Ghosts of the past, present, and future in brachiopod systematics

ABSTRACT-Three historical phases can be distinguished in the study of brachiopod systematics over the past 75 years. Prior to 1956, systematic neontologists and paleontologists struggled to reconcile differences in perceived evolutionary patterns (and thus classifications) based largely on static morphological differences observed separately among living brachiopods and among fossil brachiopods. Following 1956, patterns of morphological distribution began to be interpreted relative to the processes by which they were formed, and a more dynamic view of brachiopod phylogeny and classification resulted. Over the past decade, newer methodologies (phylogenetic systematics) have allowed older phylogenetic hypotheses to be tested and evaluated. The major challenges that brachiopod systematists now face are not unique to brachiopods; they concern improving the methods of phylogeny (and classification) reconstruction so that all the sources of data available to paleontologists can be utilized more effectively. In the future, I predict that more intensified, global fossil collecting, together with further investigation of the embryology and development of brachiopods, and molecular systematic research, will play an increasingly larger role in revising the classification currently in use.

INTRODUCTION

BRACHIOPODS ARE solitary marine coelomates, with a bivalved shell bilaterally symmetrical about a sagittal plane, composed of either calcium phosphate or calcium carbonate, secreted by dorsal and ventral mantles (folded extensions of the body wall) with mantle canals (coelomic extensions). The lophophore, a ciliated, tentacular organ for feeding (and respiration), of varying three-dimensional morphology, is suspended between the mantles. The phylum Brachiopoda is a clade (Rowell, 1981 a; Carlson, 1995; Williams et al., 1996) that may (Cohen, 2000), or may not, include the Phoronida as a derived, shell-less subclade. The group is large and diverse, with approximately 4500 named genera grouped into 109 superfamilies, 26 orders, eight classes, and three subphyla (Williams et al., 2000; see Tables 1 and 2). Fewer than 5% of the genera are extant, but they appear to be distributed widely across the clade (Williams et al., 1996). Brachiopods are a morphologically diverse group, with a broad range of life habits (pediculate, cemented, free-living); they range in depth from abyssal to intertidal, and range geographically from poles to tropics. Their stratigraphic record is rich and dense (highly "complete" sensu Foote and Sepkoski, 1999), with individuals abundant; their near-ubiquitous presence is familiar and comforting to most students of the fossil record, particularly in the Paleozoic era.

Brachiopod systematics has traditionally been dominated by evolutionary paleontologists and neontologists. With the benefit of hindsight, in the light of new data and techniques, some of the classifications adopted over the years (discussed below) may be viewed as misguided, but they were almost universally focused on characterizing perceived evolutionary patterns, rather than simply organizing observable diversity. In this sense, brachiopods have always been studied as evolving organisms and not as objects to be categorized; consideration of homology and polarity has long influenced interpretations of character distribution.

Because of recent reviews of the classification of brachiopods (Williams et al., 1996, 2000) that build on excellent older, comprehensive reviews (Muir-Wood, 1955; Williams and Rowell, 1965b), I have chosen to approach this contribution more like an essay than a conventional review article, with the hope that it will initiate discussion of general systematic issues as they apply to invertebrate paleontology, in addition to generating more focused empirical contributions to the systematics of brachiopods.

MAJOR HISTORICAL DEVELOPMENTS

Most brachiopod classification schemes have been dominated by the asserted evolutionary importance of a few key characters or character complexes (Williams, 1956). Six characters have played a particularly significant role in structuring the major groups of brachiopods: nature of valve articulation; type of shell mineralogy; relationship of the pedicle to the shell; embryological origin of the pedicle; whether or not the mantle reverses during development; and the presence or absence of an anal opening. States of these characters are generally congruent with one another (for example, see Table 1), apparently indicating patterns of common ancestry, but some incongruences (homoplasies) exist and, of course, focus disagreement among those convinced that one character is more informative about relationships than another. Such characters have thus been weighted in some fashion to emphasize their relative importance, and differences in classifications over time generally reflect these differences in opinion.

States of only the first two (or possibly three) of these six characters can be determined unambiguously from fossil material alone, reflecting the biological training of many of the earliest brachiopod systematists. However, it has long been recognized that the extant diversity of brachiopods is far lower than their extinct diversity, and reconciling these two different sources of information about relationships has haunted the construction of brachiopod classifications. In other words, extant brachiopods are known from only a small number of species, but each one can be characterized by a relatively large number of diverse types of features (Fig. 1). Extinct brachiopods, on the other hand, are known from a much larger number of species, but they are characterized by skeletal features alone. How far can we or should we generalize about the presence or absence of unfossilizable features in extinct representatives?