In scallops, the shell valves are circular, but they may be concave and similar or the left (uppermost) one may be flat. Like oysters, they also lie in a horizontal position on the substrate. However, scallops, far from being fixed, are active swimming bivalves. In early life, they use byssus threads for attachment to algae, but before they attain a size of 15 mm the majority of species have detached themselves to take up a free‐living existence on the seabed.
Cementation is another mode of attachment that evolved during the early Mesozoic era (252 mya). This adaptation arose independently in Pteriomorphia, Heterodonta and Palaeohterodonta, peaking in the Late Triassic and Jurassic periods of the Mesozoic era (252–145 mya) as a possible response to the appearance of many predatory groups (Vermeij 1977; Harper 1991). During the Triassic, another important development occurred when an ancestral unionid (Paleoheterodonta) colonised freshwater environments, thereby gaining access to a bivalve‐free ecosystem. Giribet (2008) suggests that this move may have been triggered by the evolution of a novel mode of development using microscopic glochidia larvae with fish as intermediate hosts.
Burrowing into the substrate is the habit most extensively exploited by bivalves. Contact is maintained with the surface by way of siphons that extend from the posterior end of the animal. During the Cenozoic era (up to 60 mya), soft, nutrient‐rich sediments on continental margins allowed diversification of shallow burrowing, globular, strongly ribbed forms and deep burrowers with smooth, blade‐like shells (Giribet 2008). Cockles (e.g. Cardium spp.) are shallow burrowers, while many clam species (e.g. razor clams) burrow as deep as 60 cm. The geoducks (Panopea) on the West Coast of the United States are among the deepest burrowers, digging down to a depth of over 1 m, aided by a streamlined shell for fast burrowing and fusion of the mantle edges (apart from a small gape for the large muscular foot), which prevents entry of sediment into the mantle cavity. Many bivalves that burrow deeply (>30 cm) live in permanent burrows, moving deeper as they grow larger. This lifestyle is brought to an extreme in bivalves that bore into hard substrates such as shell, coral, wood and rock and are permanently locked in their burrows, which are therefore inevitably dependent on outside sources of food. Both shallow and deep‐sea wood‐boring bivalves excavate ‘sawdust’ as their principal food source with the aid of gill‐associated symbiotic bacteria (Voight 2007).
Phylogeny and Evolution in the Mytilida
Pteriomorphia is monophyletic and contains four well‐supported marine clades: Arcida, Mytilida, Ostreida and Pectinida + Limida (Figure 1.2). The order Mytilida originated in the Cambrian era some 540–560 million years ago and is composed of a single superfamily, the Mytiloidea, which is considered to be representative of and basal to the Pteriomorphia (Morton 2015 and references therein). Information based on shell morphology and molecular data suggests that Mytiloidea is a monophyletic clade (Morton et al. 2016 and references therein). For more than five decades, there has been no consensus on the number of families contained within the Mytiloidea. For example, Boss (1982) recognised only a single family, the Mytilidae, divided into four subfamilies, Mytilinae, Crenellinae, Lithophaginae and Modiolinae. Coan et al. (2000), and subsequently Coan & Valentich‐Scott (2012), also recognised a single family, but added three additional subfamilies: Bathymodiolinae, Dacrydiinae and Septiferinae. Bieler et al. (2010) added one more, Limnoperninae, making a total of eight subfamilies within the Mytilidae. In contrast, others have recognised between three and four families in the Mytiloidea; Carter et al. (2011), for example, in a classification including fossil taxa, recognised three families, Mytilidae, Crenellidae and Septiferidae, while Morton (2015) proposed four, the Mytilidae, Crenellidae, Modiolidae and Musculidae – although he suggested that the Septiferinae might be eligible for family status subject to additional morphological and molecular analyses. Bieler & Mikkelsen (2006) initially recognised five families, the Mytilidae, Crenellidae, Musculidae, Modiolidae and Septiferidae, although as we have seen they later recognised only one, the Mytilidae (Bieler et al. 2010). Table 1.1 presents a summary of five different taxonomic classifications of extant Mytiloidea, which are only a fraction of the published classifications for this superfamily. The majority of classifications have been heavily reliant on shell features, such as shape and characteristics of the valves, hinges and dentition, with few authors to date using anatomical features (reviewed in Morton 2015) or molecular data.
The Mytilidae is the largest family within the Mytiloida, with subfamily numbers varying between four and eight (Table 1.1). The family consists of 52 genera and 411 species (www.marinespecies.org). Species are recognised as members of the family by the shell form, sculpture, hinge and muscle scars (see Soot‐Ryen 1955 for details). Mytiliform species, with a broadly triangular shape and terminal umbones, are placed in the genus Mytilus (Figure 1.4A), while those with subterminal umbones with a distinct anterior margin – the modioliform species – are consigned to the genus Modiolus (Figure 1.4B). Elongate forms that burrow in rock or mud and have parallel dorsal and ventral shell margins and subterminal umbones are placed in the genus Lithophaga (Figure 1.4C), species with pronounced anterior and posterior radiating ribs, except for a smooth median area, are placed in the genus Musculus (Figure 1.4D), and species with radiating sculpture over the whole outer shell surface are assigned to the genus Brachidontes (Figure 1.4E). Classification solely based on external shell features is problematic as these can be influenced by environmental conditions and similar forms may arise as a result of convergent adaptations to common environmental conditions. For example, a phylogenic study of four subfamilies of Mytilidae found that Crenellinae and Lithophaginae were monophyletic but Mytilinae and Modiolinae were polyphyletic4 (i.e. they do not share a common ancestor), suggesting that the mytiliform and/or the modioliform body plans have evolved independently in at least two mytilid lineages (Distel 2000). However, in a recent study using species in five subfamilies (Mytilinae, Modiolinae, Bathymodiolinae, Lithophaginae and Crenellinae), Liu et al. (2018) found that phylogenetic trees reconstructed with a combination of two mitochondrial (COI and 16S rRNA) and three nuclear (18S and 28S rRNA and histone H3) genes indicated monophyly of the family Mytilidae and the subfamily Bathymodiolinae, polyphyly of the subfamilies Modiolinae and Lithophaginae and paraphyly5 of the subfamily Mytilinae.
Table 1.1 Classification schemes in the superfamily Mytiloidea. Families within the Mytiloidea are in bold, while those in plain text are subfamilies recognized within the individual families. Source: From Morton (2015). Reprinted with permission from Taylor & Francis.
| Boss 1982 a | Coan et al. 2000 a | Bieler et al. 2010 a,b,c | Carter et al. 2011 a,b,c | †Morton 2015 a |
|---|---|---|---|---|
| Mytilidae | Mytilidae | Mytilidae | Mytilidae | Mytilidae |
| Mytilinae |
Mytilinae
|
