The challenge of understanding the Coral Triangle biodiversity hotspot

It can be argued that biogeography was born 150 years ago following Wallace's observations of terrestrial biodiversity in the Indo-Malay Archipelago, and that MacArthur and Wilson escorted the field into the modern age in 1963 with their theory of island biogeography. It is thus noteworthy that one of the most vexing issues in modern biogeography remains the origins of high marine biodiversity in the Coral Triangle, a region comprising six archipelagic nations that span the Indo-Malay Archipelago. The Coral Triangle is the global epicentre for marine biodiversity, with diversity that rivals the Amazonian rain forests, and is a terrestrial biodiversity hotspot as well. Across a broad range of taxa, species diversity peaks within this region, with sharply decreasing diversity gradients moving away from the centre. While this pattern has been well known for decades, the processes contributing to this pattern have remained elusive. Multiple hypotheses have been proposed to explain the above pattern. Ekman (1953) suggested that this region is a 'centre of origin'; speciation occurs within the Coral Triangle with the observed biodiversity gradients resulting from ocean currents limiting range expansion outwards from the centre. Ladd (1960) proposed instead that this region was a 'centre of accumulation', whereby newly formed species from islands in the Indian or Pacific Oceans disperse towards the centre of diversity, either on ocean currents or from the collision and integration of islands over geological time, uniting otherwise independently evolved faunas (McKenna, 1973). Lastly, Woodland (1983) hypothesized that the Coral Triangle is a 'centre of overlap', where Indian and Pacific sister species overlap in the centre, creating elevated biodiversity. While these three theories were derived from patterns observed in a variety of taxonomic groups across the Indo-Pacific, they often lacked a theoretical framework to test competing hypotheses. Halas & Winterbottom (2009) employ a novel phylogenetic approach (pact, phylogenetic analysis for comparing trees) to examine the origins of high marine biodiversity in the Coral Triangle, a region they call the 'East Indies Triangle' Originally, Santini & Winterbottom (2002) used Brooks parsimony analysis (BPA), wherein the phylogenies of taxa are converted into area cladograms, representing the geographical ranges of the species analysed. Their consensus of the area cladograms for 13 different taxonomic groups provided support for the centre of accumulation model. However, BPA has a major shortcoming in that it cannot easily accommodate species ranges of reticulated origin, such as is specifically predicted by either version of the centre of accumulation model. Through reanalysing Santini & Winterbottom's (2002) dataset using pact, a program designed to accommodate reticulate geographical origins, Halas & Winterbottom (2009) find no substantial support for any of the proposed theories. This inability of modern biogeographical methods to answer a decades old question is not necessarily an indictment of the state of our science but instead highlights the magnitude of the difficulty that understanding the origins of the Coral Triangle biodiversity hotspot represents. Fundamental to each of the above hypotheses is a prediction about where speciation has occurred and how the ranges of individual new species have subsequently been modified over time – knowledge of each of these processes presents a unique challenge in marine environments. Determining where and how speciation may have occurred is challenging, because most marine species have a dispersive pelagic larval stage, some of which can travel great distances, providing connectivity among geographically isolated populations, even if that potential is not always realized (Kinlan & Gaines, 2003). Pelagic dispersal challenges our notions of barriers to dispersal: what is an allopatric barrier for a highly dispersive propagule travelling on ocean currents and where are those barriers? Rare, random events may transport larvae out of their normal range, promoting diversification but, by being rare and random, such events can be difficult to document. Localizing where divergence occurs can be further complicated because long-distance larval dispersal combined with metapopulation dynamics (Kritzer & Sale, 2006) means that the ranges of marine species may be very dynamic, potentially erasing the signal of previous isolation. Documenting the dynamics of the ranges of individual species over evolutionary time is also challenging because not all marine environments and not all marine taxa preserve well in the fossil record. Broad-scale biogeographical studies of fossil marine communities provide important insights into the history and assembly of Coral Triangle biodiversity on deep time-scales (e.g. Wilson & Rosen, 1998), but may be less applicable to more recent evolutionary processes contributing to present-day patterns of biodiversity. Even if a complete series of fossils was found that spanned the Coral Triangle and adjacent oceans, it would be unclear what those fossilized remains would represent because genetic studies within the Coral Triangle are increasingly showing that many species contain multiple highly divergent lineages that could represent cryptic species (Barber et al., 2006). Unravelling these complex challenges requires integration among disciplines. As Halas & Winterbottom (2009) point out, the increasing number of phylogenies of Indo-Pacific marine species, generated largely by molecular data, will be of tremendous value to cladistic biogeographical approaches, and will allow evolutionary biologists to focus directly on the process of species formation. For example, Barber et al. (2006) demonstrated concordant patterns of lineage diversification in three species of stomatopods that could only be explained by diversification within the Coral Triangle, providing mechanistic support for a centre of origin. While patterns of genetic diversity shift across space and time, much like species ranges, a valuable aspect of genetic diversity is that it provides a direct record of those shifts. By simultaneously documenting regions of genetic divergence through phylogeographical analysis, as well as the directionality of expansion of that genetic diversity through directional gene flow analyses, it then becomes possible to directly test the predictions of the above four models on a species basis. For example, Drew et al. (2008) demonstrated strong patterns of genetic diversification in peripheral populations of five Indo-Pacific reef fish, a pattern consistent with the centre of accumulation model. Subsequent analysis shows that the Fijian clade of one species, Pomacentrus moluccensis, is basal, and that the age of clades decreases on moving into the heart of the Coral Triangle, matching the predictions of the centre of accumulation hypothesis (Fig. 1; Drew & Barber, 2009). However, while these new cladistic and genetic methods provide an opportunity to unravel more recent evolutionary processes, they complement, but are not a replacement for, traditional palaeontological techniques which provide insights on deeper time-scales. Cladogenic expansion from Fiji into the Coral Triangle in Pomacentrus moluccensis based on maximum likelihood analysis of 192 mitochondrial DNA control region sequences using Pomacentrus callanus as an outgroup (after Drew & Barber, 2009). Shaded circles indicate the geographical distribution of phylogroups (A, white; B, light grey; C, dark grey; D, black) depicted on the tree. The scale bar indicates five mutational steps. The integration of traditional biogeographical approaches with modern cladistic and genetic methods is likely to transform the field of biogeography, ushering in an era of increased understanding of the linkages of biogeographical pattern and process over a broad range of time-scales. While this 'new synthesis' in biogeography will surely increase our understanding of the processes that have contributed to high biodiversity in the Coral Triangle, it may not yield a single unified theory. Given the age, complex geological and oceanographic history, the vast geographical scale, and the magnitude of the biological diversity of the Coral Triangle, it is very likely that multiple processes have been at work, either simultaneously or with different time periods being dominated by different evolutionary processes. While the consistency of biodiversity patterns in the Coral Triangle, across taxa and on both land and sea, seems to beg for a unifying theory, Halas & Winterbottom (2009) may have hit on the answer to this persistent question of marine biogeography – that there may be no single explanation for the pattern. This commentary was stimulated by discussions with J. A. Drew and was greatly improved by comments by R. Winterbottom and R. A. Rosen. Support was provided by NSF OCE- 0349177.