Phylogenetic studies of the marine bivalve subfamily Venerinae (: ) Isabella Kappner, Rüdiger Bieler Department of Zoology (Invertebrates), Field Museum of Natural History, Chicago, IL, [email protected]

1 23 Chamelea gallina Venus verrucosa 54 Venus casina 96 Goal of present studies 52 Clausinella punctigera Results & Discussion 54 96 Venus rosalina 54 How did morphological diversity evolve over Globivenus toreuma 71 ƒ The phylogeny of the combined data (16S, COI and 80 Globivenus rugatina time in the marine bivalve subfamily 72 76 Ventricoloidea foveolata Venerinae s.novo H3) strongly suggests that Chioninae s.s. and A1 Globivenus effosa Venerinae? Globivenus isocardia Venerinae s.novo are discrete clades. Dosina zelandica A Ameghinomya antiqua • Investigate evolutionary relationships using 100 Eurhomalea lenticularis ƒ Several genera (Chamelea, Clausinella, , 100 Tawera spissa mitochondrial and nuclear gene sequences Timoclea ovata Timoclea) were probably mis-classified in the past A2 Timoclea subnodulosa • Provide a robust phylogenetic framework for Timoclea levukensis due to homoplasy in morphological features Antigona lamellaris understanding venerine morphological character 100 Lirophora mariae 100 Lirophora paphia ƒ Monophyly of Venerinae s.l. or Venerinae s.s. evolution Anomalocardia brasiliana 99 Mercenaria campechiensis (Keen, 1969) was not present in the resulting 50% 99 Mercenaria mercenaria majority rule consensus tree. • Test traditional systematic hypotheses Humilaria kennerleyi Puberella intapurpurea Chioninae s.s. Ameghinomya sp2 ƒ The two tested traditional systematic hypotheses B85 99 Chione cancellata 85 100 Chione subimbricata are rejected with p<0.001 in the constraint analysis Taxonomic controversy Chionista fluctifraga Protothaca mcgyntyi 67 and outside the 99.9% confidence set in ELW test. Callithaca tenerrima 69 Venerine shells are morphologically very similar to Protothaca staminea 100 Ruditapes decussata GB ƒ A combination of morphological characters can be those of the subfamily Chioninae and can only be 59 100 Ruditapes decussata distinguished from each other by the presence or 99 56 100 Paphia euglypta used to distinguish the two subfamilies (Fig. 3): 99 100 Paphia vernicosa absence of an anterior lateral tooth. Taxonomists Paphia dura 74 100 Katelysia sp 1 Tapetinae ¾ Venerinae have separated siphons and most 73 98 100 99 Katelysia sp 2 have been arguing for over 150 years about the 99 62 94 Katelysia rhytiphora taxa in this group have an anterior lateral tooth 96 100 98 Katelysia scalarina significance of this minute anterior lateral tooth and 62 96 Ruditapes bruguieri ¾ Chioninae have fused siphons and lack an the of this group remains controversial. Ruditapes philippinarum G B 96 Pectunculus exoleta anterior lateral tooth. 96 96 Dosinia sp S hark Bay Dosiniinae Fischer-Piette (1975) carried out the last major 96 85 Dosinia victoriae ƒ Within the Venerinae no monophyletic generic 87 Dosinia sp E sperance revision of the Venerinae. He separated the two 100 Placamen berryi groupings have been identified 100 Placamen flindersi subfamilies and revised the Venerinae as a one- 65 listeri 65 Periglypta puerpera genus subfamily (Venus). In the most recent study Macrocallista squalida Pitar rudis Problems with sequencing COI dealing with this subject, Chioninae and Venerinae Meretrix lyrata Callista chione ƒ 14 of the 56 sampled taxa did not deliver any were synonymized based on morphological features 0.1 Kappner & Bieler, Mol. Phylogen. Evol. in press. results for the partial COI gene Coan and Scott (1997). Venerinae s.l. currently Figure 2. Molecular phylogeny of Veneridae: 50% majority rule consensus tree based on a ƒ Museum material often in unknown fixatives can comprises 41 extant genera (10 former Venerinae Bayesian analysis of the concatenated dataset (16S, COI and H3). Posterior probability and 31 former Chioninae) with over 180 species. values > 0.95 are indicated by bold lines. Support values > 70 from jackknife replicates cause difficulties in amplifying longer sequences are indicated below the branches, and from bootstrap replicates above branches. Morphological trait mapping onto the molecular phylogeny of venerid bivalves: Character 1 ƒ “Barcoding” will likely be successful with fresh Methods (crenulations in internal margin), black boxes = presence, white boxes = absence; material, but could be very labor and time intensive Character 2 (anterior lateral tooth of Type I), black boxes = presence, white boxes = with older museum material. We performed a phylogenetic analysis of our 3-gene absence; Character 3 (siphons), black boxes = separated, white boxes = fused. dataset based on partial sequences of the mitochondrial 16S gene (602 bp), COI gene (569 bp) Morphological Trait Mapping Literature cited and nuclear Histone 3 gene (328 bp) including 55 Altogether 75 conchological characters as well as six Coan, E.V., Scott, P.H., 1997. Checklist of the marine bivalves of species of 37 genera of Venerinae s.l. as well as 18 characters of internal anatomy were investigated the northeastern Pacific Ocean. Sta. Barbara Mus. Nat. Hist. Contrib.Science 1, 1-28. outgroup taxa of other venerid subfamilies. (data not presented here). The morphological Fischer-Piette, É., 1975. Révision des Venerinae s.s. (Mollusques Alignments were analyzed by a mixed model characters were mapped onto the molecular topology Lamellibranches). Mem. Mus. Natn. Hist. Nat. A Zool., Paris 93, Bayesian approach with Markov Chain Monte Carlo of the concatenated analysis and three informative 1- 64. (B/MCMC) methods using MrBayes 3.1.2 and characters were found: Kappner, I., Bieler, R. in press. Phylogeny of Venus clams maximum parsimony methods using PAUP*. MrBayes (1) Presence or absence of crenulations in interior (Bivalvia: Venerinae) as inferred from nuclear and mitochondrial gene sequences. Mol. Phylogen. Evol. was set to produce 3,000,000 generations and to run shell margin (Fig. 3B) four replicas of four chains simultaneously. Trees Keen, A.M., 1969. Superfamily Veneracea. In: Cox, L.R. et al., (2) Presence (Fig. 3C1) or absence (Fig. 3C2) of were sampled every 100 generations for a total of Part N [Bivalvia], 6, volume 2: ii + pp. N491- N952. anterior lateral tooth in left valve In: Moore, R.C. (Ed.) Treatise on Invertebrate Paleontology. 30,000 trees. The software Tracer 1.2 was used to Lawrence, Kansas: Geological Society of America and determine the “burn in”. Maximum parsimony (3) Degree of siphon fusion; partially to completely University of Kansas, N670-N690. analysis was carried out with 1000 random sequence separated (Fig. 3D1) vs. completely fused (Fig. 3D2) Mikkelsen, P.M., Bieler, R., Kappner, I., Rawlings, T.A., in press. additions, and TBR branch swapping. Branch support Phylogeny of Veneroidea (Mollusca: Bivalvia) based on was examined by undertaking 300 jackknife and morphology and molecules. Zool. J. Linn. Soc. 2000 bootstrap replicates with a minimum of 10 random sequence additions per replicate. Acknowledgements Taxon sampling Thanks to the curators and staff of following museums for loan From the traditional Venerinae we sampled seven of of specimens and tissue: MNHN, Paris; IRSNB, Brussels; NMNZ, ten genera and 15 of 31 chionine genera. Outgroup Wellington, New Zealand; FLMNH, Gainesville; LACM, Los Angeles; BMNH, London. taxa were chosen from a concurrent study of our B working group (Mikkelsen et al. (in press)). For assistance during field collections we thank: Katja Defren- A Janson (Berlin, Germany), Sonia Merino (Instituto Nacional de Desenvolvimento das Pescas - Mindelo, Cabo Verde), Elisio Delgado and crew (Mindelo, Cape Verde Islands), Eurico Barros Hypothesis testing and crew (Sal, Cabo Verde), Patrice Petit De Voize (Fédération Française d'Etudes et de Sports Sous-Marins - Commission Based on our results two traditional systematic nationale de Biologie, France), Haidar el Ali (Oceanium, Dakar, hypotheses (Fig. 1) were tested. This was performed C1 Senegal), Lisa Kirkendale (FLMNH), Emily Glover and John by examining suboptimal trees present in the Taylor (BMNH), Fred Wells (Western Australian Museum, Perth), B/MCMC sample and by calculating expected and Rachel Collin (Smithsonian Tropical Research Institute, likelihood weights (ELW) with TreePuzzle 5.2. Panama). We are grateful to the following people for help in obtaining C2 D additional specimens: Claude Berthoult (Centre ORSTOM de Nouméa, New Caledonia), Alan G. Beu (Institute of Geological and Nuclear Sciences, Lower Hutt, New Zealand), Rachel Collin (Smithsonian Tropical Research Institute, Panama), Louise Crowley (American Museum of Natural History, New York), Brian Dyer (Universidad del Mar, Chile), Leonore Kappner (Bochum, Germany), Alan J. Kohn (University of Washington, Seattle), Taeko Kimura (Mie University, Mie, Japan), and Melita Peharda Fig. 1 Different traditional systematic hypotheses (a) after Keen (University of Zagreb, Croatia). (1969) and (b) after Coan and Scott (1997) (author’s interpretation). Additional funds were provided by UIC’s Provost Award for Graduate Research, FMNH Zoology Department’s Marshall Field D1 D2 Fund, and the Conchologists of America.

Figure 3. Morphological features that were found to be useful for classification and that were utilized for mapping on the molecular tree. A. Overview of internal shell, B. Crenulations of the interior shell margin, C. Presence of an anterior lateral tooth of Type I (C1) and absence of anterior lateral tooth (C2) in the left valve. D. Overview of internal morphology, D1. Separated siphons, D2. Fused siphons. A, B, C1, and D1. Venus verrucosa Linnaeus, 1758, C2, and D2. Lirophora paphia (Linnaeus, Sponsored by NSF-PEET DEB-9978119 1767).