A peer-reviewed open-access journa I ZooKeys 717: |—139 (2017) doi: 10.3897/zookeys.717.21885 RESEARCH ARTICLE #ZooKey http:/ / ZOO keys -pen soft. net Launched to accelerate biodiversity research Polyphyly of the traditional family Flabellinidae affects a major group of Nudibranchia: aeolidacean taxonomic reassessment with descriptions of several new families, genera, and species (Mollusca, Gastropoda) Tatiana Korshunova'’, Alexander Martynov’, Torkild Bakken?, Jussi Evertsen?, Karin Fletcher*, | Wayan Mudianta®, Hiroshi Saito®, Kennet Lundin’®, Michael Schrédl*!°, Bernard Picton!!!? | Koltzov Institute of Developmental Biology, RAS, 26 Vavilova Str, 119334 Moscow, Russia 2 Zoological Museum, Moscow State University, Bolshaya Nikitskaya Str. 6, 125009 Moscow, Russia 3 NTNU University Museum, Norwegian University of Science and Technology, NO-7491 Trondheim, Norway 4 Port Orchard, Washington 98366, USA 5 Universitas Pendidikan Ganesha, Bali, 81116, Indonesia 6 National Museum of Nature and Science, Amakubo 4-1-1, Tsukuba, Japan 1 Gothenburg Natural History Museum, Box 7283, S-40235, Gothenburg, Sweden 8 Gothenburg Global Biodiversity Centre, Box 461, S-40530, Gothenburg, Sweden 9 Zoologische Staatssammlung Miinchen, Miinchhausenstr. 21, D-81247 Miinchen, Germany 10 Biozentrum Ludwig Maximilians University and GeoBio-Center LMU Munich, Germany \\ National Museums Northern Ireland, Holywood, Northern Ireland, United Kingdom \2 Queens University, Belfast, Northern Ireland, United Kingdom Corresponding author: Alexander Martynov (martynov@zmmu.msu.ru) Academic editor: NV. Yonow | Received 27 October 2017 | Accepted 8 November 2017 | Published 30 November 2017 http://zoobank.ore/C19B43B1-B321-4CB1-B1B2-A246CEAC56BC Citation: Korshunova T, Martynov A, Bakken T, Evertsen J, Fletcher K, Mudianta WI, Saito H, Lundin K, Schrédl M, Picton B (2017) Polyphyly of the traditional family Flabellinidae affects a major group of Nudibranchia: aeolidacean taxonomic reassessment with descriptions of several new families, genera, and species (Mollusca, Gastropoda). ZooKeys 717: 1-139. https://doi.org/10.3897/zookeys.717.21885 Abstract The Flabellinidae, a heterogeneous assembly of supposedly plesiomorphic to very derived sea slug groups, have not yet been addressed by integrative studies. Here novel material of rarely seen Arctic taxa as well as North Atlantic, North and South Pacific, and tropical Indo-West Pacific flabellinid species is investigated morpho-anatomically and with multi-locus markers (partial COI, 16S rDNA, 28S rDNA and H3) which were generated and analysed in a comprehensive aeolid taxon sampling. It was found that the current fam- Copyright Tatiana Korshunova et al. This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. 2 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) ily Flabellinidae is polyphyletic and its phylogeny and taxonomic patterns cannot be understood without considering members from all the Aeolidacean families and, based on a robust phylogenetic hypothesis, morpho-anatomical evolution of aeolids is more complex than suspected in earlier works and requires reclassification of the taxon. Morphological diversity of Flabellinidae is corroborated by molecular di- vergence rates and supports establishing three new families (Apataidae fam. n., Flabellinopsidae fam. n.., Samlidae fam. n.), 16 new genera, 13 new species, and two new subspecies among the former Flabellini- dae. Two families, namely Coryphellidae and Paracoryphellidae, are restored and traditional Flabellinidae is considerably restricted. The distinctness of the recently described family Unidentiidae is confirmed by both morphological and molecular data. Several species complexes among all ex-“Flabellinidae” lineages are recognised using both morphological and molecular data. The present study shows that Facelinidae and Aeolidiidae, together with traditional “Tergipedidae”, deeply divide traditional “Flabellinidae.” Diag- noses for all aeolidacean families are therefore provided and additionally two new non-flabellinid families (Abronicidae fam. n. and Murmaniidae fam. n.) within traditional tergipedids are established to accom- modate molecular and morphological disparity. To address relationships and disparity, we propose a new family system for aeolids. Here the aeolidacean species are classified into at least 102 genera and 24 fami- lies. Operational rules for integration of morphological and molecular data for taxonomy are suggested. Keywords Integration of morphological and molecular data, molecular systematics, Mollusca, morphology, phylo- genetics, taxonomic revision Table of contents IDEM earGhe Vou (a) 6 ROUr vv) aR Rt i CWP 2 nea eae Ii 0 A RUPEE 5, SOO ae CPE + Ivitettals-arObrtec nese. ..2738see ae ied hui wei eet Doras, Luh ee eat 5 Clolleetime: Catan, «tater ts ace Rtsu Ween ee te cn rtp eee pe ese ae ae les oRiata esa tee alonleciesindewniaenehe 5 Morpholowicaliatialysig restricted’. .sccstssseunsueosipcbensca.nos' dsdesapesnenas ce ast COLAO i ki os J Ra Rah MR as he a Ms oo yale Me Sa tanh ys aes 46 CArpON CN AON. IN de wecbssk Ledtheconeltuatbsiy own dda lbs edaeodedubtinieesevewieuietaealnst Ones vbeieediene' 46 CUGPON EL a CHIE ie tk Mee Mies teeters ete eee ae see Psese Mwaynyes taeet ie 47 Corp heii na IDONOC HG, MOL a haps ei ceateanceasean cadet at aneasanaitndehralicanapede 48 COL DUCA LOLOS SP) AL os ce saeco plese on Tasiaie pus ices ta eac oon se sie sae eee steSt td edeoxien: 49 A UNOSEN AOC Mo its sey sh teow vidos vhans spy teibuat ait ees tiut Apeldessy dachtuntaciTp hdd Wp whites tes 51 Flabellina Gray, 1833 in Griffith & Pidgeon, 1833-1834 oo... ee eeeeeeeeseeeees 51 DL ATABADCHANGSEGLY, 150. see tee se oo vgs ences tones on vente een hatte se Pants 00 ona cl dulanieMesia outage De PISAMOFCCHIEA NAALCUSS HNO ico Scere ccs Oh eee ees es Renee Pees ee 52 Pamiily,Flabellisidac=7cerigers ca is..cs. te. alesevsvsestsSsnanenadects sh lcessssuisiesiesheeenseenie. eoeeises =e) Family: Unidentidae Millen-8¢ Hermosillo 2012) isc ceccceetuteassteevecwoncwdaveesoel en 54 PACU ONT Ms as Sai sec ad fan ia deh RU a cat ha etgedne A ocades aps aneeasoeslegnedes sieadadandatashondesaaes 56 PACT UA MIAICA SPI of dkeaules Maas Raheaies ainswsl sane sdodaian dahentesitpenas vari basdarametbenchumpeente 56 Unidentia Millen 8 Hermosillo, 2012 .......cccccscccccssecccsecccccccccucccccceccccescceeesecs 57 Dnidentia nilourossija Sp. 1. ovouinesivcsevecaunia aane tne deepeneglsen suse deadsnunecdusincsineces 58 LTV NEI A SATA TAINT LORE SIN 3 Whe oe padi ncc dens cass unpltde tas Duar ovd cnt Pada eect, sel 59 Perel VR SAIL LICHENS Ug ig Cast raete pan cease one vce azat guano ceantelet Sera eeebanalaaebtatan batters 60 PUBCON ECON NIN rezau coven eeewev ong etiaerterianeeesvitdcen ube cout Rieceundaca duvide reciutlutensee Pete t ec 61 Sie ae eee SOU eh teats 15. va RO RST ya ATE DARREN PP sundaes Ts eksat on theres 62 SO AUTAAIASIICCTESD atoll ie Cans Sts ie a ee Mala Mas clea acetabular at 62 Bairrid yen pata aed ata. Wee. S Matec he etna Meecaphtie each tic teat Se trap tied BM acicaa oe 63 APUta OCLs, Sn net Pett ks te tons See ese te Mt pete vat Moe Dartotios reacties 64 APAIA PICT, ROMANAOVICA SILDSP)s Nhs a vvsas uinninnehsneantodtasbesonsscinsishinenvagosrienbodenaad 65 TAIT E UCHLI ret eh ew sss ly ona ate yelp w skootah tap elev ee c sare ov ag soo ne we R oan oes 66 ING PIE BOW, Ma Rast se sig usysacnesscaeacvesas SRee lat ees ags AUT RM UERe catawawesup eseeaedasnesle 66 WOTSCUES TO TE Ue rch asc. one Phas RE he. Sa CEE ail RENE, sccMMh cvntsa Pacectettorntenteettes 67 4 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) POTS WACO GS EECE a5 c0 haloes Elin hls La ae cradles oR eR Clas ee Ee 67 APNE WACO IC CLASSI CATION ose sez entc es ott ok ices ote tases eaten Wistiandoen teleost 70 Pattettis Of acolictev olin oit vere teem cert A necrmaee con mecerrawr et Re acesol Basak oe, 73 Se ISTO ETS Reins hate hate ota MRR 2s otk RU ACE 2 Wc: ETN EA WL Re 76 aite Ss Aaennsostee cr aie ssescecrta ec Sie ee ee ere ems eee Dee. cree eel ar INCKITOW Le GO EMTET US Sacre inet eek fun anton he wactinky Ueetsc ph ecUipe aaah eaae ssrbaee Sra duo petant costes 125 REPETeMCeg, ety: Aan ee. Le, CSE ee ee, A ey 9, Ae 125 Subp lemehtatysrira te Cal ol Wie aes cont eid Un ree eee lta Eis ales 138 SUP ple hide Mbat walt Abe mia: De oon. Lae teats ta utes ep tecec th eet as ce tee Meicicneseee ae Dee a eee etacees 138 Supplementaty material 5 i222 tcc see, toutes. censuses. t eee sgens ooestagesbeeevetSeans sues ae eee 139 Stop lemiehtatry rater ial coh bi wins eM she ate tte aes oh Shu. a a etch See aaa 139 Introduction Flabellinids are a large family of commonly occurring cnidosac-bearing nudibranchs, especially abundant and diverse in boreal and Arctic regions. Information on taxonomy of various flabellinids can be found in several reviews of opisthobranch regional faunas (e.g., Odhner 1907, 1922, 1939; Marcus and Marcus 1963; MacFarland 1966; Beh- rens 1980, 1991; Behrens and Hermosillo 2005; Gosliner 1980; Gosliner and Griffiths 1981; Thompson and Brown 1984; Picton and Morrow 1994; Rudman 1998; Schrédl 2003) as well as separate publications specially focused on the selected taxa of the fam- ily Flabellinidae (Kuzirian 1977, 1979; Picton 1980; Garcia-Gémez 1986; Hirano and Thompson 1990; Gosliner and Willan 1991; Cervera et al. 1998; DaCosta et al. 2007; Millen and Hermosillo 2007; and many others). Currently, approximately 74 species are included in Flabellinidae. Within the family Flabellinidae, the genus-level taxonomy is confused; whereas a majority of the species (64) of this large family have been placed within the single genus Flabellina (Gosliner and Griffiths 1981), it is a questionable placement since Flabellina in the current sense encompasses morphologi- cally different Arctic and tropical species (Wagele and Willan 2000). Moreover, despite the fact that many species have been listed under the genus name Filabellina Gray, 1833, several genera like the Arctic Chlamylla Bergh, 1899 and the Mediterranean Calmella Eliot, 1910 were not synonymised with Flabellina, despite the similarity of their distinguishing characters to those of some other species currently considered to belong to Flabellina. There are no complete revisions of the family Flabellinidae or novel integrative studies on this family combining both morphological and molecular data. Available reviews are restricted mostly to some warm water species based on a morphological approach (Gosliner and Kuzirian 1990; Millen and Hamann 2006). There are some molecular data on flabellinids scattered in more recent publications on other aeolidacean groups (e.g., Carmona et al. 2013) and a recent achievement in mo- lecular and morphological studies on Mediterranean and NE Atlantic species (Furfaro et al. 2017), but there are no recent attempts to understand the broad-scope taxonomic diversity of traditional Flabellinidae and their major phylogenetic patterns. Polyphyly of the traditional family Flabellinidae affects a major group of Nudibranchia... 5 The aim of this study is to use novel material from unique and rarely seen Arctic species as well as North Atlantic, North and South Pacific, and tropical Indo-West Pacific flabellinid species to investigate their anatomy and to test the phylogenetic relationships of the current Flabellinidae in a comprehensive molecular framework. It was found that the current concept of Flabellinidae is deeply polyphyletic and that its phylogeny and taxonomic diversity cannot be understood without considering mem- bers from the majority of families in Aeolidacea. Morphological data and molecular analyses presented here lay a foundation for a modern revision and reclassification of one of the largest subgroups of nudibranchs, the Aeolidacea. Materials and methods Collecting data Material for this study was obtained from various expeditions and fieldwork, and included specimens belonging to the different taxa of the family Flabellinidae. All specimens were deposited in the Zoological Museum of Lomonosov Moscow State University (ZMMU), Norwegian University of Science and Technology (NTNU), University Museum Trond- heim (NTNU-VM), National Museums Northern Ireland, Cultra, Belfast, Gothenburg Natural History Museum (GNM), and the Bavarian State Collection of Zoology, Mu- nich (ZSM). Type specimens from the Natural History Museum of Denmark (NHMD) have been also investigated. The majority of specimens of the new Coryphellidae species were collected alive at Gulen Dive Resort (Norway, north of Bergen). Other flabellinid species used here for comparison have been collected in various locations in northern Eurasia, including Banyuls-sur-Mer, Vigo, Ireland, middle Norway, Spitsbergen, Barents Sea, White Sea, Kara Sea, Laptev Sea, Chuckchi Sea, Bering Strait, Commander Islands, Bering Sea, Japan Sea, the Pacific side of the Japanese Islands as well as South America, Vietnam, Indonesia, and the Pacific coast of the USA. All necessary permissions have been obtained during the above-mentioned collections. The Balinese nudibranch speci- men was collected under the permit of Governor of Bali No. 070/4710/IV/BPMP/2016. Morphological analysis The external morphology of specimens was studied under a stereomicroscope. For the description of internal features, we dissected both preserved and fresh specimens (when available) under the stereomicroscope. ‘The buccal mass of each specimen was extracted and soaked in 10% sodium hypochlorite solution for 1-2 minutes to dissolve con- nective and muscle tissue, leaving only the radula and the jaws. The features of the jaws of each species were analysed under the stereomicroscope and scanning electron microscope, and then drawn. The coated radulae were examined and photographed us- ing a scanning electron microscope (CamScan). The reproductive systems of different 6 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) species were also examined and drawn using the stereomicroscope. In the description of reproductive characters, we consistently apply the terms “proximal receptaculum seminis” and “distal receptaculum seminis” (= bursa sensu e.g., Gosliner and Grifhths 1981 and other authors) since it was already clearly shown that according to their internal structure, both seminal reservoirs in aeolidacean nudibranchs are actually re- ceptacula with sperm attached to the wall (Schmekel and Portmann 1982; Wagele and Willan 2000; Fischer et al. 2007). Molecular analysis In total, 126 specimens were successfully sequenced for the mitochondrial genes cy- tochrome c oxidase subunit I (COI) and 16S rRNA, and the nuclear genes Histone 3 (H3) and 288 rRNA (C1—C2 domain). Additional sequences, including outgroup specimens, were obtained from GenBank (see Supplementary material 2: Table S1 for a full list of samples, localities, and voucher references). Small pieces of tissue were used for DNA extraction with Diatom DNA Prep 100 kit by Isogene Lab, according to the producer's protocols. Extracted DNA was used as a template for the amplifica- tion of partial sequences of COI, 16S, H3 and 28S (see Suppl. material 3: Table S2 for primers). Polymerase chain reaction (PCR) amplifications were carried out in a 20-pL reaction volume, which included 4 pL of 5x Screen Mix (Eurogen Lab), 0.5 uL of each primer (10 uM stock), 1 uL of genomic DNA, and 14 uL of sterile water. The amplification of COI and 28S was performed with an initial denaturation for 1 min at 95 °C, followed by 35 cycles of 15 sec at 95 °C (denaturation), 15 sec at 45 °C (annealing temperature), and 30 sec at 72 °C, with a final extension of 7 min at 72 °C. The 16S amplification began with an initial denaturation for 1 min at 95 °C, followed by 40 cycles of 15 sec at 95 °C (denaturation), 15 sec at 52 °C (annealing tempera- ture), and 30 sec at 72 °C, with a final extension of 7 min at 72 °C. Sequencing for both strands proceeded with the ABI PRISM BigDye Terminator v. 3.1. Sequencing reactions were analysed using an Applied Biosystems 3730 DNA Analyzer. Some COI sequences were produced at the Canadian Centre for DNA Barcoding (CCDB), using their automated systems for extraction, PCR, and sequencing. Protein-coding sequences were translated into amino acids for confirmation of the alignment. All sequences were deposited in GenBank (Suppl. material 2: Table S1, highlighted in bold). Original data and publicly available sequences were aligned with the MUSCLE algorithm (Edgar 2004). Separate analyses were conducted for COI (657 bp), 16S (434 bp), H3 (327 bp) and 28S (327 bp). Gblocks 0.91b (Talavera and Castresana 2007) was applied to discard poorly aligned regions for the 16S data set and for the 28S data set (using less stringent options; in total, 11% and 5% of the positions were eliminated). An additional analysis was performed with all four concatenated markers (1745 bp). Evolutionary models for each data set were selected using MrMod- elTest 2.3 (Nylander et al. 2004) under the Akaike information criterion (Akaike 1974). The GTR + I + G model was chosen for COI, 16S, 28S, H3 and for the combined Polyphyly of the traditional family Flabellinidae affects a major group of Nudibranchia... 7 dataset. Iwo different phylogenetic methods, Bayesian inference (BI) and Maximum likelihood (ML) were used to infer evolutionary relationships. Bayesian estimation of posterior probability was performed in MrBayes 3.2 (Ronquist et al. 2012). Markov chains were sampled at intervals of 1000 generations. Analysis was started with random starting trees and 6 x 10° generations. Maximum likelihood-based phylogeny infer- ence was performed in RAxML 7.2.8 (Stamatakis et al. 2008) with bootstrap in 1000 pseudo-replications. Final phylogenetic tree images were rendered in FigTree 1.4.2. Nodes in phylogenetic trees with Bayesian posterior probability values 20.96 (pp) and bootstrap values >90% (bs) were considered well-supported, nodes with 0.90-0.95 and 80-89% accordingly were considered moderately supported (lower support val- ues were considered not significant) (e.g., Furfaro et al. 2016). The program Mega7 (Kumar et al. 2016) was used to calculate the uncorrected p-distances between all the sequences and pairwise uncorrected p-distances within and between clades. Integration of morphological and molecular data (operational rules) There is an extensive body of literature regarding the importance of an integrative approach which is targeted to employ both morphological and molecular data (e.g., Dayrat 2005; Schlick-Steiner et al. 2010; Papakostas et al. 2016; Korshunova et al. 2017a, b). In aeolidaceans clear preference is currently given to the molecular data and a classification is often constructed following the molecular phylogenetic trees (e.g., Carmona et al. 2013; Cella et al. 2016). Because there is no epistemological evidence that molecular data should have preference over morphological features (e.g., Mishler 1994; Giribet 2010; Bazsalovicsova et al. 2014; Jenner 2015; Anton et al. 2016), we have developed several operational rules: 1) Both morphological and molecular data should be utilised in the resulting classification; 2) Morphologically highly aberrant taxa (e.g., family- or genus-level) nested inside numerous taxa with disparate morphology should not be united with the rest of the related taxa but kept separate to highlight significant morphological differences; 3) Taxa for which molec- ular data persistently indicate the heterogeneous nature of a traditional taxon (e.g., family level) with apparently similar morphology (“para-” or “polyphyly”) should be separated into several taxa of the same rank; 4) Large-volume genera incorporat- ing numerous species should be avoided because they considerably obscure both morphological and molecular diversity and do not properly allow the recognition of hidden diversity. The foundation of the empirical rules outlined here is rooted in the following bio- logical facts: i) Developmental genes (e.g., Homeobox, etc.) show a considerable level of conservatism over large phylogenetic distances that imply that similar morphological features may appear in taxa which are not closely related, ii) There is recent evidence on the importance not only of genetic but also epigenetic interactions, that implies that genes can be changed not just according to inferred molecular phylogenetic trees, iii) There is compelling recent evidence that evolutionary (phylogenetic) patterns of 8 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) any groups of living organisms are extremely complicated and include numerous para- and polyphyletic events, iv) Therefore, in order to construct a classification which will reflect such complicated patterns in nature and not be constructed for merely logical or didactic purposes (e.g., “convenience”, “ease of use”, etc.) the resulting classification should be equally complex. The separation of smaller classificatory groups/units also leads to increasing objec- tivity in taxonomy. Indeed, “objectivity” is a very complicated and not an equivocal term and cannot be applied to the taxonomic field without reservation, despite the as- sertions of some authors that there is an objective interpretation of the phylogeny and morphological characters. However, instead of clarification, lumping morphologically diverse genera and families leads to a decrease of objectivity since the decision as to which genus/family should be united and which should not is an extremely subjec- tive process even if a molecular phylogeny is used as the primary justification. This subjectivity was clearly demonstrated by the recent molecular phylogeny of tergipe- did aeolidacean nudibranchs (Cella et al. 2016), when some genera were united into super-lumping groups like Tenellia, whereas other closely related genera like Tergipes and Rubramoena were instead kept separate. The uniting of several previously clearly morphologically delineated families of aeolidaceans into the single family Fionidae (Cella et al. 2016) led, in turn, to the loss of any reliable morphological diagnostic features (see discussion in Korshunova et al. 2017a). The absence of the morphologi- cal synapomorphies for “Tenellia” in the sense of Cella et al. (2016) was also inde- pendently noted most recently in Goodheart et al. (2017).Thus, an interpretation of a phylogenetic framework itself is by no means an “objective process.” A putative hypothesis-driven modern taxonomic approach apparently based on clear testable hypothesis and reproducible methodological frameworks did not necessarily lead to objective decisions and classifications. Taxonomic objectivity can be increased using consistent separation of small maximally coherent morphological and molecular taxonomic groups (taxa). It is very important to highlight that the separation of a small coherent group/unit does not specially imply a bias toward splitting in the classical taxonomic lumping/split- ting dilemma. Instead, the necessity of splitting many traditional taxa is rather to conform to the molecular phylogenies which, in many cases, are also confirmed by morphological data reflecting the extremely complicated and mosaic pattern of natural evolutionary pathways which favour splitting, but at a new level supported in a modern and integrative way. Of course, there are many pitfalls in this method as well and it deserves wider discussion, but an objective truth is that apparently objective molecular-based methods do not imply a single resulting objective clas- sification. Furthermore, since we still use a rigid binomial nomenclature and taxa hierarchy developed long prior to any phylogenetic and evolutionary conceptions as an unavoidable taxonomic rule, we should attempt to adapt an archaic system into extremely complicated phylogenetic patterns recently discovered in most of the or- ganism groups. There is an immense body of literature on the relationship between taxonomy, phylogeny, and nomenclature (e.g., Simpson 1961; Hennig 1966; Ridley Polyphyly of the traditional family Flabellinidae affects a major group of Nudibranchia... 9 1986; Queiroz and Gauthier 1992; Wagele 2005; Schuh and Brower 2009; Wiley and Lieberman 2011; Hubert and Hanner 2015; and many others). More recently a proposal regarding, for example, PhyloCode (Rieppel 2006; Queiroz 2006) was widely discussed, as well as other suggestions for linking molecular and traditional taxonomy (e.g., Galimberti et al 2012; Vences et al. 2013; Hedges 2013; Franz et al. 2015) but did not lead to any substantial changes to current taxonomic nomenclatu- ral practice, which still only insignificantly differs from the Linnaean system. Our present study does not intend to review theoretical literature on this topic; however, without setting some operational rules/criteria, we cannot accommodate our results in a practical way. These operational rules are consistently applied here (as far as possible) for the taxonomy of one of the largest and most complicated traditional families of nudi- branchs, the Flabellinidae, and for the discussion of the general classification of one the major traditional subgroups of Nudibranchia, the Aeolidacea. These rules may help taxonomists with the actual integration of molecular and morphological data instead of a truly authoritative commonly held view that a taxonomist must just follow molecular phylogenetic patterns without any settled guidelines on how to convert the phylogenetic pattern into a taxonomic system and how to integrate, in many cases, considerable molecular and morphological disparity. The consistent application of the small coherent taxonomic groups concept may also help to resolve ongoing debates on the treatment of paraphyletic groups in taxonomy (e.g., Seifert et al. 2016; Ward et al. 2016) since on one hand, this concept targets avoidance of paraphyletic taxa to a maximal degree but on the other hand, it does not mask taxonomic diversity by unit- ing smaller monophyletic groups into larger, non-diagnosable units without support of morphological apomorphies (see also Korshunova et al. 2017a). An independent support for the validity of our approach appeared in a recent study while our paper was under review (Zamora-Silva and Malaquias 2017). In that study, a revision of the traditional family Aglajidae was undertaken, and numerous new genera were proposed to accommodate complex phylogenetic patterns and morpho- logical disparity. This aligns well with the principles of preferred separation of small, morphologically and molecularly coherent taxonomic groups/units proposed here. The most recent debates in amphibian taxonomy (Scherz et al. 2017) also support the small coherent unit taxonomic approach developed here. Results Molecular phylogeny In this molecular study, 205 specimens were included, combining 404 novel sequences with 230 from GenBank. A total of 90 species was selected to represent all conventional flabellinid subgroups and all aeolid families with multi-locus data available. Bayesian Inference (BI) and Maximum Likelihood (ML) analyses based on the combined dataset 10 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) for the mitochondrial genes COI and 16S, and the nuclear genes H3 and 28S yielded similar results (except the position of the Rubramoena clade, see Discussion for details) and revealed that the current family Flabellinidae is deeply polyphyletic (Figs 1-2). Highly supported and moderately supported nodes have been analysed (Fig. 1). Es- pecially interesting is the fact that putatively highly derived Flabellina sensu lato are comprised of several definitely non-related groups in relation to other non-flabellinid groups such as Aeolidiidae, Facelinidae, Tergipedidae, and others (Fig. 2), thus render- ing traditional Flabellinidae polyphyletic. The results of molecular phylogenetic analy- ses (BI and ML) support several family-level taxa. The molecular phylogenetic analyses in combination with species delimitation analysis support the presence of several new species. Furthermore, a flabellinid species Coryphella lineata (Lovén, 1846), commonly considered to be a single species, is actually a highly heterogeneous group comprised of at least four species (three of which are new) and two new genera; species complexes were also discovered among many other lineages (Fig. 1). Taxonomy of the traditional family Flabellinidae The family Flabellinidae is a morphologically very diverse assemblage and historically several genera have been created to encompass this species diversity (Bergh 1900a; Odhner 1907; Odhner 1968). The majority of the species were nevertheless separated within two major genera, Flabellina and Coryphella. The genus Flabellina was charac- terised by elevated stalk-like groups of cerata, whereas Coryphella was diagnosed as hav- ing non-elevated cerata inserted directly to the notum. Considering the extreme diver- sity within the family, a rigid classification based on just two character states (stalked vs. non-stalked cerata) was clearly inadequate and fails to acceptably place many inter- mediate taxa. For example, other genera with a continuous ample notal margin have been described, Chlamylla Bergh, 1886 and Paracoryphella Miller, 1971 (Bergh 1886, 1900a; Miller 1971) as well as several genera with the cerata on stalks or elevations, Samla Bergh, 1900, Nossis Bergh, 1902, Tularia Burn, 1966, Flabellinopsis MacFar- land, 1966 (Bergh 1900b; Bergh 1902; Burn 1964, 1966; MacFarland 1966), but they were largely not incorporated into the broad-scale taxonomy of the family Flabellini- dae. In 1981 when describing some new “intermediate” species a novel revision of the genus-level taxonomy of the family Flabellinidae was clearly needed; however, a deci- sion was made to merge the overwhelming majority of flabellinid species under just the single oldest genus name, Flabellina Gray, 1833 (Gosliner and Grifhths 1981). This decision was justified thus: “Mayr (1969) suggested that a distinct morphological gap should exist between genera. The presence of intermediate forms with poor correlation of morphological characteristics suggests that maintenance of the generic separation of Coryphella and Flabellina is untenable.” (Gosliner and Griffiths 1981: 110). However, apart from this theoretical reasoning, which by no means represented any obligatory rule that any taxonomist should strictly follow and which also implies very subjective decisions, a thorough revision of the generic classification of the family Flabellinidae Polyphyly of the traditional family Flabellinidae affects a major group of Nudibranchia... 11 was absent and the majority of flabellinid species were merely listed without any de- tailed discussion under the name Fiabellina (Gosliner and Griffiths 1981). The present phylogenetic analysis of a broad selection of various Arctic and tropical flabellinid taxa reveals that fundamentally different flabellinid clades have been concealed under apparent “intermediate forms” (see below and Figs 1—2). One of the most remark- able results of the present analysis is that members of the traditional family “Flabellini- dae” are phylogenetically extremely heterogeneous according to the molecular data, con- firming and even extending earlier morphology-based assumptions (Wagele and Willan 2000; Martynov 2006a). The highly derived Flabellina sensu stricto are comprised of several non-related groups, and the inclusion of several disparate non-flabellinid families (Fig. 2) makes traditional Flabellinidae at least triply polyphyletic. Thus, the previously unchallenged decision to merge all diversity of the traditional family Flabellinidae has considerably masked not only the generic diversity within the traditional Flabellinidae (which is higher than currently recognised), but also Flabellinidae sensu Jato actually embraced most other aeolid family-level taxa. This conclusion is foreshadowed by the anomalous position of Flabellina babai revealed by Carmona et al. (2013) and the trees presented by Furfaro et al. (2017) while our present study was under review. Previous researchers who focused on flabellinids may have suspected that the taxon was paraphyletic; if so, they did not propose a satisfactory solution to incorporate those doubts. For example, while almost all flabellinid species were placed under the name Flabellina by Gosliner and Griffiths (1981), they did not consider several genera which already existed, e.g., Chlamylla and Calmella. As a result, while the genus Chlamylla was not synonymised with Flabellina, one species, Coryphella orientalis Volodchenko, 1941, was renamed as Flabellina orientalis (Volodchenko 1941) and a junior objective synonym Coryphella barentsi Derjugin & Gurjanova, 1926 was given a new name Flabellina in- cognita (Gosliner & Griffiths, 1981: 112). However, Coryphella orientalis Volodchenko, 1941 is actually a junior synonym of Chlamylla atypica (Volodchenko 1941; Martynov 2006a: 283) and Flabellina incognita is, according to the radular morphology and pres- ence of an external penial collar, also a member of the genus Ch/amylla. Thus, at least two species from the genus Chlamylla were listed under the genus name Flabellina, yet the genus Chlamylla was neither synonymised with Flabellina nor even mentioned in Gosliner and Griffiths (1981) or any subsequent papers. The same treatment was applied to some stalked/elevated cerata-bearing genera, e.g., Calmella and Tularia. The Australian and New Zealand genus and species Tularia bractea (Burn, 1966) which possesses all of the advanced flabellinid characters (triserial radula, cerata on raised elevations, reproduc- tive system without penial gland) was also either not discussed or not included in the genus Flabellina. Therefore, a genus-level revision of the family is highly desirable. Several authors continued to use at least the genus name Coryphella as separate from Flabellina (Thompson and Brown 1984; Roginskaya 1987; Picton and Morrow 1994; Martynov 2006a; Martynov and Korshunova 2011); however, the very poorly defined Flabellina sensu lato dominates in common usage. Most importantly, this question is not just purely theoretical or redundant, but instead deeply affects the core of practical aeolid taxonomy. For example, assessment of the phylogenetic relationship of the flabellinid species com- 12 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) plex C. lineata (Lovén, 1846), performed in the course of the present study, was impos- sible without a broad comparison between different flabellinid taxa. Our molecular phylogeny thus shows that flabellinids with elevated/stalked cerata are clearly polyphyletic and cannot therefore be maintained within the same genus or even within the same family (Fig. 2). Actually, there are also several clades of flabellinids with continuous rows of cerata, with significant genetic gaps, that are now separated. The current practice of genus separation into many different groups is very far from the simplistic consideration that the presence of some “intermediate forms” is sufficient for synonymising genera. Instead, differences used for establishing the numerous new genera in a group in most recent papers (e.g., Bourguignon et al. 2016; Zamora-Silva and Malaquias 2017) shows that genera can be distinguished by subtle morphological characters. The dominant current classification of the family Flabellinidae is also very incompatible with the approach that is widely utilised in the family Facelinidae, one of the closest to the flabellinids, where numerous genera including monotypic ones are currently widely utilised (e.g., Millen and Hamann 1992; Rudman 1998; Hirano 1999; Millen and Hermosillo 2012; see Supplementary materials). Furthermore, there are recent proposals (e.g., Carmona et al. 2013, Knutson and Gosliner 2014), although not universally accepted, that COI divergence more than 11.1 + 5.1% in molluscs is sufficient for separation of genus-level taxa. This implies broad limits of the molecular foundations of the genus-level taxa since divergence of more than ca. 5% may imply status of a separate genus for a particular species or a species complex. Thus, actual taxonomic practice of how many genera can be maintained in any group is not a field of universally accepted and clear rules, but instead a very complicated mixture of tradi- tional authority and variously interpreted morphological and molecular data. Given the great morphological and molecular diversity of the family Flabellini- dae, independently arising Flabellina-like taxa with elevated/stalked cerata with very significant molecular divergence (Fig. 1) and similar significant divergence between different taxa with continuous cerata (Fig. 2), it is impossible to maintain only the traditional pair of taxa Flabellina-Coryphella but equally impossible to merge all flabel- linid diversity into the single genus Flabellina and the single family Flabellinidae. The para- and polyphyletic nature of the genus Coryphella itself, after the exclusion of Fla- bellina, has already been noticed (Gosliner and Willan 1991). Not only molecular, but morphological differences within the family Flabellinidae are so significant that several species are formally still included under the generic name Flabellina, but actually far exceed the taxonomic diagnosis of the family Flabellinidae. For example, Flabellina rubrolineata (O’ Donoghue, 1929) and several closely related species uniquely possess a putative triaulic reproductive system (Gosliner and Willan 1991). This character nor- mally present in completely different dorid nudibranchs (Schmekel 1970, 1971), was originally described under the new genus Coryphellina O’Donoghue, 1929 but this is currently universally included within Flabellina. The recently described Flabellina god- dardi Gosliner, 2010 has a uniserial radula, a very important character, never present in any other flabellinid taxa, which instead invariably possess a triserial radula (Gosliner 2010). There were preliminary data (Gonzalez-Duarte et al. 2008) that another genus Polyphyly of the traditional family Flabellinidae affects a major group of Nudibranchia... 13 with uniserial radula, Piseinotecus, which is still traditionally placed within the separate family Piseinotecidae, is closely related to some Flabellina species. Most recently, Fur- faro et al. (2017) confirmed this and have shown that in most of the NE Atlantic and Mediterranean species previously assigned to genus Piseinotecus the true triserial radula is present, and delicate lateral denticles were just overlooked in previous studies. Thus, the only plausible alternative is a careful distinction of family-level groups and numerous genus-level taxa within the traditional family Flabellinidae employing both molecular and morphological evidence (Fig. 2). This approach is especially im- portant when the number of morphologically difficult-to-distinguish species is con- tinually growing and to attribute new species correctly requires increasingly narrower definitions of the genera used. Thus, having a morphologically disparate, large-volume genus we can never consistently describe many potential cryptic species (for usage of the term cryptic species see Korshunova et al. 2017b). Instead, narrowly-defined genera are more consistent with an approach to recognise potential species within mor- phologically homogeneous narrow genera. The theoretical and practical framework developed here is consistently applied below in detail for the Flabellinidae s. / complex. Importantly, although this study does not include molecular data on all species of traditional Flabellinidae, we present the largest taxon selection of traditional Flabelli- nidae and related groups (ranging from the North Pole to the tip of southern America through tropical regions) ever studied. Such an approach allows us to integrate mor- phological and molecular data and in most cases to suggest generic placement of spe- cies for which molecular data are not yet available. Family Paracoryphellidae Miller, 1971, reinstated Diagnosis. Body wide. Notal edge present, well-defined, continuous. Cerata not stalked, in continuous numerous rows. Rhinophores smooth to wrinkled. Anus pleu- roproctic under the notal edge. No distinct oral glands. Radula formula 1.1.1. Asym- metrically placed additional 1-3 rows of small reduced lateral teeth may be present. Rachidian teeth with strong cusp, never compressed by adjacent lateral denticles. Lat- eral teeth narrow or with attenuated process basally, usually denticulated. Commonly only single distal receptaculum seminis present. Vas deferens always long, with wide granulated or tubular prostate. External permanent penial collar present in some taxa. Penis elongated conical, internal or fully external, unarmed. Genera included. Chlamylla Bergh, 1886, Paracoryphella Miller, 1971, Polaria gen. n., Ziminella gen. n. Chlamylla Bergh, 1886 Figs 2—7 Type species. Chlamylla borealis Bergh, 1886 14 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) Diagnosis. Body wide. Notal edge present, well-defined, continuous. Cerata not stalked, continuous. Rhinophores smooth to wrinkled, longer than oral tentacles. An- terior foot corners absent. Anus pleuroproctic under the notal edge. Rachidian teeth with strong denticulated cusp; lateral denticles not clearly delineated from cusp. Lateral teeth weakly denticulated to smooth without attenuated process basally. Single distal receptaculum seminis. Wide granulated prostate. Thin, long vas deferens clearly sepa- rated from prostate. External permanent penial collar. Penis elongated conical, internal. Species included. Chlamylla borealis borealis Bergh, 1886, stat. n. (= Gonieo- lis atypica Bergh, 1899, syn. n.) (Fig. 3) (original description in Bergh 1886, 1899, 1900a), Ch. borealis orientalis (Volodchenko, 1941), comb. n. (Fig. 4) (original des- cription in Volodchenko 1941), Chlamylla intermedia (Bergh, 1899), comb. n. (Figs 5, 6) (original description in Bergh 1899, 1900a). Remarks. Bergh (1886) described the genus Chlamylla based on the species Ch. borealis with a continuous notum, wide granulated prostate, and aberrant radula (lateral teeth with very broad bases, rachidian teeth completely lacking denticles and having instead a pair of unusual long processes (Bergh 1886: Taf. 1, fig. 14). The true identity of the type species Ch. borealis with our extensive recent paracoryphellid specimens from the Arctic is a complicated question. Such radular features are completely unknown in the group of the Arctic Paracoryphellidae usually assigned to the genus Chlamylla (Roginskaya 1987; Martynov 2006a) Figs 1, 3H, 4H, 5H, 6E). However, the general shape of the body, shape of the jaws, presence of a wide granulated prostate, and penial collar in Ch. borealis agree with specimens that are currently assigned to the species Ch. atypica and Ch. intermedia. Bergh in the same work (1886) described the radula of another species “Goniaeolis typica” in detail (Bergh 1886: ‘Taf. 3, fig. 14) as a normal triserial radula. However, Bergh (1886) also noted that the radula of Ch. borealis was in a poor condition. This may imply that either Bergh studied an abnormal specimen or had damaged the radula in some way during preparation. For this study we specially investigated the holotype of Ch. borealis (NHMD GAS-2055). It is dry and heavily dissected, but a separate cut-off of the external penial collar is left; the penis and dam- aged jaws are preserved. Comparison of available information from the holotype of C/. borealis with figures from the original description (Bergh 1886: Taf. 1, fig. 22) confirms the presence of an external penial collar with caudal genital fold. According to our data only two paracoryphellid species with a complex folded penial collar are known from Arctic seas and are currently identified as Chlamylla atypica and Ch. intermedia (Figs 3, 4C, 4I, 5J, 6K, 7). Morphologically these taxa differ from the rest of the traditional fla- bellinids, forming a distinct compact clade according to the present molecular analysis (Figs 1, 2), which demonstrates significant molecular divergence (more than 11%) from other members of the family Paracoryphellidae. Based on the details in the original de- scription of Chlamylla and type material, we can conclude, under supposition that the radula was malformed or wrongly processed, that the type species Ch. borealis belongs to the same genus as known paracoryphellid species from the Arctic, currently identified as Ch. atypica and Ch. intermedia. Given that Ch. borealis was never found again, but inhabited the same region our other samples, it is highly likely that Ch. borealis is actu- Polyphyly of the traditional family Flabellinidae affects a major group of Nudibranchia... 15 ally conspecific with one of the two currently known Chlamylla species. One of these species, Ch. intermedia (Bergh, 1899) possesses an external penial collar but without any traces of a caudal genital fold (Fig. 5J), whereas Ch. borealis, according to both its original description (Bergh 1886: Taf. 1, fig. 22a) and our novel information from the holotype, possesses a short but evident caudal genital fold. Thus, we can conclude that Ch. intermedia cannot be a synonym of Ch. borealis. Another species, Ch. atypica which was described under the genus Goniaeolis from the Davis Strait, Greenland (Bergh 1899, 1900a) readily differs from other Chlamylla species by the presence of a special long ex- ternal genital fold towards the anal opening (Bergh 1899, 1900a: Tab 4, fig. 6) (Fig. 3). Because Ch. borealis also possesses a genital fold, and given a high similarity of prostate patterns between our specimens previously identified as Ch. atypica (Fig. 3J) and the fig- ure of the reproductive system with the characteristically bent prostate as in Ch. borealis in Bergh’s original description of Ch. borealis (Bergh 1886, Taf. 1, fig. 21) we therefore conclude that Ch. atypica is most likely is a junior synonym of Ch. borealis. The differ- ences between length of the genital fold of the holotype of CA. atypica (also investigated in the present study, NHMD GAS-2090) and Ch. borealis is possibly due to consider- able differences in the length of holotypes of Ch. borealis and Ch. atypica (the former is nearly two times shorter than the latter). Smaller specimens previously identified as C/. atypica s. l. may possess a considerably shorter genital fold, especially in the preserved state (Fig. 4C). Therefore, in order to preserve current usage of the genus Chlamylla that has already appeared in a number of publications on Russian nudibranch fauna, we therefore synonymise here the species Ch. atypica with Ch. borealis. The Japan Sea specimens are consistent with the Arctic specimens in the presence of the external genital fold, but due to minor differences in the radula and also a very large geographic gap we consider it as a subspecies Chlamylla borealis orientalis (Volod- chenko, 1941), comb. n. (Fig. 4). Chlamylla borealis orientalis is locally abundant dur- ing winter in Northern Japan (Hirano 1997, as Ch. atypica) and in the Russian part of the Sea of Japan (Martynoy, unpublished course work (1991) as Ch. atypica; Martynov and Korshunova 2011; present study). Several specimens of other Chlamylla without such a fold from the Arctic seas (distributed at least from the Barents Sea to Laptev Sea) have no significant molecular differences between them (Figs 1, 2) and have consistent morphology (Figs 5—G) and clearly belong to the same species. The oldest name for this Chlamylla without a genital fold is Goniaeolis intermedia Bergh, 1899 also from the Davis Strait, Greenland. The lateral teeth of Goniaeolis intermedia with a broad base and slightly attenuated lateral processes (Bergh 1899: tab 4, fig. 16) are similar to our material from the Arctic seas (Figs 5H, 6E). However, G. intermedia lacked denticles on the lateral teeth. The Arctic specimens show small, sometimes almost diminishing denticles (Fig. GE). Such denticles are not always clearly evident. Possibly this species reaches at least the Bering Strait and potentially may enter the coldest shelf waters of the NW Pacific (ie., Bering and Okhotsk Sea, Martynov 2006a). Coryphella bar- entsi Derjugin, 1924 (Derjugin 1924a, b, preoccupied by Coryphella barentsi Vayssiére, 1913 (Vayssiére 1913, see Gosliner and Griffiths 1981 suggesting a replacement name) is a possible synonym of Ch. intermedia (Bergh, 1899). 16 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) Paracoryphella Miller, 1971 Figs 2, 7-10 Type species. Coryphella islandica Odhner, 1937 Diagnosis. Body wide. Notal edge present, well-defined, continuous. Cerata not stalked, continuous. Rhinophores smooth to wrinkled, shorter than or similar in size to oral tentacles. Anterior foot corners present. Anus pleuroproctic under the notal edge. Rachidian teeth with strong cusp; lateral denticles not clearly delineated from cusp. Lateral teeth weakly denticulated without attenuated process basally. Reduced additional rows of of small lateral teeth may present. Single distal receptaculum semi- nis. Long vas deferens without separate granulated prostate. Penis not internal, perma- nently attached externally. Species included. Paracoryphella ignicrystalla sp. n. (Fig. 8), P islandica (Odhner, 1937) (Fig. 9) (original description in Odhner 1937), Paracoryphella parva (Hadfield, 1963), comb. n. (original description in Hadfield 1963) (Fig. 10). Remarks. The genus Paracoryphella and the family Paracoryphellidae were initially proposed by Miller (1971) because of the putative presence of a second asymmetrical row of lateral teeth in the original description of the species “Coryphella” islandica (see Odhner 1937). However, the latter character is not very evident compared to the true unique feature of the genus Paracoryphella, a non-retractable, permanently external pe- nis, which is attached directly to the body wall and does not possesses any penial sheath (Figs 8J, 9M, 10C). Importantly, all three known species of this genus invariably pos- sess this feature. This character is unique not only within traditional flabellinids, but also within the majority of Aeolidacea. Only a single species of the family Notaeoli- diidae also has such an external penis (Wagele 1990). Furthermore, several members of the notaspid family Pleurobranchidae with an internal shell (Martynov and Schrédl 2008) and very basal Acteonidae with an external solid shell also possess an external penis. Therefore, this character within the genus Paracoryphella may be either a basal plesiomorphy, or an ontogenetic reversion to the basal plesiomorphy. Molecular data shows that in either case it occurs within one of the most basal clades of the traditional flabellinids. Additionally, analysis of the light microscopy images of the radula of both the type species P islandica and the new species confirm the possible presence of 1-3 very reduced, asymmetrically placed, additional rows of lateral teeth (Figs 8E 9K). During preparation for the SEM study these apparently reduced additional teeth be- came fully indistinguishable. Their correspondence to the normal lateral teeth needs to be further investigated. Coryphella parva, only known from its original description from Swedish waters (Hadfield 1963), was described as having a permanent external penis and therefore is included here in the genus Paracoryphella. Here, for the first time, we have studied type material of C. parva from the Natural History Museum of Denmark (NHMD-91476) (Fig. 10) and confirmed that it possesses a non-retractable external penis (Fig. 10C) and other external features that align with the diagnosis of the genus Paracoryphella. The drawings of the radula (Fig. 10D, E) and reproductive systems (Fig. 10F) in the Polyphyly of the traditional family Flabellinidae affects a major group of Nudibranchia... 17 original description of P parva in Hadfield (1963) are also very consistent with two other species of the genus Paracoryphella. Remarkably, even though the length of living specimens of P parva do not exceed 3.5 mm (fixed not more than 2 mm, Fig. 10A, B), the animals at that size were fully mature and produced egg masses (Hadfield 1963). Because both P ignicrystalla sp. n. and the type species of the genus P islandica reach mature size in specimens at least three times larger than P parva, the latter species thus is clearly a separate one and may represent an example of a partial paedomorphosis. According to the molecular phylogenetic analysis, the genus Paracoryphella is the sister of the Chlamylla clade (Figs 1, 2, 7). We retain the genus Paracoryphella as separate because of unique morphological characteristics including a permanent external penis and also the apparent presence of additional rudimentary lateral teeth rows. Paracoryphella ignicrystalla sp. n. http://zoobank.org/3 C9I4E2ZE9-C880-40F0-9557-3B39C339B7C0 Fig. 8 Type material. Holotype, ZMMU Op-490, 11.5 mm long (fixed), The Sea of Japan, Vostok Bay, intertidal, 17.03.1994, coll. A.V. Martynov. 1 paratype, ZMMU Op-491, 5 mm long (fixed, dissected), The Sea of Japan, Vostok Bay, intertidal, 14.03.1994, coll. A.V. Martynov. 1 paratype, ZMMU Op-492, 12 mm long (fixed, dissected), The Sea of Japan, Vostok Bay, intertidal, 18.02.1990, coll. A.V. Martynov. Type locality. The Sea of Japan, Vostok Bay. Etymology. From igni (= fire, Latin) and crystallum (= ice, rock crystal, Latin), in reference to the double combination of peculiar morphological and ecological fea- tures: short flame-like cerata with icy speckles on dorsum and peculiar environmental characteristics of the type locality which combines icy sea water temperatures (down to -2 °C) in winter and warm subtropical conditions in summer (water temperature up to +26 °C) as an allusion to the George R. R. Martin “A Song of Ice and Fire” novels. Diagnosis. Continuous notal edge, colour translucent white with scattered opaque white dots, cerata orange-brown to reddish-brown, rachidian tooth with up to 12 den- ticles not clearly delineated from relatively low central cusp, lateral teeth with few distinct basal denticles, distal receptaculum seminis, penis not internal, permanently attached externally. Description. External morphology. Body wide. Foot and tail wide, anterior foot corners short. Oral tentacles long. Rhinophores ca. 1.5 times shorter than oral tenta- cles, smooth to slightly wrinkled. Dorsal cerata fusiform, relatively short, continuously attached to well-defined uninterrupted notal edge without forming clusters. Apices of cerata pointed. Notum narrow but distinct throughout both lateral sides of body. Digestive gland diverticulum fills significant volume of the cerata. Anal opening on right side below notal edge close to middle body part. Reproductive openings lateral and non-retractable penis below second ceratal row. Tail short and pointed, extending only a short distance beyond last cerata. 18 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) Colour (Fig. 8A). Background colour translucent white. Digestive gland diverticula orange-brown to reddish-brown. Small opaque white spots cover the entire dorsum, commonly on ceratal bases and less on cerata. Rhinophores and oral tentacles similar in colour to body; apical parts covered with opaque white pigment. Apical parts of cerata without opaque cap of white pigment. Jaws (Fig. 8E). Masticatory process more than one-third as long as jaw body. Edge of masticatory processes bears ca. 40-50 denticles that continue to form several re- duced rows of denticles on the body of the masticatory processes. Radula (Fig. 8F, H). Radula formula: 10-12 x 1.1.1(2—3). Rachidian tooth elon- gate-triangular with strong non-compressed cusp of nealy 1/3 of the tooth length (Fig. 8G). Rachidian tooth bears up to 15 well-defined separated (but adpressed to- wards the cusp) long lateral denticles. Cusp is not clearly delineated from the adjacent first lateral denticles. Lateral teeth (Fig. 8E H) narrowly triangular with peculiar wid- ened base and few indistinct denticles on internal edge. There are one to three rudi- mentary additional lateral teeth on the right side only. Reproductive system (Fig. 81, J). Diaulic. Hermaphroditic duct leads to strong con- voluted ampulla of about two whorls. Vas deferens is relatively long, no distinct pros- tate. No penial sheath. Penis is attached to the external body wall, vas deferens enters the base of penis from the internal side. Oviduct connects through insemination duct into female gland complex. Vagina short and indistinct. Distal receptaculum seminis. Ecology. Stony intertidal to 5—6 m. Feeds on athecate solitary hydroids. This spe- cies is locally abundant. Egg mass is white to pinkish narrow cord. Reproduction pe- riod from December to April. Development is about one month. ‘The larva is a plank- totrophic veliger with spiral shell. Distribution. Northwest part of the Sea of Japan. Remarks. Paracoryphella ignicrystalla sp. n. clearly differs from the type species of the genus P islandica (Fig. 9K, H) in having considerably shorter cusps of the rachidian teeth (Fig. 8F, G). We have not yet obtained molecular data for this Sea of Japan Para- coryphella, but regard the morphological differences sufficient to warrant a new species. Polaria gen. n. http://zoobank.org/716D049A-DF2B-489C-ABC8-B76B 1 E830073 Figs 2,75 bl Type species. Coryphella polaris Volodchenko, 1946 Etymology. After the northern Polar region, the predominant area of distribution of this genus. Diagnosis. Body wide. Notal edge present, well-defined, continuous. Cerata not stalked, continuous. Rhinophores smooth to wrinkled, longer than oral tentacles. Anterior foot corners present. Anus pleuroproctic under the notal edge. Rachidian teeth with strong smooth cusp and distinct denticles. Lateral teeth strongly denticu- lated with considerably attenuated process basally. Single distal receptaculum seminis. Polyphyly of the traditional family Flabellinidae affects a major group of Nudibranchia... 19 Long vas deferens without separate granulated prostate. No penial collar. Penis elon- gated conical. Species included. Polaria polaris (Volodchenko, 1946), comb. n. (Figs 7, 11) (original description in Volodchenko 1946). Remarks. ‘The type and single species of the genus Polaria, Coryphella polaris is the only available valid name (Volodchenko 1946) for the species Goniaeolis typica M. Sars, 1861 as incorrectly identified by Bergh (1886), Odhner (1907), and Rogins- kaya (1987, 1997). True Goniaeolis typica possesses a triserial radula but has no diges- tive gland branches or cnidosacs in the cerata and the shape of radular teeth is very dif- ferent from any traditional flabellinid (M. Sars 1861; G.O. Sars 1872; Odhner 1922). The name Coryphella polaris Voldochenko, 1946 was therefore resurrected to avoid this misidentification (Martynov 2006a). The genus Po/aria forms a separate clade within Paracoryphellidae according to the molecular phylogenetic analysis (Figs 1, 2). By combination of relatively long vas deferens without distinct prostate (Figs 7, 10H, I), retractable penis without external collar, and lateral teeth with strongly attenuated process (Fig. 10G), the genus Polaria morphologically differs from all other genera of the family Paracoryphellidae. Ziminella gen. n. http://zoobank.org/62A50536-3E18-4A03-8FD2-B27D43357628 Figs 2, 12-15 Type species. Eolis salmonacea Couthouy, 1838 Etymology. In honour of Olga Zimina, scientist at Murmansk Marine Biology Institute; she made a considerable contribution in collecting Arctic paracoryphellid species for this study. Diagnosis. Body wide. Notal edge present, well-defined, continuous. Cerata not stalked, continuous. Rhinophores smooth to wrinkled, similar in size to oral tentacles. Anterior foot corners present. Anus pleuroproctic under the notal edge. Rachidian teeth with strong denticulated cusp; lateral denticles not clearly delineated from cusp. Lateral teeth weakly denticulated to smooth without attenuated process basally, sig- nificantly smaller than rachidian teeth. Receptaculum seminis not evident. Long vas deferens without separate granulated prostate. No penial collar. Penis folded or elon- gated conical. Species included. Ziminella abyssa sp. n. (Fig. 12), Ziminella circapolaris sp. n. (Fig. 13), Ziminella japonica (Volodchenko, 1941), comb. n. (Fig. 14) (original de- scription in Volodchenko 1941; lectotype designated in Martynov 2013), Ziminella salmonacea (Couthouy, 1838), comb. n. (Fig. 15) (original description in Couthouy 1838; redescription in Kuzirian 1979). Remarks. Ziminella considerably differs morphologically from the genus Chlamylla by the absence of a granulated prostate and external penial collar (Fig. 7) and by the shape of the radular teeth, from the genus Paracoryphella by the presence of a penial 20 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) sheath and by the shape of the radular teeth, from the genus Polaria by the absence of a very long, conspicuous distal receptaculum seminis and by the shape of the radular teeth. On the molecular tree, both species of the genus Ziminella place as a well-separat- ed clade within the family Paracoryphellidae (Figs 1, 2). Four species currently known within the genus Ziminella are well-separated by the penial morphology (Z. salmonacea possesses a slightly folded penis, whereas Z. abyssa and Z. japonica have an entire conical penis; for discussion see Martynov 2013). Molecular data (Figs 1, 2) published here for the first time show that three species are placed together within a larger clade. Ziminella abyssa sp. n. http://zoobank.org/B22A7B85-F09B-4604-BE8C-6D5B560ABBDD Fig. 12 Type material. Holotype, ZSM Mol-20100647, 19 mm long (fixed), R/V “Akademik Lavrentyev’, sta. B5—10, 24.08.2010, The Sea of Japan, depth 2676 m. 8 paratypes, ZMMU Op-248, up to 20 mm long (fixed), R/V “Vityaz”, sta. 6657, 15.06.1972, The Sea of Japan, depth 3560 m. 9 paratypes, ZMMU Op-249, up to 24 mm long (fixed), R/V “Vityaz’, sta.7462, 29.05.1976, The Sea of Japan, 41°19.8'N, 131°15.2'E, depth 3290 m. 57 paratypes, ZMMU Op-250, up to 25 mm long (fixed), R/V “Vityaz”, sta. 7463, 29.05.1976, The Sea of Japan, 40°45.0'N, 131°16.0'E, depth 3300 m. 1 paratype, ZMMU Op-252, 9 mm long (fixed), R/V “Vityaz’”, sta. 7476, 06.06.1976, The Sea of Japan, 40°39.5'N, 132°32.8'E, depth 3425 m. 7 paratypes, ZMMU Op- 253, up to 15 mm long (fixed), R/V “Vityaz’, sta. 7479, 07.06.1976, The Sea of Ja- pan, 39°09.1'N, 131°03.2'E, depth 3120 m. 2 paratypes, ZMMU Op-255, up to 14 mm long (fixed), R/V “Vityaz”, sta. 7492, 12.06.1976, The Sea of Japan, 38°21.8'N, 134°49.5'E, depth 3020-3030 m. 2 paratypes, ZMMU Op-256, up to 19 mm long (fixed), R/V “Vityaz’, sta. 7494, 14.06.1976, The Sea of Japan, 38°48.0'N, 136°04.6'E, depth 2740 m. Two paratypes, ZMMU Op-257, 15 mm long (fixed), R/V “Vityaz”, sta. 7496, 01.07.1976, The Sea of Japan, 40°36.60'N, 139°00.002'E, depth 3340 m. 48 paratypes, ZMMU Op-258, up to 21 mm long (fixed), R/V “Vityaz”, sta. 7517, The Sea of Japan, 42°28.2'N, 138°20.9'E, depth 3620 m. 6 paratypes, ZMMU Op- 259, up to 18 mm long (fixed), R/V “Vityaz’, sta. 7519, 02.06.1976, The Sea of Japan, 41°28.8'N, 136°06.1'E, depth 3460 m. 11 paratypes, ZMMU Op-264, up to 12 mm long (fixed), R/V “Akademik Lavrentyev’ (SoJaBio), sta. A3-11, 14.06.2010, The Sea of Japan, 44°47.6338'N, 137°15.3182'E, depth 1494-1525 m. Type locality. The Sea of Japan. Etymology. From abyssum (= depth, Latin) in reference to abyssal habitat of the new species, one of the deepest among aeolidacean nudibranchs. Diagnosis. Continuous notal edge, lateral branches of digestive gland (ceratal ba- sis) dark violet, rachidian tooth with up to 30 (and more) fold-like or fork-shaped denticles clearly delineated from central cusp, lateral teeth with few denticles on teeth edge, receptaculum seminis not evident, penis elongate conical. Polyphyly of the traditional family Flabellinidae affects a major group of Nudibranchia... 21 Description. External morphology (Fig. 12A, B). Body wide. Foot and tail wide, anterior foot corners short. Rhinophores similar in size to oral tentacles, smooth to slightly wrinkled. Dorsal cerata elongate, thin, continuously attached to well-defined uninterrupted notal edge without forming clusters. Apices of cerata pointed. Notum narrow but distinct throughout both lateral sides of body. Digestive gland diverticu- lum fills significant volume of the cerata. Anal opening on right side below notal edge close to middle body part. Reproductive openings on right side. Tail long and pointed, extending only short distance beyond last cerata Colour. Background colour translucent milky white. Digestive gland diverticula probably reddish. Rhinophores light orange at the bases and pale on most of the rest of the length. Lateral branches of digestive gland (that shine through lateral dorsum sides) characteristically dark violet (estimated from a field colour photograph). Jaws (Fig. 12C, D). Jaws broad, yellowish in colour. Masticatory processes of jaws covered with several compound denticles. Radula (Fig. 12E—H). Radula formula up to 40 x 1.1.1 (in adult specimen 20-24 mm in length). Rachidian tooth with strongly protracted non-compressed cusp of ca. 1/3 of the tooth length (Fig. 12E). Rachidian tooth bears up to 30 (and more) special, fold-like or fork-like lateral denticles, which are sometimes greatly disordered. Larger denticles typically intermingled with the smaller ones. Cusp is clearly delineated from the adjacent first lateral denticles. Lateral teeth (Fig. 12F—H) narrowly triangular with few small denticles on internal edge. Reproductive system (Fig. 12J). Diaulic. Hermaphroditic duct leads to long, rela- tively narrow convoluted ampulla of several whorls. Vas deferens extremely long, with- out distinct prostate. Penial sheath elongated. Penis entire, conical. Oviduct connects through insemination duct into female gland complex. No distal or proximal recep- taculum seminis detected. Ecology. Deep sea basins at 1494-3620 m depth, soft bottom. This species is most common in the deepest parts of the Sea of Japan at about 3000 m. Upper bathymetric limit needs to be refined. Feeds on sea anemones of the family Edwardsiidae. Distribution. Central deepest basins of the Sea of Japan. Remarks. By presence of an entire copulative organ Z. abyssa sp. n. is similar to Z. japonica (Volodchenko, 1941), but clearly differs in the pattern of the rachidian radular teeth. While Z. japonica has regular lateral denticles of the rachidian teeth, all similar in size, no more 20 in number even in very large specimens (30 mm), Z. abyssa sp. n. has highly irregular lateral denticles of the rachidian teeth, different in size, fold- and fork- shape (Fig. 12F, G), sometimes even placed disorderly (Fig. 12H), up to at least 30 (and more) in number even in specimens which are much smaller than in the previous spe- cies with specimens reaching 20-24 mm in length. Remarkably, the irregularity of the rachidian teeth in Z. abyssa sp. n. persists even in very small juveniles (2 mm), in which larger denticles already intermingle with the smaller ones (Fig. 121). The morphological differences between these two species agree with the considerable bathymetric differences: Z. japonica is a shelf and upper bathyal species (ca. 100-500 m), whereas Z. abyssa sp. n. inhabits predominantly the deepest (approximately 3000 m) basins of the Sea of Japan. 22 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) Ziminella circapolaris sp. n. http://zoobank.org/E376DC16-2736-4109-B09F-81B77B326A0B Fig. 13 Type material. Holotype, ZMMU Op-598, 35 mm long (fixed), Arctic Ocean, Franz Josef Land, Wiltona Island, 23.08.2013, depth 18-25 m, collected by O.V. Savinkin. 1 paratype, ZMMU Op-482, 14 mm long (fixed), Arctic Ocean, Franz Josef Land, Northbrook Island, 26.08.2013, depth 18-23 m, collected by O.V. Savinkin. 1 Para- type, ZMMU Op-483, 12 mm long (preserved), Arctic Ocean, Franz Josef Land, Pio- neer Island, 17.08.2013, depth 21-23 m, collected by O.V. Savinkin. Type locality. Franz Josef Land. Etymology. From circa (= near, Latin) and polaris (= polar, Latin) in reference to the proximity of the habitat of the new species (Franz Josef Land) to the North Pole. Diagnosis. Continuous notal edge, colour yellowish, cerata reddish-brown, few apical white dots, rachidian tooth with up to ten denticles clearly delineated from cen- tral cusp, lateral teeth with numerous denticles (up to 24) which cover whole edges of lateral teeth, receptaculum seminis not evident, penis folded. Description. External morphology. Body wide. Foot and tail wide, anterior foot corners short. Rhinophores similar in size to oral tentacles, smooth to slightly wrinkled. Dorsal cerata elongate, thick, continuously attached to well-defined uninterrupted no- tal edge without forming clusters. Notum narrow but distinct throughout both lateral sides of body. Digestive gland diverticulum fills significant volume of cerata. Anal open- ing on right side below notal edge close to middle body part. Reproductive openings on right side. Tail long and pointed, extending only a short distance beyond last cerata. Colour (Fig. 13A, B). Background colour yellowish-white. Digestive gland diver- ticula reddish-brown. Rhinophores light orange-yellowish. Lateral branches of diges- tive gland (that shine through lateral dorsum sides) not distinct. Jaws (Fig. 13E, F). Jaws broad, yellowish in colour. Masticatory processes of jaws covered with several simple denticles. Radula (Fig. 13G-—J). Radula formula up 26 x 1.1.1 (specimens 20-24 mm in length). Rachidian tooth elongate with strongly protracted, pointed, non-compressed cusp of almost 1/3 of tooth length (Fig. 13G, I). Rachidian tooth bears five to ten lateral denticles. Larger denticles not intermingled with smaller ones. Cusp clearly delineated from the adjacent first lateral denticles. Lateral teeth (Fig. 13H, J) narrowly triangular with 19-24 denticles on internal edge. Reproductive system (Fig. 13K, L). Diaulic. Hermaphroditic duct leads to long, rela- tively narrow convoluted ampulla of several whorls. Vas deferens long, without distinct prostate. Penial sheath elongated. Penis folded. Oviduct connects through insemination duct into female gland complex. No distal or proximal receptaculum seminis detected. Ecology. Soft bottom with stones 18 to 23 m. Distribution. Franz Josef Land. Remarks. According to the molecular phylogenetic analysis Ziminella circapolaris sp. n. forms a separate sister clade to Z. salmonacea (Fig. 1). Specimens from the West Polyphyly of the traditional family Flabellinidae affects a major group of Nudibranchia... 23 Atlantic North American coast (type locality of Z. salmonacea) and Spitzbergen belong to Z. salmonacea, according to molecular data, whereas specimens from Franz Josef Land belong to the separate species Z. circapolaris sp. n. This is concordant with a considerable level of endemism of the nudibranch fauna of Franz Josef Land recently demonstrated for other nudibranch groups (Martynov and Korshunova 2017). Mor- phological analysis reveals differences in the denticulation of the lateral teeth between Z. salmonacea and Z. circapolaris sp. n.: in the first species denticulation on the lateral teeth tends to be restricted to the first half of the length of lateral teeth or the teeth can be completely smooth, whereas in the new species denticulation runs up to the very end of the lateral teeth, and the teeth are always denticulated. Kuzirian (1979) studied the radulae of 15 specimens of Z. salmonacea from near the type locality and all of them have considerably smoother lateral teeth than Z. circapolaris sp. n. Family Flabellinopsidae fam. n. http://zoobank.org/F3E7E3B7-E77F-484A-B629-FFDDA9FDCD59 Diagnosis. Body relatively wide. Notal edge discontinuous. Cerata in separate clusters on broad extensions. Rhinophores perfoliated or granulated. Anus pleuroproctic under reduced notal edge. No distinct oral glands. Radula formula 1.1.1. Rachidian teeth usually with cusp compressed by adjacent lateral denticles. Lateral teeth narrow or with attenuated process basally, denticulated or smooth. Single distal receptaculum seminis. Vas deferens long, with or without distinct prostate. External permanent penial collar absent. Penis elongated conical, internal unarmed. Genera included. Baenopsis gen. n., Flabellinopsis MacFarland, 1966. Remarks. One of the unexpected results of the present molecular analysis is the most basal position of the species Flabellina iodinea (Cooper, 1863) in relation to the families Paracoryphellidae, Coryphellidae, and Flabellinidae (Figs 1, 2). This is prima facie not very consistent with the morphological data since F iodinea is similar to the members of the family Flabellinidae s. stv. as it possesses a medium-sized body with cerata on lateral extensions. However, this may be a case of morphological con- vergence since F iodinea's relatively broad, flap-like lateral modifications of the notal edge and absence of distinct oral glands is different from the majority of the Flabel- linidae s. str. Furthermore, the vas deferens in F iodinea is considerably longer than in the taxa of the Flabellinidae s. str. On all trees obtained, Flabellina iodinea invari- ably appears most basally and separate from all other families. Therefore, in this case our results speak for preference of the molecular data in revealing the very separate position of this particular taxon. It is most likely that the Flabellina-like external ap- pearance has evolved in F iodinea independently from Flabellinidae s. str. by parallel modifications of a continuous ancestral notal edge into flap-like ceratal structures. We therefore resurrect the genus Flabellinopsis MacFarland, 1966, which was previously proposed for the species Aeolis (Phidiana?) iodinea Cooper, 1863. On the present mo- lecular tree, the NE Pacific Flabellinopsis iodinea clustered together with a NE Atlantic 24 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) species, “Flabellina” baetica Garcia-Gomez, 1984, in the same clade (Figs 1, 2, 7). The NE Atlantic species has ceratal and reproductive morphology somewhat similar to the genus Flabellinopsis but possesses very peculiar folded and granulated rhinophores and smooth lateral teeth (Garcia-Gomez 1984). Therefore, rather than uniting it into the genus Flabellinopsis making that morphologically not very consistent, we propose a separate new genus for Flabellina baetica (see below). The diversity of the family Fla- bellinopsidae fam. n. is more considerable than currently understood. On our tree, an undetermined species of “Piseinotecus” sp. appeared as possibly related to this family (Fig. 1). Furthermore, Furfaro et al. (2017) have shown that “Piseinotecus” soussi Tam- souri, Carmona, Moukrim and Cervera, 2014 (original description in Tamsouri et al. 2014) has appeared basally to “Flabellina” baetica and Facelina quatrefagesi (Vayssi€re, 1888). Further investigations need to be made on these taxa. Baenopsis gen. n. http://zoobank.org/7 COE306D-0BC4-4286-814B-983D01598489 Figs:2, ' ' a. 4 Figure 29. [taxia falklandica (Eliot, 1907), comb. n. South Pacific, Chile, Canal Artilleria, ZSM Mol- 20070592, specimen 12 mm length (fixed): A dorsal view (live) B ventral view (live) C lateral view (live) D details of cerata (live) E jaws, light microscopy F details of masticatory process of jaw, light microscopy G radular teeth, posterior part, SEM H reproductive system (non mature), light microscopy. Abbreviations: fgm female gland mass nc continuous notal edge pvd prostatic vas deferens. Scale bars: G = 30 um. Photos of living specimens by M. Schrédl, other photos and SEM images by A.V. Martynov. 106 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) Aes r Figure 30. Microchlamylla gracilis gracilis (Alder & Hancock, 1844), comb. n. Norwegian Sea, Gulen Dive Center. ZMMU Op-502, living specimen 23 mm length: A dorsal view B lateral view C ventral view D details of cerata E dissected anterior part (pharynx removed) and rhinophores, SEM F jaw, SEM G details of masticatory process of jaw, SEM H radular teeth, posterior part, SEM I reproductive system, SEM. Abbreviations: a anus am ampulla fgm female gland mass go genital opening nd discontinuous notal edge og oral glands ps penial sheath pvd prostatic vas deferens r rhinophores rsd distal recep- taculum seminis. Scale bars: E, | = 1 mm; F = 300 um; G, H = 30 um. Photos and SEM images by TA. Korshunova, A.V. Martynov. Polyphyly of the traditional family Flabellinidae affects a major group of Nudibranchia... 107 Figure 31. Microchlamylla gracilis zfi subsp. n. Arctic, Franz Josef Land. ZMMU Op-501, specimen 11 mm length (fixed): A dorsal view (live) B lateral view (live) C ventral view (fixed) D details of cerata E egg masses on hydroids F jaws, SEM G radular teeth, posterior part, SEM H reproductive system. Ab- breviations: a anus am ampulla fgm female gland mass nd discontinuous notal edge pvd prostatic vas deferens rsd distal receptaculum seminis. Scale bars: F = 300 um; G = 30 um; H = 1 mm. Photos of living specimens and eggs by O.V. Savinkin, other photos and SEM images by A.V. Martynov 108 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) Figure 32. Occidentella athadona (Bergh, 1875), comb. n. North West Pacific, Kamchatka. ZMMU Op- 498, living animal 19 mm length: A dorsal view B lateral view (fixed) C ventral view D details of cerata E jaw, SEM, scale bar = 1 mm F details of masticatory process of jaw, SEM G details of masticatory pro- cess of jaw, light microscopy H radular teeth, posterior part, SEM I reproductive system, light microscopy J reproductive system, SEM. Abbreviations: a anus agf additional glandular formation (modified pros- tate), am ampulla fgm female gland mass go genital opening ps penial sheath rsd distal receptaculum seminis rsp proximal receptaculum seminis vd vas deferens (very rudimentary, rapidly transits to greatly enlarged modified prostate (agf). Scale bars: E = 1 mm; F,H = 30 um; J = 300 um. Photos and SEM im- ages by T.A. Korshunova, A.V. Martynov. Polyphyly of the traditional family Flabellinidae affects a major group of Nudibranchia... 109 Figure 33. Calmella cavolini (Vérany, 1846). Mediterranean Sea, Banyuls-sur-Mer. ZMMU Op-485, liv- ing specimen 12 mm in length: A dorsal view B ventral view C details of cerata D dissected anterior part (pharynx removed) and rhinophores E jaw, SEM F details of masticatory process of jaw, SEM G radular teeth, posterior part, SEM H I reproductive system, light microscopy. Abbreviations: a anus aM ampulla cs ceratal stalks fgm female gland mass go genital opening ogp oral gland penetrating into basis of cerata r rhinophores ps penial sheath pvd prostatic vas deferens. Scale bars: D, E = 300 pm; F, G = 10 um. Photos and SEM images by T.A. Korshunova, A.V. Martynoy. 110 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) PARAFLABELLINA Oy CALMELLA CY CARRONELLA FLABELLINA PY CORYPHELLINA EDMUNDSELLA Figure 34. Schematic outline of the reproductive systems of the taxa of the family Flabellinidae inte- grated with molecular phylogenetic data. Colour indication of reproductive system characters: ampulla — green; body wall — gray; distal receptaculum seminis — red; female gland mass — yellow; female genital opening — orange; penis and male genital opening — dark blue; penial sheath — pale blue; prostatic vas deferens — turquoise; proximal receptaculum seminis — pink. Polyphyly of the traditional family Flabellinidae affects a major group of Nudibranchia... 111 Figure 35. Carronella enne sp. n. NE Atlantic, off Ireland. ZMMU Op-526, specimen 5 mm (fixed) in length: A dorsal view (live); B details of cerata C rhinophore D dorsal view (fixed) E ventro-lateral (fixed) F jaw, SEM G details of masticatory process of jaw, SEM H details of masticatory process of jaw, light microscopy I radular teeth, posterior part, SEM J reproductive system, light microscopy. Abbreviations: a anus am ampulla ce ceratal elevations fgm female gland mass g gonad go genital opening nd dis- continuous notal edge ps penial sheath pvd prostatic vas deferens r rhinophores rsd distal receptaculum seminis rSp proximal receptaculum seminis. Scale bars: F = 300 um; G, I = 30 um. Photos of living specimens by E. Schwabe, other photos and SEM images by A.V. Martynov. 2 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) Figure 36. Carronella pellucida (Alder & Hancock, 1843), comb. n. Norwegian Sea, Gulen Dive Center. ZMMU Op-513, living specimen 18.5 mm in length: A dorsal view B lateral view C ventral view D details of cerata E dissected anterior part (pharynx removed) and rhinophores, SEM F jaw SEM G details of masti- catory process of jaw, SEM H radular teeth, posterior part, SEM I reproductive system, SEM J reproductive system, light microscopy K dissected penial sheath and penis, SEM. Abbreviations: a anus am ampulla ce ceratal elevations fgm female gland mass go genital opening nd discontinuous notal edge ogp oral gland penetrating into basis of cerata r rhinophores p penis ps penial sheath pvd prostatic vas deferens rsd distal receptaculum seminis rsp proximal receptaculum seminis. Scale bars: E = 1 mm; F, K = 300 um; G = 30 um; H = 30 um; I = 1 mm. Photos and SEM images by T.A. Korshunova, A.V. Martynov. Polyphyly of the traditional family Flabellinidae affects a major group of Nudibranchia... 113 Figure 37. Coryphellina exoptata (Gosliner & Willan, 1991), comb. n. Vietnam, Nhatrang Bay. ZMMU Op-116, living specimen 9 mm in length: A dorsal view B ventral view (fixed) C details of cerata D dis- sected anterior part (pharynx removed) and rhinophores, SEM E jaw, SEM F details of masticatory process of jaw, SEM, scale bar = 30 pm G radular teeth, posterior part, SEM H, I reproductive system, light mi- croscopy. Abbreviations: a anus am ampulla fgm female gland mass go genital opening nd discontinuous notal edge ogp oral gland penetrating into basis of cerata ps penial sheath r rhinophores rsp proximal receptaculum seminis pvd prostatic vas deferens. Scale bars: D = 1 mm; E = 300 um; F = 30 um. Photos of living specimens by O.V. Savinkin, other photos and SEM images by A.V. Martynov. 114 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) Figure 38. Coryphellina lotos sp. n. Japan, Pacific Honshu, Osezaki. ZMMU Op-515, living specimen 15.5 mm in length: A dorsal view B lateral view C ventral view D details of cerata E rhinophores, close up F jaw SEM G details of masticatory process of jaw, SEM H details of masticatory process of jaw, light microscopy I radular teeth, posterior part, SEM J reproductive system, light microscopy. Abbreviations: a anus am ampulla fgm female gland mass go genital opening nd discontinuous notal edge pvd pros- tatic vas deferens r rhinophores rsp proximal receptaculum seminis. Scale bars: F = 100 um; G = 20 um; I =50 um. Photos and SEM images by T.A. Korshunova, A.V. Martynov. Polyphyly of the traditional family Flabellinidae affects a major group of Nudibranchia... 115 Figure 39. Coryphellina rubrolineata O'Donoghue, 1929. Vietnam, Nhatrang Bay. ZMMU Op-132, living specimen 15 mm in length: A dorsal view B ventral view (fixed) C details of cerata D dissected anterior part (pharynx removed) and rhinophores, SEM, scale bar 1 mm E rhinophores, close up, SEM F jaw, SEM G details of masticatory process of jaw, SEM H radular teeth, posterior part, SEM I, J re- productive system, light microscopy K bilobed proximal receptaculum seminis. Abbreviations: a anus am ampulla fgm female gland mass go genital opening nd discontinuous notal edge ogp oral gland penetrating into basis of cerata ps penial sheath rsp proximal receptaculum seminis pvd prostatic vas deferens r rhinophores. Scale bars: D, F = 300 um; G, H = 30 um. Photos of living specimens by O.V. Savinkin, other photos and SEM images by A.V. Martynov. 116 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) - Figure 40. Edmundsella pedata (Montagu, 1815), comb. n. Norwegian Sea, Gulen Dive Center. ZMMU Op-540, living specimen 28 mm in length: A dorsal view B lateral view C ventral view D details of cerata E jaw, SEM F details of masticatory process of jaw, SEM G radular teeth, posterior part, SEM H reproduc- tive system SEM I reproductive system, light microscopy J penial sheath dissected. Abbreviations: a anus am ampulla ce ceratal elevations fgm female gland mass go genital opening nd discontinuous notal edge Ps penial sheath pvd prostatic vas deferens rsd distal receptaculum seminis rsp proximal receptaculum seminis. Scale bars: E, H, J = 100 um; F = 20 um; G = 50 um. Photos and SEM images by T:A. Korshu- nova, A.V. Martynov. Polyphyly of the traditional family Flabellinidae affects a major group of Nudibranchia... 117 Figure 41. Pacifia amica sp. n., NE Pacific, Port Orchard, holotype ZMMU Op-614: A live, dorsal view, 6.5 mm in length B lateral view C details of cerata D jaws, lateral view E details of masticatory processes of jaw F posterior rows of radula, SEM G anterior rows of radula H details of posterior radular teeth I reproduc- tive system in situ, light microscopy. Abbreviations: am ampulla fgm female gland mass go genital opening pr prostate ps penial sheath rsd distal receptaculum seminis rsp proximal receptaculum seminis vd vas deferens. Scale bars: D = 100 pm E =2 um; F, G, H = 10 um. Photos of living specimens by K. Fletcher, other photos and SEM images by A.V. Martynov. 118 Tatiana Korshunova et al. / ZooKeys 717: 1-139 (2017) —