Towards a new synthesis of evolutionary relationships and classification of scleractinia

ABSTRACT-The focus of this paper is to provide an overview of historical and modern accounts of scleractinian evolutionary relationships and classification. Scleractinian evolutionary relationships proposed in the 19th and the beginning of the 20th centuries were based mainly on skeletal data. More in-depth observations of the coral skeleton showed that the gross-morphology could be highly confusing. Profound differences in microstructural and microarchitectural characters of e.g., Mesozoic microsolenine, pachythecaliine, stylophylline, stylinine, and rhipidogyrine corals compared with nominotypic representatives of higher-rank units in which they were classified suggest their separate (?subordinal) taxonomic status. Recent application of molecular techniques resulted in hypotheses of evolutionary relationships that differed from traditional ones. The emergence of new and promising research methods such as highresolution morphometrics, analysis of biochemical skeletal data, and refined microstructural observations may still increase resolution of the "skeletal" approach. Achieving a more reliable and comprehensive scheme of evolutionary relationships and classification framework for the Scleractinia will require close cooperation between coral biologists, ecologists, geologists, geochemists, and paleontologists.

INTRODUCTION

THE FOUNDATION for our understanding of scleractinian evolutionary relationships was established in the 19th and the beginning of the 20th centuries, when numerous papers dealing with anatomy, physiology, behavior, and skeletal macro- and microstructures were published (e.g., Bourne, 1887, 1899; Pratz, 1882; Fowler, 1885, 1886, 1887, 1888, 1890; Duerden, 1902, 1904). A few authors (Duerden, 1902; Matthai, 1914) used anatomical features of coral polyps as key-characters in classification, but the two most advanced 19th century classifications (Milne Edwards and Haime, 1857; Ogilvie, 1897) were established using complex and relatively accessible skeletal characters.

Milne Edwards and Haime's (1857) classification was based mainly on macroscopic skeletal characters and was rooted in the broad zoological and paleontological knowledge of the two authors. Their greatest achievement was providing a uniform terminology for structures and a framework of classification into which any new form could be placed. This scheme, originally bipartite (Aporosa and Perforata) and then tripartite [Duncan (1885) included Fungida of equal rank with the Aporosa and Perforata], survived until Vaughan and Wells's revision (1943). In contrast, Ogilvie's (1897) classification was developed mainly using observations of skeletal microstructures of some representatives of Recent and fossil corals. Special attention was paid to the structure and pattern of the distribution of septal trabeculae. Though acknowledged by contemporary workers, the microstructural observations of Ogilvie (1897) were not implemented in scleractinian studies until Vaughan and Wells's (1943) revision, in which many of Ogilvie's (1897) original drawings were reproduced.

During the 20th century, four main systems of scleractinian classification (Vaughan and Wells, 1943; Wells, 1956; Alloiteau, 1952; Chevalier, 1987) attempted to assimilate new data on the diversity and skeletal structures of Mesozoic and Cenozoic faunas. The authors of these systems contributed significantly to our knowledge of the Scleractinia, and were active in the systematics of both Recent and fossil corals. At this time, because the main taxonomic criteria were skeletal, there was no discrepancy between zoological and paleontological classifications. The agreement between zoological and paleontological classifications is noteworthy, because different classifications existed for many other groups of organisms (e.g., molluscs) depending on the field of the researcher.

The emergence of molecular techniques at the end of 20th century resulted in hypotheses of evolutionary relationships that were independent and different from traditional ones based on the skeletal data (Romano, 1996, 2000; Veron, 1996). New interpretations of coral biomineralization challenged the traditional concept of physiochemical control of skeletal formation (Cuff, 1997, 1998; Gautret, 1997) During the same period, new records and revisions of Recent azooxanthellate (Cairns, 1979, 1982, 1984, 1995, 1998, 1999; Zibrowius, 1980; Stolarski, 2000), zooxanthellate (Hoeksema, 1989; Wallace, 1999; Veron, 2000) and fossil coral faunas (e.g., Beauvais, 1982; Morycowa, 1971; Roniewicz, 1989; Loser 1989, 1994; Baron-Szabo and Steuber, 1996; Baron-Szabo, 1997, 1998, 2000; Lathuiliere, 2000a, 2000b) significantly improved understanding of scleractinian morphological and microstructural diversity through time. Last but not least, new hypotheses about the processes underlying scleractinian evolution (e.g., the concept of reticulate evolution by Veron, 1995) and about the integration of spectra of morphological variation into scleractinian classification (e.g, Lathuiliere, 1996) were proposed. All of these new observations and hypotheses concerning Recent and fossil scleractinians have not yet been discussed and summarized in the form of a Treatise on Invertebrate Paleontology or Zoology. However, it has become crucial for today's workers to learn how best to reconcile the increasingly growing discrepancy between molecular vs. morphological systems of scleractinian phylogeny and classification. Will it be possible to maintain the unity of zoological and paleontological classification in the future? Will it be possible to establish the level of macro- and microstructural skeletal observations that will reliably explain relationships between fossil and Recent taxa? Will the observations of Recent forms correspond with those suggested by molecular analyses?