MYRTEINAE, Chavan, 1969

Taylor, John D., Glover, Emily A., Smith, Lisa, Dyal, Patricia & Williams, Suzanne T., 2011, Molecular phylogeny and classification of the chemosymbiotic bivalve family Lucinidae (Mollusca: Bivalvia), Zoological Journal of the Linnean Society 163 (1), pp. 15-49 : 33

MYRTEINAE CHAVAN, 1969

Myrteinae comprise a diverse, poorly studied group of lucinids, usually associated with deeper water, offshore habitats. Chavan (1969) included 11 genera within his subfamily Myrteinae : five of these ( Myrtea , Gonimyrtea , Monitilora , Lucinoma , and Prophetilora ) with living species and six others including only fossil species. Our molecular results indicate that Monitilora and Lucinoma group in separate subfamilies, Monitilorinae and Codakiinae, respectively, and are discussed in detail below. The position of Gonimyrtea and Prophetilora remains uncertain, pending further molecular analysis. We analysed six species classified in the genera Myrtea , Notomyrtea , and Gloverina , the latter a recently described genus ( Cosel & Bouchet, 2008) included here in the Myrteinae for the first time. Our results ( Figs 1–5 ) show that Myrteinae form a highly supported clade sister to all other lucinids, except the Pegophyseminae and Leucosphaerinae .

In our new concept, Myrteinae share some common morphological features, such as thin, elongate shells and regularly spaced commarginal lamellae that are often elevated above the dorsal shell margin. Some species have radial folds between the lamellae, but never strong radial ribbing. Additionally, they have short, anterior adductor muscle scars, narrow hinges with small to obscure cardinal teeth, and in some species a clearly asymmetric lunule.

Other newly described genera from the deep sea, including Elliptolucina, Taylorina , Rostrilucina , Graecina ( Cosel, 2006; Cosel & Bouchet, 2008), and Jorgenia ( Taylor & Glover, 2009a) , may also belong in Myrteinae , but no suitable material for molecular analysis was available. Similarly, some shallow water genera, such as Myrtina and Solelucina ( Glover & Taylor, 2007) , have shell characters consistent with inclusion in Myrteinae .

Myrteinae are poorly documented in the fossil record, but some Myrtea -like bivalves are present throughout the Cenozoic, and Lucina blanckenhorni Chavan, 1947 from the late Cretaceous (Campanian 83.5-70.6 Ma) may be part of a ‘ Myrtea ’ stem group. Somewhat earlier, Paramyrtea sabulosa Kendrick & Vartak, 2007 , from the mid-Cretaceous (Cenomanian) of India, has shell characters compatible with inclusion in the Myrteinae . The chronogram ( Fig. 5 ) indicates a late Cretaceous radiation for the subfamily, with a possible separation from other lucinids earlier in the Cretaceous.

Figures

Figure 1. Molecular phylogeny of the Lucinidae based on the single gene analysis of 18S rRNA produced by Bayesian analysis using MRBAYES. Support values are Bayesian posterior probabilities (%). Details of the taxa are given in Table 1.

Figure 2. Molecular phylogeny of the Lucinidae based on the single gene analysis of 28S rRNA produced by Bayesian analysis using MRBAYES. Support values are Bayesian posterior probabilities (%). Details of taxa are given in Table 1.

Figure 3. Molecular phylogeny of the Lucinidae based on the single gene analysis of cytochrome b produced by Bayesian analysis using MRBAYES. Support values are Bayesian posterior probabilities (%). Details of taxa are given in Table 1.

Figure 4. Molecular phyogeny of Lucinidae produced by Bayesian analysis of the combined data set from concatenated sequences of the nuclear 18S rRNA and 28S rRNA, and mitochondrial cytochrome b genes, using Thyasiridae species as out-groups. Phacoides pectinatus is excluded from this tree because of its unstable position within the large clade. The inset to the right shows details of the larger clades. Support values are Bayesian posterior probabilities (%). Details of

Figure 5. Chronogram generated by BEAST for species of Lucinidae, and calibrated using ten fossil records. The position of each calibration point is indicated by the numbers in circles (see the Material and methods and Discussion sections for details about the fossils). Error bars for node ages are the 95% highest posterior density intervals; clades with less than 50% support have no error bars; the root node has node bar trimmed by 50% for clarity. New subfamily divisions are shown on the right-hand side, with details presented in Table 3.