Book Reviews

Arthropod Fossils and Phylogeny
edited by Gregory D. Edgecombe
Columbia University Press, New York. 1998, 347 pages
ISBN 0-231-09654-2, $125.00 (U.S.).

Reviewed by Jonathan Adrain
Department of Palaeontology, The Natural History Museum, Cromwell Road, London, SW7 5BD, UK.

The most speciose of extant animal phyla, nearly ubiquitous in modern ecosystems, arthropods are studied in more ways, for more reasons, in more biological disciplines, than any other group. The relationships between major arthropod groups (chelicerates, crustaceans, hexapods, myriapods, trilobites) and other players (onychophorans, tardigrades) have preoccupied workers for decades, with further fuel added by the explosion of interest in molecular systematics. Arthropods have been central to the development of new methodologies (they were, after all, the launch vehicle for phylogenetic systematics), to debates over phylogenetic practice, fossils versus extant taxa, and molecules versus morphology. In spite of all of this attention, there is consensus on very few high-level phylogenetic questions.

Sidnie Manton's (1977) "The Arthropoda" is a legitimate classic of comparative anatomy. Yet it is a measure of how much and how quickly things have changed that its central premise - arthropod polyphyly - is these days nearly an irrelevance, its perception of differences washed away by a tidal wave of shared similarities (Briggs 1990). Almost everyone now agrees arthropods are monophyletic, but while probably a majority of present-day studies (bolstered by molecular results) unite crustaceans with atelocerates (hexapods and myriapods) in the classic Mandibulata, there are also many that align crustaceans with chelicerates and trilobites. Supporters of Mandibulata can't agree whether Atelocerata is monophyletic and sister to Crustacea (though this position has strong majority support), or whether crustaceans are sister to hexapods (Friedrich and Tautz 1995; Dohle 1998; Kraus 1998), with the myriapods potentially paraphyletic (e.g., Akam and Averof 1993, Averof and Akam 1995). Some molecular studies (e.g., Turbeville et al. 1991; Ballard et al. 1992) have united the chelicerates with the myriapods, or with the crustaceans and insects to the exclusion of the myriapods.

In part, these conflicting results are influenced by the fundamentally different classes of evidence on which they are based, by the relative infancy of the field of molecular systematics, and by the need for more extensive sampling; both genetic and taxonomic. But they are also based almost exclusively on study of extant arthropods. They are attempts to find evidence in present-day animals of events that happened more than half a billion years ago (e.g., Wray et al. 1996). In recent years, though, the explosion of interest in high-level arthropod phylogeny has been paralleled by the discovery and description/redescription of several spectacular early arthropod faunas, and by a rekindling of the debate on the nature and meaning of this early evidence (not least due to Gould's [1989] book Wonderful Life).

This is the subject of Greg Edgecombe's fascinating new book: an examination of the influence of fossils and of extinct taxa on hypotheses of arthropod relationship. The book consists of six essay-style chapters on fossil arthropods along with a shorter initial chapter based on extant taxa and a brief introduction by Edgecombe. Illustrations are mainly line art text-figures, as is appropriate to the analytical reviews contained in most chapters. Two chapters draw heavily on fossils from the Lower Cambrian Chengjiang fauna, and it's worth noting that these contain excellent halftones; in several cases of previously unpublished specimens.

Chapter One, "Molecular Systematics and Arthropods" by Ward Wheeler, has, in terms of this book, something of a thankless task: to present a state-of-the-art arthropod phylogeny based on extant diversity. Wheeler is a principal advocate of the "total evidence" approach (Wheeler et al. 1993). Following a brief tour of the molecular arthropod analyses appearing in the past decade most of the chapter is concerned with a combined molecular/morphological analysis of 21 euarthropods as well as a tardigrade, two onychophorans, and six annelid/molluscan outgroup taxa. The result of most relevance to the rest of the book is the strong support for Mandibulata (Crustacea plus a monophyletic Atelocerata).

Chapter Two, "An Arthropod Phylogeny Based on Fossil and Recent Taxa" by Matthew Wills, Derek Briggs, Richard Fortey, Mark Wilkinson, and Peter Sneath, is a 73 page tour-de-force. A morphological analysis including 25 extant, 28 Cambrian, and 9 other fossil genera, and based on 97 comprehensively discussed characters, this chapter supports previous papers in Science (Briggs et al. 1992) and elsewhere (Wills et al. 1994, 1995) with an expanded and fully documented cladistic framework. The chapter also includes a succinct summary of the major issues in arthropod phylogeny and a discussion of some of the critical early fossil groups. The analysis itself includes runs with the full, unweighted dataset, with either extant or fossil terminals excluded, and with the full dataset using successive approximations weighting (SAW). The result is in stark contrast to strictly neontologic studies. Mandibulata receives no support. Instead, chelicerates, trilobites, and their early stem groups form an arachnomorph clade sister to Crustacea - the biramous Schizoramia. It's of extreme interest that the Wills et al. analysis recovers this grouping even with fossil taxa excluded. This is powerful evidence that consideration of early representatives may focus morphological character analysis on features relevant to deep history.

Chapter Three, "Cambrian Lobopodians: Morphology and Phylogeny" by Lars Ramsköld and Chen Junyuan is a gem. Onychophorans have long been identified as waddling around near the base of Euarthropoda, but extant forms are all terrestrial. Although a smattering of fossil marine "lobopodians" have accumulated over the years (including the spectacular, and spectacularly misinterpreted, Hallucigenia from the Burgess Shale), it is really only in the 1990s that a cogent picture of their morphology has taken shape, due mainly to Ramsköld's work on the Chengjiang lagerstätte (Ramsköld 1992a, 1992b; Ramsköld and Hou 1991; Chen et al. 1995). Here, the Cambrian lobopodians are discussed in turn, and analyzed together with Recent Onychophora. It's a superb turn of descriptive precision and comparative anatomy, that results in a dazzling display of the clarity that can be introduced with a skilled cladistic hand. The authors convincingly locate the extant onychophorans as ingroup Cambrian "lobopodians" on the basis of four synapomorphies. The morphology of the Cambrian marine animals must now be accounted for in any discussion of the systematic position of Onychophora.

It's difficult to know what to say about Chapter Four, "Chengjiang Arthropods and Their Bearing on Early Arthropod Evolution" by Jan Bergström and Hou Xianguang. Where most of the rest of the book is based on a rigorous cladistic framework, here the approach is an intuitive "key-character" assertion of hand-mapped characters. One also can't read this chapter without being repeatedly reminded of the factionalized way in which description of the Chengjiang fauna has proceeded. We get one side of the coin here (see also Hou and Bergström 1997), featuring the fauna's discoverer (Hou). Yet the book's editor (Edgecombe) is part of a team (Chen et al. 1994, 1995, 1997; Ramsköld and Edgecombe 1996; Ramsköld et al. 1996, 1997) also doing work on exactly the same taxa. These latter studies are much more in keeping with the relatively unified outlook of the rest of the book, though most of them appeared after the volume went to press. For their part, Hou and Bergström at least provide great entertainment. Seemingly everyone who has taken a critical view of their recent work (and a few bystanders as well) gets a good thrashing. This is an impressively long list, and the chapter features bracing, variably justified, commentary on Ramsköld (p. 153), Whittington and all of his Burgess Shale group (p. 153), National Geographic (p. 153), Lauterbach (p. 167), Schram (pp. 168, 177), Budd (p. 176), Wägele (p. 178), Kukalová-Peck (p. 178), Shear (p. 179), and even Richard Dawkins (p. 180). That the criticism of Ramsköld's work ("...Ramsköld (1992) introduced a new error by omitting the head from his reconstruction [of Hallucigenia] and interpreting the tail as the head.") is soundly demolished in the preceding chapter of the book (Ramsköld and Chen, p. 112) is a measure of the isolation of some aspects of Hou and Bergström's position.

Chapter Five, "Early Arthropod Phylogeny in Light of the Cambrian 'Orsten' Fossils" by Dieter Walossek and Klaus Müller, is concerned with some of the most amazing and beautifully documented of all arthropod fossils. The Swedish Upper Cambrian "Orsten" forms feature exquisitely preserved three-dimensional soft parts (e.g., Müller and Walossek 1985, 1987, 1988; Walossek 1993) and rank among the most important extinct arthropods ever discovered. Most of this chapter is devoted to detailed character analysis, developing an overview of early crustacean phylogeny according to the stem group/crown group concepts of Ax (1985). Many of the "Orsten" taxa are considered stem-group crustaceans. These are of potentially critical importance to the Mandibulata debate as they reveal that certain putative key synapomorphies of Mandibulata (e.g., the presence two sets of maxillae) do not, in fact, occur in stem-group crustaceans.

Chapter Six, "Fossils and the Interrelationships of Major Crustacean Groups" by Fred Schram and Cees Hof, comprises another large-scale cladistic analysis, this time concentrating on Crustacea and involving 67 terminals and 90 characters. Beginning with a detailed overview of previous ideas on crustacean phylogeny and new sources of evidence (including a frank account of the senior author's transformation from evolutionary systematic polyphylist (Schram 1978) to cladist), Schram and Hof offer a generous 14 pages of character analysis culminating in unweighted parsimony analyses of both the entire dataset and the terminals represented by either at least partially extant or entirely extinct taxa. A less usual aspect of the analysis is that many of the terminals are not species (though a significant portion are), but rather are abstractions of large clades. Here, as one might expect, the codings are not presumptive basal states for the clades, but instead account for the range of states developed within the terminal group, resulting in quite a few polymorphic codings (p. 269). Of the 9700 equally parsimonious trees generated by analysis of the entire dataset, slightly more than half group Walossek and Müller's stem-group Crustacea in a clade, while the remainder contain some variation of a stem-group Hennigian comb. Very interestingly, reanalysis with nonfossilized soft-part characters removed (and, consequently, the deletion of many missing codings for the fossils) yielded much stronger support for the stem-group hypothesis.

Finally, Chapter Seven, "Fossil Taxa and Relationships of Chelicerates" by Paul Selden and Jason Dunlop, is, by the standards of the rest of the book, a very good half of a chapter. The authors present a thorough review of the phylogenetic status of the chelicerate groups and near-relatives (with summaries of apomorphies) and also of previous attempts to derive phylogenies with fossil information, culminating in an evolutionary tree mapping the cladistic results of Shultz (1990) and Wills et al. (1995) against stratigraphy. What the chapter doesn't contain is a novel phylogenetic analysis of chelicerate relationships. For that, you have to go to another book, where Dunlop and Selden (1998) present an 18-terminal, 25-character cladogram. To be fair, this probably reflects different lead-in times to publication. Despite the closely spaced appearance of the books, and Selden and Dunlop's chapter is a useful and thorough precursor to Dunlop and Selden's. It holds up well as a good review in its own right. Still, other chapters in Edgecombe's book provide the reader with both the background and the analysis.

What's the net effect? Just plain grand. This is an engrossing book, full of the issues that have helped to set the paleontological agenda in Science and Nature in recent years, with spirited overviews written by the main investigators. As a compelling and multifaceted argument for the critical role of fossils in the recovery of deep history there has simply been nothing like it published previously. It also functions as a readable introduction to the main issues in arthropod evolution, as a thorough review of the primary literature of the last 20 years, as a vehicle for some novel, extended phylogenetic analyses, as a treasure trove of character information, and even, in places, as a guilty pleasure for those of us who enjoy sampling the socio(path)ology of science. It is required reading for anyone interested in arthropod systematics. But it is its resounding parade of hard evidence for the core relevance of paleontology gives this book a broader significance and makes it something special.

See References.

Copyright: Palaeontologia Electronica, 15 March 1999