocene. Even if phylogenetics were to prove to be useful to study the causes of diversification in New Caledonia, or to test if supposedly Gondwanan groups follow a museum model, alternative hypotheses (mountain refugia, island refugia, long-distance dispersal) about the presence of relict groups all explain the same phylogenetic pattern, and so the hypotheses cannot be separated on the evidence. In this context, the presence of numerous relict groups in New Caledonia remains puzzling. Editor: John Lambshead