Downloaded from http://sp.lyellcollection.org/ by guest on September 26, 2021

Unravelling the evolutionary biology of the : a multidisciplinary approach

E. M. HARPER l, J. D. TAYLOR 2 & J.A. CRAME 3 1 Department of Earth Sciences, Downing Street, Cambridge CB2 3EQ, UK (e-mail: emh21 @cus.cam.ac.uk) 2 Department of Zoology, The Natural History Museum, London SW7 5BD, UK British Antarctic Survey, Madingley Road, Cambridge CB3 0ET, UK

Bivalves have been important members of marine taxonomic diversification of the bivalves (Pojeta communities since the early Palaeozoic, in terms of 1978) and the rostroconchs (Runnegar 1978) are both their numerical abundance and diversity. They still widely cited. However, in 1977 the Treatise are particularly prevalent in shallow shelf volumes (Cox et al. 1969; Stenzel 1971) were still sediments, but they have also conquered the very much in vogue as a reliable data source, intertidal zone as well as the deep sea, where they although even then there was a feeling that it was in are successful predators and key components of need of a comprehensive revision (Yonge 1978). some vent communities. They have also invaded This sentiment has been echoed ever since, most freshwater systems a number of times, where today strongly by Johnston & Haggart (1998) in their they are important (and costly) foulers. In terms of introduction to Bivalves: An Eon of Evolution. general community structure, bivalves are Paleobiological Studies Honoring Norman D. important as prey items for a range of different Newell. The Royal Society volume was also written predatory groups, and as major space occupiers, at a time when cladistic studies were virtually particularly on hard substrata where space may be unknown and there was not the wealth of molecular limited. techniques at hand to tackle questions of The abundance and diversity of both Recent and phylogeny. The enormous changes which have bivalves have made them attractive subjects taken place in the way in which evolutionary for both zoologists and palaeontologists, and both processes are thought about has changed during the disciplines have contributed to the present system last 20 years, as have the number of taxa, both of classification and the understanding of their Recent and fossil, that have been discovered during phylogenetic relationships. However, somewhat this period, both of which have inspired this inevitably, the focus of the two groups has been volume of multidisciplinary papers from both rather different, with zoologists concentrating on palaeontologists and zoologists. anatomical characters such as those associated with the gills and stomach, whilst palaeontologists have Bivalve classification and phylogeny necessarily dwelt on hard-part characters such as dentition and shell microstructure. It is becoming Despite the long scientific interest in bivalves, the increasingly clear, however, that convergence and relationships between the higher taxa are perhaps parallelism is rife within the class and more surprisingly unresolved. Certainly the Treatise integrated approaches are necessary to unravel classification is now dated, but even more recent these. schemes are often deeply conflicting, probably The last major attempt to integrate the palaeo- because of their reliance on single or small groups ntological and zoological approaches to bivalve of characters (Starabogatov 1992; Morton 1996; evolution was more than 20 years ago, at the Royal Salvini-Plawen & Steiner 1996; Waller 1998; Society of London meeting in 1977. The resulting Amler 1999). The relatively recent boom in volume of the Philosophical Transactions of the molecular techniques is of great potential here and Royal Society of London, Volume 284, is still a two papers in this volume have used them to try to benchmark volume. In particular, an early attempt elucidate the relationships between higher taxa. at a cladistic analysis of the Pteriomorphia by Steiner & Hammer have used an analysis of 18S Waller (1978) and papers on the origin and rDNA sequences in order to tackle the phylogenetic

From: HARPER,E. M., TAYLOR,J. D. & CRAME,J. A. (eds) The Evolutionao' Biology of the Bivalvia. Geological Society, London, Special Publications, 177, 1-9. 1-86239-076-2/00/$15.00 9The Geological Society of London 2000. Downloaded from http://sp.lyellcollection.org/ by guest on September 26, 2021

2 E.M. HARPER ET AL. relationships of the pteriomorphs, both in terms of character refinement may reduce the numbers of the major pteriomorph taxa (such as the ostreoids, possible models. The paper is accompanied by full pinnoids and pterioids, etc.), and between the character descriptions and the data matrix which Pteriomorphia as a whole and the other major will allow such future re-analysis. Despite the large bivalve clades (e.g. protobranchs, heteroconchs, number of trees produced, the consensus trees do anomalodesmatans). Their researches supports the introduce a number of interesting conclusions, monophyly of the , Heteroconchia some of which support the conclusion of previous and Pteriomorphia, and float the idea that the studies, such as Waller (1998), on the basal position Heteroconchia and Pteriomorphia may be sister of the Mytiloidea amongst the pteriomorphs, and taxa. Within the pteriomorphs, there is strong the basal dichotomy within the palaeotaxodonts support for two major clades, ([Pinnoidea between the Nuculanoidea and the Nuculoidea + (Ostreoidea + Pterioidea)] and [(Anomioidea + Solemyoida. They also support the view of shared Plicatuloidea) + (Limioidea + Pectinoidea)]. ancestry of the Nuculoidea and the palaeo- Further molecular studies are presented by heterodonts. In other areas, however, the authors Campbell, also using the 18S gene, who come to rather different conclusions to those investigated the relationships between a wide range reached by others. The most striking of these is the of bivalve superfamilies, orders and subclasses. rejection of the popular notion that the Bivalvia Gratifyingly, he found all superfamilies, which were derived from the rostroconchs, preferring have been established on morphological grounds, instead a stenothecoid monoplacophoran origin. appeared monophyletic, as were most orders. Only Elsewhere they are at odds with Waller (1998), the Myoida appear polyphyletic, thus confirming rejecting a link between the Anomalodesmata and the suspicions of some previous workers (e.g. the Heteroconchs, preferring to place the former as Morris et al. 1991). a sister group to the pteriomorphs. Again, there are Palaeontologists face the fascinating but difficult striking anomalies between the results of this challenge of trying to test various models of bivalve analysis and those of molecular studies. Cope has phylogeny with the known fossil record. At face used a more traditional approach, based on the value the task should be relatively simple as the culmination of detailed research of a range of early bivalves have, by virtue of their robust calcareous bivalve faunas. He bases his scheme on valves, an excellent fossil record, but this has not dentition, shell microstructure and inferred gill proved to be the case. It is particularly difficult grade. Cope identifies the evolution of the because following the good fossil record of the filibranch gill, which he infers from dental changes, earliest widely accepted bivalves (Fordilla and as providing the initial impetus for the Pojetaia), which are relatively abundant from the diversification of the bivalves in either the latest Tommotian of North America, Europe and or earliest Ordovician. He recognizes Australasia, there is an unfortunate gap in the fossil two subclasses; (1) the Protobranchia- including record of the class that spans from the middle the Nuculoidea and chemosymbiotic forms such as Middle Cambrian-Early Ordovician, some 4% of the solemyoids; (2) the Autolamellibranchiata - the entire evolutionary history of the class (Harper from which he derives three principal stocks, the 1998). Unhappily, this gap in the fossil record Trigonioida, the Anomalodesmata and the appears to cover a period of major evolutionary Heteroconchia, which he believes then gave rise to changes within the class, for on their reappearance the Pteriomorphia. the bivalves are larger and more diverse both Two other papers in the volume deal with the taxonomically and in life habits. None the less, new phylogenetic relationships of enigmatic higher discoveries in the last decade have improved bivalve taxa. In a major new synthesis, Skelton & current knowledge of early Palaeozoic bivalve Smith present a first attempt at a comprehensive faunas. These findings have inspired two papers in phylogeny for the , an extinct group of this volume which tackle the problem in two rather sessile, epifaunal bivalves that flourished in warm different ways. temperate and tropical seas from the Late Carter et al. have undertaken an ambitious to the Late . Their cladistic study was cladistic analysis of 62 bivalve taxa of Palaeozoic based on 33 skeletal characters and 31 species, age, including a number of equivocal forms such as representing each of the higher taxa recognized in Tuarangia, using 117 characters derived almost previous classification schemes. One of the ten exclusively from the hard-parts. The large number most parsimonious trees obtained from their study of characters (and states) and taxa has limited this was selected as a working hypothesis. This showed study to a heuristic analysis, and the analysis has many of the basic phylogenetic relationships revealed 411 most parsimonious trees. Doubtless an established by early workers, such as Douvill6 and increase in computer power in the next few years MacGillavry, were perfectly sound. An outer shell will allow more exhaustive searches and further layer of fibrillar prismatic calcite is a diagnostic Downloaded from http://sp.lyellcollection.org/ by guest on September 26, 2021

INTRODUCTION 3 synapomorphy for all rudists and attachment to the disjunct geographic distributions, they have substrate by one or the other valve provides a traditionally been placed within a monophyletic fundamental division into two stocks. Other family, the Etheriidae. Bogan & Hoeh have features shown by the cladogram include the integrated new cytochrome C oxidase subunit I monophyletic status of the uncoiled rudists, as well DNA sequences from these three taxa with existing as established families such as the Radiolitidae and data from other unionid taxa from around the Hippuritidae. Important new light is shed on the world. Their findings do not support a close origin of the latter family by its sister-group relationship between the three genera, suggesting relationship with the polyconitid Tepeyacia, rather that the cemented habit has arisen at least from the Albian of Mexico. There is every prospect twice. They also pose the rather interesting that future iterations of this study will refine rudist question as to why the cemented habit might have phylogeny even further. evolved in the first place, noting that there are no Deep-sea carnivory is just one of some rather recorded freshwater cementers in either North unusual lifestyles found amongst members of the America or China, the major hubs of unioinoid bivalve subclass Anomalodesmata. This group diversity. comprises 14 extant families of morphologically Comparative morphology continues to be an disparate bivalves, many from the deep sea, whose important tool in our efforts to determine classification and phylogenetic relationships have phylogenetic relationships amongst bivalves. A been highly confused. Harper et al. use a cladistic major benefit from the introduction of cladistic analysis of a set of anatomical and shell characters methods in phylogenetic reconstruction has been in in an attempt to determine relationships amongst the objective definition of characters and their this diverse group. One conclusion from this states. Concomitantly, there has been a search for analysis is that carnivory has arisen independently new types of characters and two papers show how within two separate clades. Trees generated from detailed morphological studies of bivalve sperm, shell characters alone were incongruent with those using transmission electron microscopy, can derived from the total evidence of anatomy and generate sets of phylogenetically important shell, suggesting that although anomalodesmatans characters. In the first paper, Healy et al. use sperm have a long fossil record it may be difficult to characters to explore relationships within the large integrate into such an analysis. bivalve subclass Pteriomorphia, e.g. showing how Bivalves play a major role in freshwater the Ostreoidea and Limoidea are probably closely communities, where they are significant in water linked, as are the Pterioidea, Pinnoidea and filtering, and their larval phases are important Pectinoidea. By contrast, sperm characters suggest parasites of fish, whilst their propensity to spread that the Arcoidea and Limopsoidea are less closely has made them important biofoulers (Aldridge related than generally supposed. At a more detailed 1999). Because of the nature of freshwater bodies, level, Keys & Healy use characters of sperm many taxa are endemic to specific regions and ultrastructure to explore relationships amongst the hence it is probable that they, above all other living species of giant clams - Tridacninae. Sperm bivalves, may show extreme degrees of morpho- data supports a close relationship between giant logical convergence. Similarly, many geographi- clams and cardiid bivalves, but perhaps indicates a cally restricted taxa are at great danger from separate origin of Hippopus and the other Tridacna extinction. Lydeard et al. have used the species. Amongst the species of Tridacna, T. mitochondrial cytochrome C oxidase subunit I and squamosa shares more characters with T. gigas 16S rRNA gene sequences to examine the than with the other species in the subgenus, phylogenetic relationships of five species of Gulf Chametrachea, where it is usually placed; a Coastal unionid generally accepted as belonging to conclusion supported by some molecular data. three different genera. These are particularly important taxa as they are all considered Bivalve form and function endangered. Their analyses found that the five taxa were all distinct and the 'genera' in which they had The great diversity of bivalve life habits is reflected previously been dispersed in were polyphyletic. in a wide range of morphological variation in terms The threat of morphological convergence ham- of both shell and anatomy. The seminal work of pering classification is an issue with the three Stanley (1970) provided a firm basis for the use of unionid genera of cementing 'freshwater oysters' shell characters to infer the life habits of extinct [Etheria (Africa), Acostaea (South America) and taxa and few have approached the elegance of the Pseudomulleria (India)]. Despite the fact that the work of C. M. Yonge in demonstrating the variation cemented habit within the Bivalvia as a whole is in anatomical features across the class. Despite the well known to be polyphyletic (Yonge 1979) and established interest in bivalve form and function, that these freshwater cementers have highly there remains a surprising amount to do and, in Downloaded from http://sp.lyellcollection.org/ by guest on September 26, 2021

4 E.M. HARPER ET AL. particular, to synthesize the information. It is too. However, the most distinctive bivalve important to document the morphological diversity within the assemblage is a large, new taxon which displayed by the class and to investigate the is referred to the anomalodesmatan family constraints and preadaptations which have favoured Modiomorphidae. This new bivalve reaches some body plans and not others. c. 300 mm in length by 120 mm in height, and has Taylor & Glover report that all the species of a shell which is, in places, up to 28 mm thick. It Lucinidae studied to date harbour sulphide- seems to be yet another new giant bivalve taxon oxidizing, chemosymbiotic bacteria within their associated exclusively with cold seeps. Members of gills. Although chemosymbiosis in living lucinids the Pectinidae possess the anatomically most has been recognized for some time, it seems that advanced eyes of any bivalve family and it would little thought has gone into its mechanics - if seem only logical to conclude that they have sulphide-laden water is drawn into the mantle developed these to provide an advanced early cavity alongside the inhalant respiratory current, warning system of vicinal predators. However, in a how is oxidization prevented before reaching the wide-ranging review of the functional significance chemosymbionts? Taylor & Glover's careful work of pectinid pallial eyes, Morton suggests that their revealed a number of adaptations by which the two true purpose is still largely unknown. He first adds sites of respiration and chemosynthesis are to our anatomical knowledge of pectinid eyes by partitioned, e.g. by the detached anterior muscle describing those of the boreal species Patinopecten which separates the oxidized water that come in yessoensis and then goes on to show that in almost anteriorly from the sulphide-rich waters which are all taxa studied an escape response can only be drawn in posteriorly. The presence of mantle gills elicited by either tactile or chemical stimuli. Visual on the internal surface of the anterior of the mantle clues rarely, if ever, provoke active swimming. Is of many taxa underlines this 'switch' in the location the pectinid eye an example of evolutionary of normal respiratory activity. Mantle gills had been overdesign? If pectinid and spondylid eyes are described previously, but only superficially and indeed homologous, it would suggest that they their complexity and true significance had never evolved before the two families diverged in the been realized. This underlines the continued Early Jurassic. Sophisticated pallial eyes may have importance of anatomical studies of seemingly been retained since then for a number of poorly familiar bivalves. Such is the importance of the understood functions. chemosymbiotic relationship, that evolutionary Few organisms have evolved the capability of syntheses of the lucinoids should take account of it digesting wood, but the teredinid bivalves, the and it is therefore extremely useful to now have shipworms, have been successful in this mode of morphological characters which may be used as feeding, dating back to the earliest Cretaceous, a markers for the habit. Many of these features are time when angiospermous wood would have demonstrable in the Ilionia suggesting that become abundant in coastal habitats. Indeed, in the chemosymbiosis within the lucinoids is Recent ecosystems teredinids have been the extremely ancient. scourge of human-made coastal structures and In recent years it has become apparent that the boats made of wood. However, relatively little is very distinctive chemosymbiotic bivalve communi- known about the evolution of this very specialized ties associated at the present day with both hot habit. Lopes et al. have described and compared vents and cold seeps can be traced back through the functional anatomy of the digestive system of considerable periods of time. Such faunas seem to two wood boring taxa, Neoteredo reynei (Bartsch, have been particularly widespread in the 1920) and Psiloteredo healdi (Bartsch, 1931), Cretaceous and Kelly et al. record a further which are common in Brazilian mangroves. occurrence in the Early Cretaceous (Late Important differences between the two suggest that Barremian) of Wollaston Forland, Northeast N. reyenei, with its enlarged wood-packed appendix Greenland. Sedimentological evidence indicates and anal canal, is more dependant on wood for its that this is a methane cold-seep complex which nutrition than is P.healdi, in which these structures accumulated in a mid- to outer-shelf setting. It is are small and where active palps have been characterized by a large number of limestone retained, suggesting the importance of suspension mounds, 1-3 m in diameter and up to 1.8 m high, feeding. There is evidence that during ontogeny that occur within an otherwise monotonous, wood may increase in importance in the diet of the mudstone-dominated sequence. The mounds reveal latter. an internal structure of calcite-cemented tube Gill structure in bivalves has long been the focus systems, laminated calcite crusts and void fills. A of functional and evolutionary studies, and large new species of the lucinid genus Cryptolucina Beninger & Dufour describe and discuss the fine is a dominant element within the fauna, and there structure of the abfrontal surface of the gill are occasional specimens of the cryptodont filaments. From the supposed primitive function as Downloaded from http://sp.lyellcollection.org/ by guest on September 26, 2021

INTRODUCTION 5 a mucociliary cleaning surface in protobranch number of prodissoconchs from Middle Jurassic bivalves, two evolutionary trajectories involving oysters. Careful examination of their morphology cilia and mucocytes are proposed. One involves the revealed that the ancestral larval hinge most likely progressive reduction in density of cilia and evolved from a mytilid-like provinculum, to mucocytes, and the other involves reduction of cilia produce a secondarily symmetrical hinge line. He but retention or increase in acid mucopoly- also suggests that the small size of the pro- saccharide-secreting mucocytes. dissoconch I, along with the small P I/P II ratios, Many arcoid bivalves are characterized by a confirm that the planktotrophic development style duplivincular ligament, wherein the lamellar layer seen in modern oysters is in fact the plesiomorphic is embedded in the fibrous one in a distinctive state for the entire superfamily. series of chevrons. The one prominent exception to The tube-dwelling clavagellid bivalves are some this type of organization occurs in the family of the most enigmatic of all bivalves and as Noetiidae, which is characterized by vertical, indicated by Harper et al., their position within the chevron-shaped strips of lamellar ligament. At first subclass Anomalodesmata is extremely equivocal, sight it would appear that these two types of their extreme morphological adaptations apparently ligament must be the product of two quite distinct masking useful phylogenetic signals. Savazzi has groups, but chance observations on a series of investigated the little-known clavagellid Bryopa. Limopsis ligaments suggested to Thomas et al. that Unlike most clavagellids, which are thought to be striking differences in the form of the arcoid facultative semi-endolithic tube dwellers, Bryopa ligament can be produced by only a very modest appears to be fully and obligatorily endolithic. change in the developmental process. To develop Despite the fact that the left valve is permanently this idea further they have hypothesized that, along attached to the tube, the bivalve appears to migrate the mantle isthmus, alternate zones of activation forwards within the substratum throughout growth. and inhibition control the production of lamellar Despite the fact that no living , or indeed ligament that forms either oblique sheets or vertical soft parts, were available for study, Savazzi was layers; only a single, simple change of instructions able to use morphodynamic principles to suggest to a narrow field of cells is required to change from how this could occur. He envisages a mechanism by one pattern to the other. Tests provided by two which the shell elongates anteriorly, sliding the soft contrasting sets of mathematical models indicate parts forward, whilst the posterior portion of the that this may indeed be the case and it may be right valve is continually being resorbed. The concluded that disparate growth patterns can in fact strong inequivalvy which results is seen in none of be produced by only subtle character shifts. the other clavagellids and, unfortunately, it seems With the current revolution in the understanding that Bryopa is too derived to provide the much of which taxonomic groups may or may not be needed information as to the affinities of the assigned to the subclass Cryptodonta (Johnston & superfamily. Collom 1998), considerable importance is now Although the mineralization of the external attached to a proper understanding of the shells of bivalves has been extensively studied and Palaeozoic order Praecardioida. Yancey & Heaney used as a phylogenetically useful character (Taylor present some timely information on the et al. 1969, 1973; Carter 1990), some internal Lunulacardiidae, the youngest praecardioid family tissues are also mineralized and freshwater mussels in North America. Their material comes from the (Unionoida) often have large accumulations of Pennsylvanian (Upper ) Buckhorn calcium granules in their tissues. Byrne describes Asphalt biota of south-central Oklahoma, and the fine structure and chemistry of abundant comprises a unique assemblage of small bivalves. orange, iron-rich calcium granules in the Some of these are definitely juveniles but others are Australian freshwater mussel Hyridella. adults that can be assigned to an unusual new taxon. Comparison of granule distribution and com- Characterized by a distinct change in shape from a position between different unionoids suggests the veneriform juvenile to a subtrigonal adult, this new existence of phylogenetically significant differ- form possesses both stout teeth and a prominent ences. ligament. Indeed, were it not for these latter Teranota is a relatively newly described taxon, characters, it could easily be confused with a known only from the Middle of conocardioid rostroconch ! Germany. These scimitar-shaped bivalves are The morphology of larval shells of both living interpreted by Rogalla & Amler as belonging to and fossil bivalves has the potential to supply the enigmatic family Orthonotidae, whose affinities useful phylogenetic information, particularly are debated as belonging to either the modio- where, as in the case of the oysters, the post-larval morphids, anomalodesmatans or even the shell is prone to a large degree of ecophenotypic heteroconchs. Rogalla & Amler are persuaded of an variation. Malchus has recognized and studied a anomalodesmatan affinity but support a close link Downloaded from http://sp.lyellcollection.org/ by guest on September 26, 2021

6 E.M. HARPER ET AL. with the modiomorphids. The discovery of quality and completeness of the source data from specimens in (apparently) life position, embedded regional levels. An excellent example of how these anterior-end downwards at angles of 60-80 ~ to data are acquired is the analysis by Mikkelson & the sediment surface, has allowed them to Bieler of the bivalve fauna of the Florida Keys. reconstruct these as endobyssate bivalves with no Prior to their survey only 163 species were true siphons. recorded from the area. Using a combination of original collecting, search of museum collections Biodiversity and biogeography and scouring of literature records this list has now been increased to 325 species. Interestingly, but The documentation and analysis of patterns of rather worrying, primary literature sources biodiversity on regional and global scales, and, in recovered only 44% of the species, although this particular, the latitudinal diversity gradient, have increased to 73% when 'grey' literature was generated much interest, a plethora of hypotheses included. Museum collections were the most and considerable controversy. Marine bivalves important source of records, picking up 77% of the have been often used in such analyses by both species including 62 not found by other methods. biologists and palaeontologists. A new analysis of The message from this paper is that a diversity regional bivalve faunas by Crame shows that both database for a particular area based solely on latitudinal and longitudinal gradients are not as literature records is likely to be a severe regular in form as supposed. There is a distinct step underestimate of the true diversity. between 20 and 30~ in the latitudinal diversity Molecular techniques are now being used in gradient for the northern hemisphere and in the biogeographical analysis to test hypotheses con- southern hemisphere the bivalve fauna of Australia cerning the area of origin and movements of forms a distinct hotspot of diversity. The causes of species. The smooth blue mussel, Mytilus these large-scale patterns are likely multiple and galloprovincialis, has an unusually widespread complex, and Crame critically reviews various distribution in temperate and subpolar waters of hypotheses emphasizing that historical processes in both the northern and southern hemispheres. the Neogene, such as the closure of Tethys and the Introduction by humans may account for some of northward movement of Australia into the tropics, the distribution pattern but mussels are present in have been important in shaping the biodiversity shell middens predating European arrival in patterns of a region. southern Australia and natural dispersal also seems Jablonski et al. have also been investigating the likely. Daguin & Borsa used nuclear DNA markers latitudinal gradient in marine bivalve diversity, to determine the genetic relationships and tease paying particular attention to that present on the apart the history of Mytilus galloprovincialis eastern Pacific continental shelf. Here, their very populations around the world. They demonstrate comprehensive data set also picks out a sharp the genetic distinctiveness of the Australian popu- (50%) fall in the number of species at the edge of lations and reject the hypothesis that these mussels the tropics, followed by a much gentler decline into were recently introduced by humans. By contrast, Arctic regions. They believe that this gradient may the close similarity of samples from the be a positive function of available energy and, Mediterranean, the North Pacific and Chile using a residuals analysis to counter the effects of supports the hypothesis that mussels in the latter spatial autocorrelation within the data set, demon- two areas have been introduced recently. strate that mean sea-surface temperature is a highly significant predictor of bivalve diversity. It is also Ecological and evolutionary trends apparent that both infaunal and epifaunal taxa increase in diversity from the poles to the tropics, Bivalves have a crucially important role to play in although at somewhat different rates. Along the the study of evolutionary lineages. Crampton & Pacific coast of North America their ratio changes Maxwell argue that outline shape should be a key significantly with latitude and there is preliminary morphological character and proceed to develop a evidence from the Mesozoic fossil record to new method for studying it based on Fourier shape suggest that it may also change with time. Clearly, analysis. An outline trace of either a growth stage this is a dynamic gradient driven by the differential or the complete adult valve is digitized, yielding a diversification of the two functional groups. Do suite of Fourier coefficients that describe a factors such as the mass extinction event at the K-T spectrum of harmonically related trigonometric boundary and global climatic change have a role to curves, or harmonics. These coefficients can then play in driving these clade dynamics? be used to compare synthetic average morphologies Analysis of global patterns of marine bio- at a variety of ontogenetic and evolutionary stages. diversity as exemplified in the Crame and In a detailed study of the shallow-burrowing Jablonski et al. papers is highly dependent on the crassatellid genus, Spissatella, it was shown that Downloaded from http://sp.lyellcollection.org/ by guest on September 26, 2021

INTRODUCTION 7 growth is strongly allometric. Ontogenetic changes significant elevation of whole-organism metabo- in shape of individuals are invariably far greater lism at low temperatures. The restricted temper- than total evolutionary changes within the genus ature envelope and low metabolic rates of polar over c. 20 Ma. An overall strong correspondence ectotherms are adaptations that have evolved over between individual species ontogenies and evolu- at least the last 15 Ma, if not considerably longer. tionary changes in adult form strongly suggest that Edelaar has been investigating whether the heterochrony has been the dominant evolutionary depth of burrowing in the deposit-feeding tellinid mechanism throughout the lifespan of Spissatella. Macoma balthica might be regarded as a trade-off There is good evidence to suggest that paedo- between safety from predators and access to its morphosis and peramorphosis occurred in food source. Obviously, the deeper Macoma approximately equal frequencies from the latest burrows the safer it is from surface-dwelling Oligocene onwards. predators, but the more remote it is from its food. Johnson et al. have highlighted the potential of The key question of interest here is whether the Queen Scallop, Aequipecten opercularis, as a organisms such as Macoma can adjust to optimum tool to investigate Late Cenozoic palaeoenviron- values by modifying their phenotype (such as depth mental change. Occurring at the present day from of burrowing). To try and answer this, a series of Norway to the Adriatic, it has rapid, year-round laboratory experiments was devised to investigate growth rates that ensure recovery of well-resolved depth of burrowing with, firstly, food (finely records of seasonal variation, and a fossil record chopped spinach) either present or absent, and then that stretches back into the Miocene. Somewhat a common predator (the shorecrab, Carcinus surprisingly, an oxygen isotope palaeothermometry maenas) either present or absent. The results study on mid-Pliocene shells indicated seasonal showed a significant decrease in burrowing depth temperatures similar to those of the present day. when food was added but then a significant This stands in marked contrast to a variety of other increase when a predator was introduced. On isotope and general assemblage data which average, individual Macoma burrowed 57% deeper suggests that the mid-Pliocene was globally in the presence of the shorecrab. Such behavioural warmer than now. Perhaps the southern North Sea plasticity may have a number of important Basin was in some way exempt from this general ecological and evolutionary implications. For trend, or maybe even these well-preserved fossils example, pronounced changes in the shape or suffered a degree of cryptic diagenetic alteration. density of growth tings might reflect predation More promising in this particular respect may be intensity as much as a low food supply. the study of microgrowth increments, for these do Ensis directus, a recent ballast-water invader show markedly higher maxima and minima (i.e. along European North Sea coasts, shows periodic related to extreme summer and winter temper- mass mortalities that are still poorly understood. atures) in the mid-Pliocene. Carbon isotope and Cad~e has noticed how, at certain times during the various trace element data also offer considerable winter, large numbers of shells become exposed by potential for studying variation in seasonality at least half their length above the sediment surface patterns. and then cannot re-burrow. At such times they The evolution of physiological characters has become easy prey for herring gulls, who have been investigated experimentally by Peck & perfected a technique of alternate shaking and Conway. They were particularly concerned with dropping to break open the shell. This invariably the concept of metabolic cold adaption (MCA), causes damage to just the middle part of the valve which is the hotly debated subject of whether the and the production of a unique type of shell metabolic rates of polar ectotherms are elevated to fragment. Many types of shell fragment are now compensate for the physiological constraints known to be specific to individual predators and imposed by living at low temperatures. The recent thus have considerable potential for palaeoeco- demonstration that the muscles of certain polar fish logical studies. Presumably, avian predators exert a contain twice the number of mitochondria as significant predation pressure on intertidal and temperate species is taken by some to be evidence shallow subtidal bivalves and, therefore, the for the operation of MCA. The approach adopted in recognition of diagnostic shell damage caused by this study was to look at rates of oxygen con- these predators, which is demonstrably different sumption in two Antarctic bivalve species and then from taphonomic damage, is extremely useful in to compare the results with those obtained previ- any future attempts to document this activity from ously from a range of lower latitude taxa. In this the fossil record. way it was possible to show, quite conclusively, Marine mytilid mussels are probably the most that rates of oxygen consumption decline with familiar and well-studied of all bivalves. They have increasing latitude. Together with a similar study of a huge biomass and are major space-occupying Q10 coefficients, the data indicate that there is no organisms upon rocky intertidal shores throughout Downloaded from http://sp.lyellcollection.org/ by guest on September 26, 2021

8 E.M. HARPER ET AL. the world, where they are often important com- mantle edges to mimic fish, complete with fin-like ponents of intertidal food webs. Seed et al. review structures and 'eyes', and also extruded structures the attributes that contribute towards the ecological containing the glochidian larvae which mimic fish success of mytilid mussels. They then go on to and arthropod larvae (Haag et al. 1995). show how mussel beds themselves form major A remarkable recent discovery is that the worm- habitats for many other organisms, with the like Xenoturbella, a ciliated, virtually composition and structure of these communities organless bag from northern Europe, is in fact a remarkably similar in widely separated parts of the highly modified shell-less protobranch bivalve world. Because of their abundance and accessi- (Israelsson 1999). This umasking of Xenoturbella bility, as well as their importance as food, mussels extends present conceptions of the morphological have been widely used as sentinel indicators of range of bivalves. environmental change and pollution. Seed et al. Palaeontologists also continue to unearth demonstrate, by detailed studies of shell growth wonderful and extraordinary animals which truly and trace element analysis, how mussel shells can extend our concepts of morphological disparity be used as chronometers of environmental change. amongst bivalves, e.g. the discovery of a new Their work provides a model for the interpretation family, the Wallowaconchidae, of large-vaned and of growth patterns in fossil shells. chambered bivalves (Yancey & Stanley 1999), which may have harboured microbial symbionts. Whilst debate still continues over the Future directions life habits of enigmatic extinct taxa such as the It would be easy to presume that the long tradition rudists and inoceramids, new discoveries are of bivalve study by both zoologists and palaeont- gradually extending our knowledge of early ologists, and their basic familiarity with the class, Palaeozoic bivalve evolution and continually has led to a comprehensive understanding of their extending, or otherwise modifying, the geological evolutionary history. However, this is far from the ranges of known groups. The most recent published case. The start of the twenty-first century offers compilation of familial ranges (Skelton & Benton many exciting challenges to the understanding of 1993) is strikingly different from that published in the evolutionary biology of the Bivalvia. the Treatise and even then its authors admit that its Exciting discoveries continue to be made quality is rather 'patchy'. amongst living bivalves, and two in particular have One of the most obvious short-term challenges is changed the concepts of trophic adaptations and to produce a comprehensive compilation of all the evolution of the class. Bivalves had long been available phylogenetic information and geological regarded as having two main feeding strategies - ranges and the present authors echo the calls by suspension and deposit feeding - with a few species Yonge (1978) and Johnston & Haggart (1998) for nutritionally dependent on having photosymbiosis an updated version of the Treatise. with unicellular algae. It is only 20 years since the discovery of predatory behaviour amongst bivalves References and this is now recognized as a major feeding strategy amongst at least three groups of deeper ALDRIDGE, D. C. 1999. The morphology, growth and water bivalves (Morton 1981). Similarly, explor- reproduction of Unionidae (Bivalvia) in a Fenland ation of hydrothermal vents and cold-seep sites waterway. Journal of Molluscan Studies, 65, 47-60. AMEER, M. R. 1999. Synoptical classification of fossil and over the last 20 years has revealed communities of Recent Bivalvia. Geologica et Palaeontologica, 33, bivalves nutritionally dependent on symbiotic 237-248. sulphur- and methane-oxidizing bacteria. It is CARTER, J. G. 1990. Skeletal Biomineralization: Patterns, remarkable that this chemosymbiosis has now been Processes and Evolutionary Trends. Van Nostrand recognized in at least six different clades and many Reinhold, New York. species of bivalves, living not only in the deep sea Cox, L. R. & 24 others. 1969. In: MOORE, R. C. (ed.) but also in shallow water and even intertidal Treatise on Invertebrate Paleontology. Part N. habitats (Distel 1998). Comparative morphology (1-2), Bivalvia. Geological Society of suggests that this chemosymbiosis is an ancient America, Boulder, CO, and University of Kansas Press, Lawrence, KS. nutritional strategy which may date back to the DISTEL, D. L. 1998. Evolution of chemautotrophic Ordovician. endosymbioses in bivalves. Bioscience, 48, The biology of these humble animals continues 277-286. to surprise, as witnessed by the extraordinary HAAG, W. R., BUTLER, R. S. & HARTFIELD, P. D. 1995. An behavioural and morphological adaptations extraordinary reproductive strategy in freshwater employed by freshwater unionid bivalves to ensure bivalves: prey mimicry to facilitate larval dispersal. transfer of their larvae to host fish species. These Freshwater Biology, 34, 471-476 remarkable adaptations involve extensions of the HARPER, E. M. 1998. The fossil record of bivalve Downloaded from http://sp.lyellcollection.org/ by guest on September 26, 2021

INTRODUCTION 9

molluscs. In: DONOVAN, S. K. & PAUL, C. R. C. Rostroconchia, Scaphopoda and Bivalvia. In: (eds) The Adequacy of the Fossil Record. John BENTON, M. J. (ed.) The Fossil Record 2. Chapman Wiley and Sons, Chichester, 243-267. & Hall, London, 237-263, ISRAELSSON, O. 1999. New light on the enigmatic STAROBOGATOV, Y. I. 1992. Morphological basis for Xenoturbella (phylum uncertain): ontogeny and phylogeny and classification of Bivalvia. Ruthenica, phylogeny. Proceedings of the Royal Society of 2, 1-25. London, 266B, 835-841. STANLEY, S. M. 1970. Relation of shell form to life habits JOHNSTON, E A. & COLLOM, C. J. 1998. The bivalve of the Bivalvia. Geological Socie~ of America heresies - inoceramidae are Cryptodonta, not Memoir, 125, 1-296. Pteriomorphia. In: JOHNSTON, P. A. & HAGGART, J. STENZEL, H. B. 1971. Oysters. In: MOORE, R. C. (ed.) W. (eds) Bivalves: An Eon of Evolution. Treatise on Invertebrate Paleontology. Part N. PaleobioIogical Studies Honoring Norman D. Mollusca (3), Bivalvia. Geological Society of Newell. Calgary University Press, Calgary, America Boulder, CO, and University of Kansas 347-360. Press, Lawrence, KS. -- & HAGGART, J. W. 1998. In: JOHNSTON, E A. & TAYLOR, J. D., KENNEDY, W. J. & HALL, A. 1969. The HAGGART, J. W. (eds) Bivalves: An Eon of shell structure and mineralogy of the Bivalvia. Evolution. Paleobiological Studies Honoring Introduction, Nuculacea-Trigonacea. Bulletin of the Norman D. Newell. Calgary University Press, British Museum (Natural History), Zoology Series, Calgary, xi-xii. Supplement, 3, 1-125. MORRIS, N. J., DICKINS, J. M. & ASTAFIEvA-URBA1TIS,K. & -- 1973. The shell structure and 1991. Upper Palaeozoic anomalodesmatan bivalves. mineralogy of the Bivalvia. II. Lucinacea- Bulletin of the British Museum of Natural History Clavagellacea, Conclusions. Bulletin of the British (Geology Series), 47, 51-100. Museum (Natural History), Zoology Series, MORTON, B. 1981. Prey capture in the carnivorous Supplement, 22, 253-284. septibranch Poromya granulata (Bivalvia: WALLER, T. R. 1978. Morphology, morphoclines and a Anomalodesmata: Poromyacea). Sarsia, 66, new classification of the Pteriomorphia (Mollusca: 241-256. Bivalvia). Philosophical Transactions of the Royal -- 1996. The evolutionary history of the Bivalvia. In: Socie~ of London, Series B, 284, 345-365. TAYLOR, J. D. (ed.) Origin and Evolutionar3, -- 1998. Origin of the molluscan Class Bivalvia and a Radiation of the Mollusca. Oxford University Press, phylogeny of the major groups. In: JOHNSTON, E A. Oxford, 337-356. & HAGGART, J. W. (eds) Bivalves: An Eon of POJETA, J. 1978. The origin and taxonomic diversification Evolution. Paleobiological Studies Honoring of pelecypods. Philosophical Transactions of the Norman D. Newell. Calgary University Press, Royal Society of London, Series B, 284, 225-243. Calgary, 1-45. RUNNEGAR, B. 1978. Origin and evolution of the Class YANCEY, T. E. & STANLEY, G. D. 1999. Giant alatoform Rostroconchia. Philosophical Transactions of bivalves in the Upper Triassic of western North the Royal Society of London, Series B, 284, America. Palaeontology, 42, 1-23. 319-333. YONGE, C. M. 1978. Introductory remarks. Philosophical SALVINI-PLAWEN, L. v. & STEINER, G. 1996. Transactions of the Royal Society of London, Series Synapomorphies and plesiomorphies in higher B, 284, 201. classification of the Mollusca. In: TAYLOR, J. D. -- 1979. Cementation in bivalves. In: VAN DER SPOEL, (ed.) Origin and Evolutionary Radiation of the S., VAN BRUGGEN, A. C. & LEVER, J. (eds) Mollusca. Oxford University Press, Oxford, 29-51. Pathways in Malacology. Bohn, Scheltema, SKELTON, E W. & BENTON, M. J. 1993. Mollusca: Holkema and Junk, Utrecht, 83-106.