Non-technical Summary
The Mollusca is the second largest animal phylum, which consists of two major groups that diverged during the Early Cambrian. These are the Aculifera (chitons and relatives) and Conchifera (snails, clams, squid, and their relatives). Remarkably, the Conchifera are early 100 times more diverse in terms of number of species than the Aculifera. This fact raises the question of what accounts for this remarkable difference in diversity. One hypothesis is that mate choice promotes species formation by making isolation among populations more likely. We argue that, although mate choice plays a role in Conchifera, the primary difference between the two major molluscan groups is the relationship between the soft body and the enclosing shell. In Aculifera, the body and shell are tightly coupled, whereas in the Conchifera, the body is much more flexible, extending into and out of the shell and permitting many forms of locomotion and other activities that are not possible in Aculifera. Decoupling of body parts is thus an important factor enabling evolutionary exploration of many modes of life that are unavailable to groups in which parts are tightly coupled. Although there are groups within Conchifera that are strikingly poor in species and in the diversity of forms, this conservatism appears to be a secondary consequence of ecological restriction to habitats or modes of life where high metabolic rates, which are associated with divergent forms, are strongly constrained. The points we make about mollusks likely apply to many other cases of divergence between a species-rich and a species-poor group as well.
Background
A recurrent pattern in evolutionary history is the striking inequality between sister clades in taxonomic, ecological, and phenotypic diversity (Vrba Reference Vrba, Eldredge and Stanley1984; Mitter et al. Reference Mitter, Farrell and Wiegmann1988; Wiens et al. Reference Wiens, Lapoint and Whiteman2015). The widespread occurrence of so-called pectinate phylogenies, in which successive branch points are between one lineage and a much larger clade, is one indication of this pattern, which appears at all scales of phylogenetic inclusion. Most studies of this inequality have sought to identify rare phenotypic, genetic, or ecological breakthroughs (“key innovations”) that occur in one but not the other clade and that correlate with species proliferation. These include herbivory in animals (Mitter et al. Reference Mitter, Farrell and Wiegmann1988) and habitat specialization (Vrba Reference Vrba, Eldredge and Stanley1984). Norm-breaking innovations, however, result in diversification in only a minority of cases (Wainwright and Price Reference Wainwright and Price2016; Vermeij Reference Vermeij2023b), and causal links between the trait in question and subsequent evolution remain elusive.
The phylum Mollusca exemplifies this contrast. Its two great lineages, Aculifera and Conchifera, share a common ancestry but diverged radically in evolutionary outcome (Chen et al. Reference Chen, Baeza, Chen, Gonzalez, Gonzalez, Greve, Kocot and Marbizu2025). Aculifera, now represented by around two thousand known living species, comprises the clades Caudofoveata, Polyplacophora, and Solenogastres, and is characterized by a serial organization with a dorsal sclerotome and ventral foot (when present). All are strictly marine benthic animals that are primarily grazers and predators of small animals. The Conchifera, with more than 120,000 living species, comprises Monoplacophora, Cephalopoda, and the much larger Megalopodifera (Chen et al. Reference Chen, Baeza, Chen, Gonzalez, Gonzalez, Greve, Kocot and Marbizu2025). The Megalopodifera comprises the living classes Bivalvia, Gastropoda, and Scaphopoda, as well as the extinct Helcionellida, Hyolitha, Paragastropoda, and Rostroconchia. It is characterized by a veliger larva, a retractable body in a one- or two-piece shell, and a large mobile foot (Chen et al. Reference Chen, Baeza, Chen, Gonzalez, Gonzalez, Greve, Kocot and Marbizu2025). About 95% of living molluscan species belong to Megalopodifera, which collectively occur in nearly all marine, freshwater, and terrestrial habitats (Ponder and Lindberg Reference Ponder and Lindberg2020). Conchiferan modes of life without aculiferan equivalents include suspension feeding, photosymbiosis, predation by drilling and envenomation, swimming, aerial locomotion, endoparasitism, algal gardening, and life in the plankton and neuston. In both scope and ecological reach, Conchifera, and in particular Megalopodifera, are extraordinary in their diversity not only relative to Aculifera but relative to almost all other metazoan lineages. Outside the arthropods, no other clade has explored as many environments or as many distinct adaptive architectures.
Within Mollusca, the comparison between the two clades Aculifera and Conchifera and among some lineages within these radiations provides a framework for understanding the origins of morphological disparity. It should be noted that aculiferan diversity is not small—chitons alone exceed the species richness of cephalopods, and the clade is larger in species richness than other midsized major clades such as Holothuroidea, Nemertea, Chiroptera, or Chondrichthyes. Aculifera seems limited mainly in that it is typically set against the extraordinary radiation of Conchifera, but it is precisely this contrast that enables an exploration of conchiferan exceptionalism.
One hypothesis is that the evolution of species-specific signals associated with mate choice and parent-controlled gamete fertilization promotes diversification. According to this hypothesis, minor variations in these cues arising locally can lead to genetic isolation and thus speciation and ultimately to taxonomic diversification (West-Eberhard Reference West-Eberhard1983; Bush et al. Reference Bush, Hunt and Bambach2016; Vermeij and Grosberg Reference Vermeij and Grosberg2018). Display signals need not be gender specific in order for sexual selection to be important or for speciation and diversification to be promoted by sexual selection. If correct, this hypothesis would mean that neither mate choice nor species recognition more generally is important in Aculifera, whereas both could play some role in the diversification of Conchifera.
Fertilization and species-recognition traits as they relate to isolation may be only one expression of a more general causal link between diversity and the tendency of some lineages to stray far from the phenotypic legacy of the common ancestry of sister clades. One characteristic of mate choice and of species recognition more generally is that these require activity or motility and the ability to sense signals at a distance. The ability to move and thus in some degree to choose and modify an animal’s habitat could create self-imposed genetic isolation. It may thus be adaptive activity and signaling as they relate to promoting genetic isolation that explain contrasting diversity between sister clades. In the molluscan case, Conchifera (and especially those clades within Conchifera with high diversity) should exhibit more traits indicating motility and signaling at a distance than Aculifera.
Based on an analysis of phenotypes that are unique to one or the other of the two sister clades, we propose that the key lies in the morphological relationship between body and shell. In Aculifera, body and armor remained tightly coupled: even when the shell was lost, locomotor options and sensory elaborations were restricted. By contrast, Conchifera evolved a dynamic integration of body and shell, enabling retraction, extension, and modular remodeling. This shift enabled conchiferans to gain access to phenotypes and modes of life that entail higher levels of activity, including mate choice involving communication at a distance. Aculifera, by contrast, have not transcended the ancestrally tight connection between shell and body, and have been more limited in the range of potential activity levels.
To evaluate these hypotheses, it is instructive to compare derived phenotypes in sister clades that differ in their realized morphological and ecological diversity. These phenotypes should be observable or inferable in living as well as fossil representatives. Restricting comparisons to extant taxa can be highly misleading, because extinct members of clades often occupied phenotypic space that has since been vacated. Mollusca is ideal for such a comparative study because of its rich fossil record, high phenotypic diversity, and great age.
Methods
Using our knowledge of the literature and our direct observations of character states throughout the Mollusca, including extinct clades, we tabulated skeletal and some anatomical features and conditions and assessed their phylogenetic distribution and times of first appearance. Our approach to traits has much in common with the identification and evaluation of character states as employed in the construction of morphology-based phylogenies. We emphasized functionally interpretable phenotypes in order to shift attention toward individual-level phenomena and away from clade-wide properties such as disparity, insofar as these exist and have biological meaning.
Of particular interest are phenotypes related to morphological decoupling and activity. Morphological decoupling is indicated by the dynamic relationship between the soft parts and the shell and by regional differentiation among parts of the rigid shell. Indications of enhanced activity include an extensible foot, retraction and extension of the body relative to the shell, deviations of the mantle edge to form spines and marginal corrugations and crenulations, gluing objects to the shell exterior, formation of the bivalve byssus, inching locomotion, envelopment of the shell by the mantle or foot, epipodial extensions, autotomy, envenomation, photosymbiosis, and remote sensation. These traits are connected to physiological specializations and therefore also indicate higher per capita metabolic rates. Also included within the indicators of active lifestyles are strong ontogenetic departures from logarithmic spiral growth of the shell as a whole and of the growing apertural margin (Vermeij Reference Vermeij2002). Motility and remote sensation enable mate choice at a distance. Motility, remote sensation, metabolic power, and morphological decoupling together influence potential degrees of freedom for phenotypic diversification.
Results and Discussion
Clade-Specific Traits
We identified 19 derived morphological states that have evolved in Conchifera but not in Aculifera, and only 6 traits that arose in Aculifera but are absent in Conchifera (Table 1). This list of adaptive phenotypes is illustrative of innovations within each of the two major molluscan clades, but it is not comprehensive and, like other measures of disparity, does not completely describe the phenotypic diversity that is realized in each clade. Our choice of traits directly links disparity within and between clades to adaptive potential and constraint and thus avoids more abstract approaches to measuring disparity based on multivariate analyses of shape.
Phenotypes unique to either Aculifera or Conchifera.
Table 1. Long description
The table is divided into two sections. The first section lists Aculifera phenotypes: Aesthetes (Cambrian, Vinther Reference Vinther2009; Li et al. Reference Li, Connors, Kolle, England, Speiser, Xiao, Aizenberg and Ortiz2015, Neoloricata), Eyespots in shell (Permian?, Varney et al. 2024, Neoloricata), Shell eyes (Cretaceous, Varney et al. Reference Varney, Speiser, Cannon, Aguilar, Eernisse and Oakley2024, Neoloricata), Body and shell not independent (Cambrian, Peel Reference Peel2020, Aculifera), Conglobation (Triassic?, Connors et al. 2012; Sigwart et al. 2019, Neoloricata), Inching locomotion (Silurian, Sutton et al. 2025, stem Aculifera). The second section lists Conchifera phenotypes: Shell gland (Cambrian, Ponder and Lindberg Reference Ponder and Lindberg2020, Conchifera), Body mobile relative to shell (Cambrian, Ponder and Lindberg Reference Ponder and Lindberg2020, Conchifera), Bilateral asymmetry of body/shell (Cambrian, Ponder and Lindberg Reference Ponder and Lindberg2020, Derived Gastropoda, Cephalopoda, Bivalvia), Epipodial extensions (Cambrian, Ponder and Lindberg Reference Ponder and Lindberg2020, Gastropoda, Bivalvia), Shell enveloped by mantle/foot (Ordovician, Vermeij Reference Vermeij and Briggs2005, Gastropoda, Cephalopoda, Bivalvia), Internal shell resorption (Carboniferous, Vermeij 2020d, Gastropoda, Scaphopoda, Bivalvia), External shell spines (Silurian, Vermeij and Thomson 2026, Gastropoda, Cephalopoda, Bivalvia), Plicated shell wall (Silurian, Vermeij and Thomson 2026, Gastropoda, Cephalopoda, Bivalvia), Mature modifications of shell (Ordovician, Vermeij and Signor 1992; Vermeij and Thomson 2026, Gastropoda, Cephalopoda, Bivalvia), Attachment by byssus or cementation (Ordovician, Vermeij and Thomson 2026, Bivalvia, juliid Gastropoda), Deliberate attachment of objects to shell (Ordovician, Vermeij and Thomson 2026, Gastropoda, Bivalvia), Respiratory openings in shell (Cambrian, Vermeij and Thomson 2026, Gastropoda, Bivalvia), Photosymbiosis (Silurian, Vermeij 2013, Gastropoda, Bivalvia), Organs for drilling prey (Ordovician, Zhang et al. 2020, Gastropoda, Cephalopoda), Hydrostatic and foldable foot (Devonian, Gainey and Stasek Reference Gainey and Stasek1984, Gastropoda, Bivalvia), Chalky shell deposits (Jurassic, Vermeij 2014, Bivalvia), Organs for envenomation (Cretaceous?, Vermeij and Thomson 2026, Gastropoda, Cephalopoda), Autotomy of body parts (Post-Paleozoic, Stasek 1967, Gastropoda, Bivalvia), Toxic secretions (Post-Paleozoic?, Ponder and Lindberg Reference Ponder and Lindberg2020, Gastropoda, Cephalopoda). Each row provides the phenotype, geological period of first appearance, literature reference, and the taxonomic group in which it occurs.
Mate Choice and Species Recognition
The hypothesis that mate choice and species recognition promote diversification is weakly supported by the available evidence from mollusks, but is by itself insufficient to account for extreme Conchiferan diversity. In the clade Aculifera, chitons are either broadcast spawners or spermcasters (Sirenko Reference Sirenko2015), with little scope for individual mate choice. Modest sexual dimorphism in adult body size and growth rate is known in species of Chiton and Acanthopleura (Glynn Reference Glynn1970) and in foot pigmentation in Chiton tuberculatus (Crozier Reference Crozier1920), but there is no indication of active mate choice or communication at a distance in these chitons or any other aculiferans.
Although most conchiferans practice broadcast spawning (including all known members of large clades such as patellogastropods, vetigastropods, and bivalves), in which gamete fertilization takes place beyond the control of parents, terrestrial gastropods as well as the most diverse radiations of marine gastropods and coleoid cephalopods choose individual mates, often (but not always) associated with external sexual dimorphism, with fertilization taking place in or close to one of the parents. Visual mate choice is unlikely in pulmonates, in which the eyes do not discriminate objects at a distance (Zieger and Meyer-Rochow Reference Zieger and Meyer-Rochow2008). It is also unlikely in monoplacophorans, scaphopods, and bivalves, which lack cephalic eyes (von Salvini-Plawen Reference Salvini-Plawen2008). Simultaneously, hermaphroditic pulmonates and opisthobranchs evolved elaborate mating rituals during obligate outcrossing, but available evidence indicates that tactile cues rather than long-distance signals are involved (Schilthuizen Reference Schilthuizen2003, Reference Schilthuizen2005; Anthes and Michiels Reference Anthes and Michiels2007; Anthes et al. Reference Anthes, Schulenburg and Michiels2008). Sexual dimorphism in most bivalves is linked to brooding of juveniles inside the shell of adult females and is reflected in differences in shell form (Heaslip Reference Heaslip and Westermann1969). Species recognition of freshwater unionids by fishes on which the bivalves’ glochidial larvae parasitize is linked to sexual dimorphism in some species but is rare in the group and therefore not a contributor to high diversity in unionids (Barnhart et al. Reference Barnhart, Haag and Roston2008). Some speculations link species-specific sculpture to species recognition and perhaps diversity in marine as well as terrestrial gastropods (Ewers Reference Ewers1967; Schilthuizen Reference Schilthuizen2003), but Hendricks (Reference Hendricks2025) found no support for the role of highly varied color patterns in the hyperdiverse marine family Conidae in species recognition. Thus, although mate choice at a distance may be more frequent in Conchifera than in Aculifera, their role in promoting species diversity in either group is minor at best. It remains possible that mate choice by direct contact could strongly promote diversification in many gastropod clades, in which species often differ in details of genital anatomy.
Activity and Anatomical Decoupling
Although all mollusks have muscles and exhibit some degree of activity, even in sedentary taxa, the Aculifera and Conchifera differ strikingly in the relationship between the rigid exoskeleton (shell) and the soft parts. In Aculifera, these two parts are intimately and inflexibly coupled, limiting locomotor ability and other forms of activity that could create or enhance self-imposed habitat choice and genetic isolation. Conchifera, by contrast, generally show greater decoupling of the shell and soft parts, and are thus much less constrained in the adaptations that are potentially accessible to them.
The most salient phenotype in shell-bearing aculiferans is their serially arranged musculature connected to the serialized valves of the shell. This musculature imparts anteroposterior flexibility in polyplacophorans and is retained in aplacophorans (Caudofoveata and Solenogastres) lacking a shell. Longitudinal flexure enabled inching locomotion (unique for mollusks) as inferred in the Silurian genus Punk (Sutton et al. Reference Sutton, Sigwart, Briggs, Gueriau, King, Siveter and Siveter2025) as well as movement over irregular rock surfaces and the pliant blades of seagrasses and large algae (Clark Reference Clark2002; Sigwart et al. Reference Sigwart, Vermeij and Hoyer2019). Some chitons excavate and occupy deep cavities (Chelazzi et al. Reference Chelazzi, Focardi, Deneuborg and Innocenti1983; Barbosa et al. Reference Barbosa, Byrne and Kelaher2008), a habit for which skeletal flexibility is also well suited. Gastropod limpets occupy all these habitats as well, but in addition they, unlike chitons, often specialize for life on the convex surfaces of other shells, even though limpets have an inflexible, one-piece shell (Vermeij Reference Vermeij2017, Reference Vermeij2020b). Flexibility of the greatly enlarged anterior part of the girdle surrounding the valves contributes to the ability of at least three independently evolved genera of chitons to entrap small mobile prey (McLean Reference McLean1962).
Another phenotypic result unique within Mollusca is conglobation, the ability of many chitons to curl the body and shell into a ball so that the anterior and posterior ends meet, although without locking (coaptive) devices as seen in fossil trilobites (Vermeij Reference Vermeij1987; Sigwart et al. Reference Sigwart, Vermeij and Hoyer2019). As is true for terrestrial arthropods and mammals and for trilobites that curl into a ball, conglobating chitons tend to be slow animals that rely on passive defenses (Vermeij Reference Vermeij1987). Chitons retain sensory contact with their surroundings, including in the enrolled configuration, by virtue of three related sense organs (aesthetes, eyespots, and shell eyes) distributed in the valves of the shell.
Finally, although fleeing on foot from slow predators such as sea stars occurs in some species of chitons (Vermeij Reference Vermeij2020c), motility and other forms of activity in Aculifera are constrained by the tight link between body and shell as well as by the meiofaunal or deep-sea habits of species without shells. No part of the body can substantially extend beyond the confines of the shell.
The ancestral condition in Conchifera is the presence of a one-piece symmetrical shell that is tightly linked to the soft parts, initiated by the shell gland, and growing at the apertural end (Ponder and Lindberg Reference Ponder and Lindberg2020). This condition characterized many Helcionelloidea and Hyolitha and the still-extant Monoplacophora. An early conchiferan innovation was the ability of the body to retract and extend into and out of the shell despite the attachment of the body by muscles to the shell’s inner surface (Ponder and Lindberg Reference Ponder and Lindberg2020; Dzik Reference Dzik2025). This motility of the body relative to the shell offered new degrees of freedom and activity. Unlike in Aculifera, soft parts including the mantle, foot, head, and their sensory and mechanical extensions in Conchifera provide a dynamic integration between shell and body, allowing for far-reaching modifications and elaborations of the shell and body. These include internalization of the shell by the foot or mantle, the formation of shell spines by highly extensible parts of the mantle (Chirat et al. Reference Chirat, Moulton and Goriely2013), other external and internal shell sculpture (ribs, ridges, keels, and shell-wall plicae) (Vermeij Reference Vermeij2002), asymmetry and other deviations and spatial differentiation and modularity of shell form from the ancestral bilaterally symmetrical condition (Vermeij Reference Vermeij2002), attachment of foreign objects to the shell exterior by the mantle or foot (Vermeij Reference Vermeij2014; Vermeij and Thomson Reference Vermeij and Thomson2026), and shell modifications to accommodate eyes, siphons, respiratory openings, and byssal threads in bivalves. This is complemented by modifications to the soft body, such as the cephalopod funnel; the gastropod proboscis; and in Megalopodifera, a hydrostatic foot that can dilate and fold for retraction, swimming, rapid crawling, or burrowing (Trueman Reference Trueman1968; Miller Reference Miller1974; Gainey and Stasek Reference Gainey and Stasek1984). Adaptive modifications of the shell were aided by, and in turn aided, the application of forces associated with motility (Vermeij Reference Vermeij2002).
The muscular decoupling of the shell and body likely originated within Helcionellida and was elaborated independently in Megalopodifera, which are characterized by a large foot that can extend outside the shell (Chen et al. Reference Chen, Baeza, Chen, Gonzalez, Gonzalez, Greve, Kocot and Marbizu2025), and shell-bearing Cephalopoda. In both of these major conchiferan clades, decoupling enabled phenotypes that indicate or enhance activity involving motility of the soft parts relative to the shell (Table 1). The body of shell-bearing cephalopods (as seen in Nautilus and inferred in other fossil forms) extends beyond the shell aperture but is not dramatically extensible as in Megalopodifera; this constraint is solved independently in cephalopods by a pathway of shell internalization and loss.
The reduction or loss of the heavy external shell in many gastropod lineages and coleoid cephalopods further eliminated many mechanical constraints and promoted replacement of passive armor with toxicity, greater maneuverability and speed, and even aggression. This apparently enabled the radiation of modern, fast-moving coleoid cephalopods, but also the shell did not constrain the radiation of the ammonites, which comprise the majority of total cephalopod diversification. Shell loss also occurred in Aculifera, but the body types that distinguish Caudofoveata and Solenogastres remain limited variations on the worm theme. Most Aculifera without shells live in the deep sea or among sediment grains in the meiofaunal environment, habitats where the evolution of power-intensive activity is comparatively constrained by low productivity and slow competition and predation. Aplacophorans lack photoreceptors (von Salvini-Plawen Reference Salvini-Plawen2008), and members of Caudofoveata and some Solenogastres lack a locomotor foot (Kocot et al. Reference Kocot, Todt, Mikkelsen and Halanych2019; Bergmeier et al. Reference Bergmeier, Ostermair and Jorger2021).
Another indication of greater phenotypic potential in some conchiferan lineages is the regional anteroposterior differentiation of the rigid shell into sectors that differ in size, shape, and sculpture. Such modularity has evolved in early bivalves, caenogastropod and heterobranch gastropods, and perhaps in some shell-bearing cephalopods, but not in scaphopods, nonsiphonate gastropods (Patellogastropoda, Vetigastropoda, and Neritimorpha), or nautilidan cephalopods (Vermeij Reference Vermeij2002; Johnson et al. Reference Johnson, Fogel and Lambert2019). In the latter groups, the apertural margin traces a circular or oval contour that conforms more or less to a logarithmic spiral configuration without regional differentiation.
Ancestral conchiferans lacked most of the traits that would later enable many lineages to depart phenotypically from the initial morphology. The potential for greater phenotypic diversity in conchiferans did emerge very early in the history of the clade: 5 conchiferan-only phenotypes, including the motility of the body relative to the shell, were established during the Cambrian; another 6 followed in the Ordovician, 4 in the Siluro-Devonian, and one in the Carboniferous (Table 1). Only 4 additional phenotypes probably appeared after the Paleozoic: chalky deposits in fast-growing oyster-like bivalves (Jurassic), envenomation in some neogastropod and octopod coleoid lineages (likely during the Late Cretaceous), and toxicity and autotomy with unknown times of origin but again, judging from phylogenetic relationships, post-Paleozoic. It is notable that all conchiferan traits except the shell gland evolved independently in many lineages, especially during the later Mesozoic and Cenozoic (Vermeij and Thomson Reference Vermeij and Thomson2026). Earliest instances of these phenotypes long preceded the major taxonomic expansions of bivalves and of internally fertilizing gastropods and cephalopods beginning in the Mesozoic. Initial colonization of environments not previously occupied by conchiferans also took place during the Paleozoic, hundreds of millions of years before diversity in these environments began to rise. Adult conchiferans colonized the pelagic realm (Cambrian), infaunal and chemosynthetic environments (Ordovician), endolithic habitats (Silurian), fresh water (Devonian), and dry land (Carboniferous) (Ponder and Lindberg Reference Ponder and Lindberg2020; Vermeij and Thomson Reference Vermeij and Thomson2026). In short, phenotypic novelty and potential were temporally disconnected from their full realization.
General Discussion
The two hypotheses considered here as potential explanations for the difference in diversity between Aculifera and Conchifera (especially its subgroup Megalopodifera) are not mutually exclusive, and indeed likely reinforce each other. Mate choice and species recognition are enabled by motility and sensory capacities (Vermeij and Grosberg Reference Vermeij and Grosberg2018). Importantly, these are local interaction and activities that not only promote adaptation and genetic isolation, but also enhance intense selection for greater activity, especially if they are accompanied by power-intensive display (Vermeij Reference Vermeij2023a). Moreover, activity enables individuals to exercise greater choice of where to live, what to do, and what to eat. Phenotypes such as body-shell decoupling that expand the range of activity and choice, including mate choice, should therefore indirectly establish evolutionary feedbacks that result in greater self-imposed genetic isolation and thus potentially speciation.
Not all conchiferan clades are ecologically or morphologically diverse. Scaphopods and monoplacophorans lack photoreceptors (von Salvini-Plawen Reference Salvini-Plawen2008) and, like protobranch bivalves, are morphologically conservative and show no indication of individual mate choice or sexual dimorphism. Nautilidan cephalopods have been conservative since their Devonian origin despite being relatively active animals that copulate and that exhibit some sexual dimorphism (Ward Reference Ward1987). Numerous conchiferan lineages evolved in or became restricted to environments where the pace of life (metabolic rate and activity) is slow. Furthermore, as the convergent evolution of the limpet form in nearly 60 gastropod lineages shows (Vermeij Reference Vermeij2017), there have been many reversals from a dynamic relationship between body and shell to a more tightly constrained one. Even in limpets and other conchiferans in which body and shell have become more tightly coupled, however, phenotypes have arisen that are unknown in aculiferans, including external asymmetry, a plicated shell wall, spines, and the capacity to extend the mantle over the shell edge (Vermeij Reference Vermeij2017). We emphasize that the question is not why some conchiferan lineages have remained at low extant diversity, but why many clades within Conchifera (and especially in Megalopodifera) have become hyperdiverse.
This heterogeneity in degree of coupling between shell and body within derived Conchifera underscores the important point that clades are unlikely to possess universal trends. Instead, clades are evolutionary branches whose only unifying characteristic is descent of members from a common ancestor, with or without ancestral traits being retained, as inferred retrospectively with phylogenetic methods (Vermeij Reference Vermeij2020a). Within-clade heterogeneity would be expected to increase with more speciation events, because with every evolutionary split within a clade there is the likelihood of some phenotypic change. This expectation also implies a certain level of circularity, or at least feedback, in the relation between innovation and taxonomic diversity.
It is important to emphasize that mate choice, species recognition, and decoupling of the body and shell do not inevitably lead to diversification, but instead provide the potential to do so. Realization of this potential depends on a host of other factors related to opportunities for isolation. For example, the greater activity and metabolic power associated with body–shell decoupling can be fully realized only in environments where higher activity levels are sustained by ecosystems in warm productive settings where selection by predators, mates, and other biological agents is intense. Causes and effects among these factors are intertwined because of the pervasive action of feedbacks among natural selection, individual and collective agency by organisms, morphological decoupling, and the productive capacity of ecosystems in which interactions take place. Isolation is also affected by whether unsuitable habitats can be crossed by dispersal stages and thus by the capacity of populations separated by potential barriers to remain interconnected or to become mutually isolated.
How might our hypothesis be falsified? If the mechanisms we propose are correct, they should apply to other highly unequal pairs of sister groups with appropriate accommodations for their particular traits. If, for example, the much less diverse clade of the pair could be shown to engage in more mate choice at a distance or if its morphology indicates greater decoupling among parts than the more diverse sister clade, the generality of the hypothesis could rightly be questioned. In science, however, hypotheses are much more likely to be modified as more data become available than they are to be discarded. The requirement that a hypothesis be formulated so that it can be falsified in principle strikes us as imposing a binary choice between acceptance and rejection. Such binary options are more applicable to facts and to mathematics than to empirical interpretations, where support for or modification of a hypothesis typically depends on either additional data or a reinterpretation of the evidence. For biological systems, phenomena and their expression depend on more than single factors, and these are rarely independent of each other; their effects entail feedbacks. In short, the framework of falsification seems inappropriately constraining in biology and probably in empirical science generally. Likewise, strictly alternative hypotheses are uncommon; much more likely is that potential explanations for a phenomenon complement each other, as we believe they do in the present study.
Conclusions
Our analysis emphasizes that the exceptional diversity of Conchifera cannot be explained as a simple case of sister-clade inequality. Fossils show that both major molluscan clades experimented with body plans, yet only within Conchifera did mollusks establish novel architectural systems that could be elaborated repeatedly and convergently. The body–shell relationship of Megalopodifera and Cephalopoda was reorganized into a modular framework permitting innovations in locomotion, feeding, defense, and reproduction. These traits not only enabled ecological expansion but also fostered feedbacks between activity, mate choice, environmental permissiveness, and genetic isolation that promoted speciation in many lineages.
In a broader context, Conchifera illustrate one of the very few instances—alongside insects and the broader pancrustacean clade as well as tetrapod vertebrates—in which a metazoan lineage attained extraordinary ecological and morphological scope. This success was not the inevitable result of novelty alone but of innovations that could be generalized, recombined, and repeatedly applied across environments. The molluscan clade underscores the importance of functional integration rather than isolated key innovations in explaining why animal lineages come to dominate the history of life. In mollusks, as in other plant and animal clades, mate choice and sexual selection do not invariably promote speciation (Kraaijeveld et al. Reference Kraaijeveld, Kraaijeveld-Smit and Maan2011); instead, it is the combination of modularity, activity, and greater feedback between organism and environment that opens the door to diversification.
Diversity is not simply about innovation, speciation, or phylogeny, but instead must be understood as the intersection of locally generated adaptation, susceptibility to genetic isolation, and extrinsic conditions such as temperature and productivity. The historical trajectories of clades reflect general evolutionary feedbacks between organism and environment as modulated by evolutionary legacy and the potential to depart from that legacy.
Acknowledgments
We thank T. J. Thomson for valuable technical assistance.
Author Contribution
G.J.V. wrote most of the paper. J.D.S. wrote part of it and contributed important insights.
Competing Interests
The authors have no conflicts of interest.
Open access
