A molecular phylogeny of the Indo-West Pacific species of Haloa sensu lato gastropods (Cephalaspidea_ Haminoeidae)_ Tethyan vicariance, generic diversity, and ecological specialization

https://doi.org/10.1016/j.ympev.2019.106557 Received 19 October 2018; Received in revised form 23 March 2019; Accepted 5 July 2019 ⁎ Corresponding author. E-mail addresses: Trond.Oskars@uib.no, trondoskars@gmail.com (T.R. Oskars). Molecular Phylogenetics and Evolution 139 (2019) 106557 Available online 06 July 2019 1055-7903/ © 2019 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/BY/4.0/). T The genus Liloa has been regarded as a valid genus by most authors (Higo et al., 1999, 2001; Too et al., 2014; Gosliner et al., 2015) and its monophyly was suggested by Too et al. (2014) based on the shared presence of unique traits of the gizzard plates and male reproductive system, and was confirmed by Oskars et al. (2019) using molecular phylogenetics. On the other hand, the genus Haloa has largely been regarded as a synonym of Haminoea (Rudman, 1971a; Burn and Thompson, 1998) and only few authors have referred to it as valid after its original introduction (Habe, 1952; Higo et al., 1999, 2001; Hori, 2001). Several genera were throughout the first half of the 20th century introduced for IWP “Haminoea”. Iredale (1929) erected the genus Penthominea for the Australian species Bulla wallisii Gray, 1824 and Kuroda and Habe (1952) erected the genera Lamprohaminoea for Bulla cymbalum Quoy & Gaimard, 1832, Parahaminoea for Haminaea binotata Pilsbry, 1895a, and Vitreohaminoea for Bulla vitrea A. Adams, 1850. However, Kuroda and Habe (1952) did not include any generic descriptions or reference to other works, which rendered their generic names invalid according to ICZN (1999) art.13 (see Oskars et al., 2019). Habe (1952) raised Haloa to genus level based on the presence of a secondary cusp on the inner edge of the lateral teeth and synonymized Penthominea and Parahaminoea with Haloa without including any justification. However, because Habe (1952) did not find any additional cusp in H. vitrea he maintained Vitreohaminoea for the latter species, yet as a subgenus of Haloa. Habe (1952) additionally introduced the subgenus Sericohaminoea for the new species Haloa (Sericohaminoea) yamagutii Habe, 1952 based on the presence of conspicuous spiral striae and a thicker columellar callus. Rudman (1971a) after studying specimens of H. crocata from Hawaii (type species and type locality of Haloa) could not find a secondary cusp in the inner lateral teeth and consequently saw no reason to separate Haloa or the genera introduced by Habe (1952) and Kuroda and Habe (1952) from Haminoea. Haloa (sensu Oskars et al., 2019) is the most diverse genus of the family Haminoeidae with an estimated 22 species (MolluscaBase, 2018c) nearly all inhabiting tropical shallow waters of the IWP with few lineages extending into temperate latitudes, such as Haloa japonica (Pilsbry, 1895a) and H. maugeansis (Burn, 1966). The genus includes dull-coloured species associated with sandy-mud flats, algae, and seagrass meadows (e.g. Haloa japonica, H. natalensis (Krauss, 1848); Rudman, 1971a; Gosliner, 1987; Gibson and Chia, 1989; Miranda and Renzo-Perissinotto, 2012; Tibiriçá and Malaquias, 2017), colourful species some with flamboyant patterns typically inhabiting coral reefs (e.g. Haloa cymbalum (Quoy & Gaimard, 1832), H. ovalis (Pease, 1868), H. cyanomarginata (Heller and Thompson, 1983); Rudman, 1999a, 1999b, 1999c; Cruz-Rivera and Paul, 2006; Gosliner et al., 2008, 2015; Tibiriçá and Malaquias, 2017; Pittman and Fiene, 2018), and mangrove dweller species (e.g. H. fusca (Adams, 1850); Riek 2013, 2014; Gosliner et al., 2015; Cobb, 2018). Species of Haloa (sensu Oskars et al., 2019) have demonstrated potential for invading and establish viable populations in regions outside their native ranges. For example, Hanson et al. (2013a) found that Haloa japonica from the western Pacific has replaced the native Haminoea vesicula (Gould, 1855) in Boundary Bay, Canada. The same has been reported by Macali et al. (2013) for the Mediterranean Sea in the Laguna di Sabaudia, Italy where the native Haminoea species were outcompeted and replaced by the invasive H. japonica. The Red Sea endemic colourful species Haloa cyanomarginata is regarded a Lessepsian immigrant and has now spread all over the Mediterranean Sea (Koehler, 2003; Yokes, 2003; Rudman, 2003a; Zenetos et al., 2004, 2008, 2010; Mifsud, 2007; Crocetta and Vazzana, 2009; FernándezVilert et al., 2018). Its impact in local faunas is not yet clear, but according to Mollo et al. (2008) the presence of brominated secondary metabolites in H. cyanomarginata is an effective feeding deterrent for native predators. Additionally, the invasive H. japonica has been identified in California to be a potential human hazard as vector of Schistosoma parasites that can cause cercarial dermatitis or swimmer’s itch in humans (Brant et al., 2010; Hanson et al., 2013b). Several molecular phylogenetic studies had previously included representatives of Haminoea sensu lato and Smaragdinella (e.g. Wägele et al., 2003; Malaquias et al., 2009; Oskars et al., 2015). However, none of these studies produced a clear pattern of relationships between the Atl+ EP and IWP species and the genus Smaragdinella, most likely because of the reduced diversity of taxa included. Consequently, the monophyly of the genus Haminoea was never questioned. It is only with the work by Oskars et al. (2019) that the phylogeny of the Haminoeidae is comprehensively studied and became clear that Haminoea as traditionally defined is not a natural group but instead encompasses three independent monophyletic lineages (Haloa, Haminoea, Smaragdinella) all supported by several morphological synapomorphies (Oskars et al., 2019). An interesting finding suggested by the latter authors was the possible occurrence of three clades in Haloa each of them with unique morphological and ecological features, which the authors have informally named Haloa clade 1 with dull-coloured species, Haloa clade 2 only with colourful species, and Haloa clade 3 containing mangroveassociated lineages. In the present contribution, we further explore the systematics and taxonomic composition of the IWP Haloa based on an expanded taxon set with specimens from all previously recognized subclades and biogeographic regions. An integrative approach combining molecular phylogenetics (based on a combination of mitochondrial and nuclear genes) together with detailed conchological and morpho-anatomical characters is used to define clades, establish relationships, and infer putative synapomorphies. 2. Methods 2.1. Sampling Novel material used in the present study was collected by the authors or obtained from the collections of the Department of Natural History, University Museum of Bergen, Norway (ZMBN), and loans from the Museum National d’Histoire Naturelle, Paris (MNHN), The Natural History Museum, London (NHMUK), The California Academy of Science (CAS), The Australian Museum, Sydney (AMS), Museum Victoria, Australia (MV), The Auckland Museum, New Zealand (AM), The Natural History Museum of Florida (UF), the Bavarian Collections of Zoology (ZSM), and the Museum of Zoology, University of Michigan, USA (UMMZ). Outgroup taxa consisting of species from six haminoeid genera, namely Atys, Aliculastrum, Bullacta, Diniatys, Liloa, Phanerophthalmus, were included in the analyses together with representatives of mini haminoeids (Carlson et al., 1998; Oskars et al., 2019). The tree was rooted with the cephalaspidean species Bulla cf. peasiana Pilsbry, 1895a. In total, this study includes 237 specimens (137 IWP Haloa, 31 Atlantic/EP Haminoea, 8 Smaragdinella, 60 other Haminoeidae and 1 Bullidae) a total of 858 sequences were gathered for analyses from 222 specimens of 84 species (Table 1). 2.2. DNA extraction, amplification, and sequencing DNA was extracted from tissue obtained from the foot or parapodial lobes using the Qiagen DNeasy® Blood and Tissue Kit (catalogue no. 69504) following the protocol recommended by the manufacturer. DNA from alcohol fixed samples more than 100 years old of “Haloa” fusca (A. Adams, 1850) from the collections of the MNHN, Paris (spcs TH42, TH83, TH84; see Table 1) collected by C. Semper in 1875 in the Philippines, was extracted with the E.Z.N.A.® Mollusc DNA Kit (Omega Bio-tek, D3373-01). The manufacturer’s protocol was followed with minor modifications. The tissue was lysed by incubating overnight at 37 °C and was then homogenized in a Qiagen TissueLyzer at 15 Hz for 25 s. The lysate was then incubated at 60 °C for 1 h, before continuing T.R. Oskars and M.A.E. Malaquias Molecular Phylogenetics and Evolution 139 (2019) 106557


Introduction
The family Haminoeidae (115 species; MolluscaBase, 2018a) is the most diverse radiation of the gastropod Order Cephalaspidea (779 species; MolluscaBase, 2018b), with species present worldwide mainly inhabiting soft bottoms and algae or seagrass mats in shallow waters of temperate and tropical latitudes. Recently Oskars et al. (2019) found Haminoea Turton and Kingston (1830), the type genus of the family, to be polyphyletic. Haminoea sensu lato was formed by three distinct clades, namely a clade containing all Indo-West Pacific (IWP) species, sister to the intertidal rock-dweller haminoeid genus Smaragdinella A. Adams (1848), and these two clades sister to a radiation including all Atlantic (Atl) and eastern Pacific (EP) species of Haminoea plus one single species from South Africa living in temperate waters of the Indian Ocean. In order to reflect the phylogeny, the authors have reinstated the genus Haloa Pilsbry (1921) for the IWP species and maintained the Atl + EP species in the genus Haminoea sensu stricto (type species Haminoea hydatis Linnaeus (1758) from the Mediterranean Sea) (Oskars et al., 2019).
The subdivision of Haminoea was earlier suggested by Pilsbry (1921) based on the shape of the columella of the shell. The author introduced the subgenera Haminaea proper (type species Bulla hydatis), Haloa (type species Haminea crocata Pease, 1860 from Hawaii), and Liloa (type species Haminea curta tomaculum Pilsbry, 1917 from Hawaii). According to Pilsbry (1921) both Haminoea and Haloa had a concave columella but differed by the fact that in Haminoea the columella is tightly fused to the last whorl of the shell, whereas in Haloa it is separated from the shell whorl by a furrow. Liloa was distinguished by having only slightly concave columella and a narrower shell with spiral striae throughout the shell.

T
The genus Liloa has been regarded as a valid genus by most authors (Higo et al., 1999(Higo et al., , 2001Too et al., 2014;Gosliner et al., 2015) and its monophyly was suggested by Too et al. (2014) based on the shared presence of unique traits of the gizzard plates and male reproductive system, and was confirmed by Oskars et al. (2019) using molecular phylogenetics.
On the other hand, the genus Haloa has largely been regarded as a synonym of Haminoea (Rudman, 1971a;Burn and Thompson, 1998) and only few authors have referred to it as valid after its original introduction (Habe, 1952;Higo et al., 1999Higo et al., , 2001Hori, 2001).
Several genera were throughout the first half of the 20th century introduced for IWP "Haminoea". Iredale (1929) erected the genus Penthominea for the Australian species Bulla wallisii Gray, 1824 andKuroda andHabe (1952) erected the genera Lamprohaminoea for Bulla cymbalum Quoy & Gaimard, 1832, Parahaminoea for Haminaea binotata Pilsbry, 1895a, and Vitreohaminoea for Bulla vitrea A. Adams, 1850. However, Kuroda and Habe (1952) did not include any generic descriptions or reference to other works, which rendered their generic names invalid according to ICZN (1999) art.13 (see Oskars et al., 2019). Habe (1952) raised Haloa to genus level based on the presence of a secondary cusp on the inner edge of the lateral teeth and synonymized Penthominea and Parahaminoea with Haloa without including any justification. However, because Habe (1952) did not find any additional cusp in H. vitrea he maintained Vitreohaminoea for the latter species, yet as a subgenus of Haloa. Habe (1952) additionally introduced the subgenus Sericohaminoea for the new species Haloa (Sericohaminoea) yamagutii Habe, 1952 based on the presence of conspicuous spiral striae and a thicker columellar callus. Rudman (1971a) after studying specimens of H. crocata from Hawaii (type species and type locality of Haloa) could not find a secondary cusp in the inner lateral teeth and consequently saw no reason to separate Haloa or the genera introduced by Habe (1952) and Kuroda and Habe (1952) from Haminoea.
Species of Haloa (sensu Oskars et al., 2019) have demonstrated potential for invading and establish viable populations in regions outside their native ranges. For example, Hanson et al. (2013a) found that Haloa japonica from the western Pacific has replaced the native Haminoea vesicula (Gould, 1855) in Boundary Bay, Canada. The same has been reported by Macali et al. (2013) for the Mediterranean Sea in the Laguna di Sabaudia, Italy where the native Haminoea species were outcompeted and replaced by the invasive H. japonica. The Red Sea endemic colourful species Haloa cyanomarginata is regarded a Lessepsian immigrant and has now spread all over the Mediterranean Sea (Koehler, 2003;Yokes, 2003;Rudman, 2003a;Zenetos et al., 2004Zenetos et al., , 2008Zenetos et al., , 2010Mifsud, 2007;Crocetta and Vazzana, 2009;Fernández-Vilert et al., 2018). Its impact in local faunas is not yet clear, but according to Mollo et al. (2008) the presence of brominated secondary metabolites in H. cyanomarginata is an effective feeding deterrent for native predators. Additionally, the invasive H. japonica has been identified in California to be a potential human hazard as vector of Schistosoma parasites that can cause cercarial dermatitis or swimmer's itch in humans (Brant et al., 2010;Hanson et al., 2013b).
Several molecular phylogenetic studies had previously included representatives of Haminoea sensu lato and Smaragdinella (e.g. Wägele et al., 2003;Oskars et al., 2015). However, none of these studies produced a clear pattern of relationships between the Atl + EP and IWP species and the genus Smaragdinella, most likely because of the reduced diversity of taxa included. Consequently, the monophyly of the genus Haminoea was never questioned. It is only with the work by Oskars et al. (2019) that the phylogeny of the Haminoeidae is comprehensively studied and became clear that Haminoea as traditionally defined is not a natural group but instead encompasses three independent monophyletic lineages (Haloa, Haminoea, Smaragdinella) all supported by several morphological synapomorphies (Oskars et al., 2019).
An interesting finding suggested by the latter authors was the possible occurrence of three clades in Haloa each of them with unique morphological and ecological features, which the authors have informally named Haloa clade 1 with dull-coloured species, Haloa clade 2 only with colourful species, and Haloa clade 3 containing mangroveassociated lineages.
In the present contribution, we further explore the systematics and taxonomic composition of the IWP Haloa based on an expanded taxon set with specimens from all previously recognized subclades and biogeographic regions. An integrative approach combining molecular phylogenetics (based on a combination of mitochondrial and nuclear genes) together with detailed conchological and morpho-anatomical characters is used to define clades, establish relationships, and infer putative synapomorphies.

Sampling
Novel material used in the present study was collected by the authors or obtained from the collections of the Department of Natural History, University Museum of Bergen, Norway (ZMBN), and loans from the Museum National d'Histoire Naturelle, Paris ( Outgroup taxa consisting of species from six haminoeid genera, namely Atys, Aliculastrum, Bullacta, Diniatys, Liloa, Phanerophthalmus, were included in the analyses together with representatives of mini haminoeids (Carlson et al., 1998;Oskars et al., 2019). The tree was rooted with the cephalaspidean species Bulla cf. peasiana Pilsbry, 1895a. In total, this study includes 237 specimens (137 IWP Haloa, 31 Atlantic/EP Haminoea, 8 Smaragdinella, 60 other Haminoeidae and 1 Bullidae) a total of 858 sequences were gathered for analyses from 222 specimens of 84 species (Table 1).

DNA extraction, amplification, and sequencing
DNA was extracted from tissue obtained from the foot or parapodial lobes using the Qiagen DNeasy® Blood and Tissue Kit (catalogue no. 69504) following the protocol recommended by the manufacturer. DNA from alcohol fixed samples more than 100 years old of "Haloa" fusca (A. Adams, 1850) from the collections of the MNHN, Paris (spcs TH42, TH83, TH84; see Table 1) collected by C. Semper in 1875 in the Philippines, was extracted with the E.Z.N.A.® Mollusc DNA Kit (Omega Bio-tek, D3373-01). The manufacturer's protocol was followed with minor modifications. The tissue was lysed by incubating overnight at 37°C and was then homogenized in a Qiagen TissueLyzer at 15 Hz for 25 s. The lysate was then incubated at 60°C for 1 h, before continuing Table 1 List of specimens used for phylogenetic analyses, with sampling localities, voucher numbers, and GenBank accession numbers (numbers marked with asterisk * are novel sequences generated for this study). Taxa highlighted in bold are the type species of genera
Polymerase chain reactions (PCR) for the COI and 28S genes followed the protocols described by , whereas for the 16S and H3 genes followed the protocols described by Oskars et al. (2015). PCR reactions for the 12S gene were performed in 50 μL volume, including 19.5 μL Sigma water, 5 μL CoraLLoad buffer, 5 μL dNTP, 10 μL Q-solution, 5 μL MgCl, 2 μL of each of the primers, 0.5 μL Taq, and 1 μL DNA. PCR cycling for 12S rRNA consisted of 40 cycles with initial denaturation at 95°C for 3 min, denaturation at 94°C for 45 s, annealing phase with an optimal annealing temperature of 49.4°C for 45 s, ramp to 72°C for 1°C/sec. and extension at 72°C for 2 min, followed by a final extension at 72°C for 10 min. For samples that did not amplify with Qiagen Taq, additional 25 μL reactions were set with TaKaRa Ex Taq Polymerase HS (250 U) (Cat. number: RR006A), following the protocol described by Oskars et al. (2015). The cycling was the same as above, but a hot start step of 94°C for 5 min was included at the beginning of the cycle. For samples extracted with E.Z.N.A. Mollusc DNA Kit the same protocols were followed, but 2-2.5 µL DNA were used instead.
The quality and quantity of PCR products were assessed by gelelectrophoresis following standard methods (see Eilertsen and Malaquias, 2013). For PCR products that yielded weak bands or double bands the products were reamplified by Gel Stabbing (Rees, D. pers. com.) where a sample of gel with the desired product was extracted with a cut pipette tip, and re-run under the original PCR protocol.
Successful PCR products were purified according to the EXO-SAP method described in Eilertsen and Malaquias (2013). Sequence reactions were run on an ABI 3730XL DNA Analyser (Applied Biosystems).

Morpho-anatomical methods
Animals were carefully separated from the shells with the aid of forceps. The male reproductive system, buccal bulb, and gizzard were dissected out of the animals. Shells were imaged with a DSLR camera equipped with macrolens. The reproductive systems were drawn using a stereo microscope fitted with a drawing tube. The anterior digestive system was dissected and the gizzard plates and gizzard bristles were extracted. The gizzard plates, radulae, and jaws were cleaned using a solution containing 180 µL buffer ATL and 20 µL of proteinase K (obtained from the Qiagen DNeasy® Blood and Tissue Kit catalogue no. 69504) incubated at 56°C for approximately 4-6 h (protocol modified from Holznagel (1998) and Vogler (2013)). For formalin fixed specimens, gizzard plates, radulae, and jaws were immersed in a 10-30% solution of sodium hydroxide (NaOH) until free of tissue. The gizzard plates, gizzard bristles and jaws were critical-point dried after cleaning to maintain their natural shape, and mounted together with the radulae on metallic stubs using carbon sticky tabs and coated with gold-palladium. The samples were scanned and imaged with a Zeiss Supra 55VP scanning electron microscope.

Phylogenetic analyses
Sequencher (v. 4.10.1, Gene Codes Corp.) and Geneious (v. 8.1.8 Biomatters Ltd., Kearse et al., 2012) were used to inspect, edit, and assemble the chromatograms of the forward and reverse DNA strands. All sequences were blasted in GenBank to check for contamination. Single gene sequences were aligned with Muscle (Edgar 2004a(Edgar , 2004b implemented in Geneious. Alignments were trimmed to a position where at least 50% of the sequences had nucleotides and missing positions at the ends were coded as missing data (?). All sequences were deposited in GenBank (Table 1).
Blocks of ambiguous data in the single gene alignments were identified and excluded using Gblocks with stringent and relaxed settings (Talavera and Castresana, 2007;Kück et al., 2010) (Table S1). The results obtained with the relaxed settings for the 16S and 12S genes showed no difference from the un-masked datasets, thus the latter were used for analyses. Saturation was tested for the first, second, and third codon positions of the protein coding genes COI and H3 using MEGA7 (Kumar et al. 2016) by plotting general time-reversible (GTR) pairwise distances against total substitutions (transitions + transversions). The JModeltest software (Darriba et al., 2012) was used to find the best-fit model of evolution for each single gene dataset under the Akaike information criterion (Akaike, 1974) (Table S2).
Eleven individual gene analyses were initially preformed: COI ( Concatenations were based on sequences yielded from the same specimen with the single exception of the samples of "Haloa" fusca from the Philippines collected by C. Semper in 1875 because only spc. TH42 yielded a COI sequence and spcs TH83 and TH84 were the only ones to yield H3 sequences. Yet, sequenced specimens came from the same lot (i.e. collected together at the same spot, same time and their morphoanatomical study confirmed conspecificity. Bayesian inference analyses (BI) using MrBayes Ronquist and Huelsenbeck, 2003) were run on the initial single gene datasets (Figs S1-S11) and all-genes concatenated dataset ( Fig. S12; 2844 bp). The analyses used three parallel runs of 4 million generations for the single gene analyses and 8 million generations for the concatenated dataset, with sampling every 100 generations. The concatenated dataset was partitioned by gene and each partition was run under the best-fit model of evolution (Table S2). Convergence of runs was inspected in Tracer v1.7 (Rambaut et al., 2018) with a burn-in set to 25% by comparing the likelihood of trees drawn by the independent runs. Posterior probabilities (PP) higher than 0.95 were considered statistically significant Alfaro et al., 2003). A Maximum Likelihood analysis (ML) of the concatenated dataset ( Fig. S13; 2844 Kearse et al., 2012). The analysis was partitioned by gene and run under the "rapid bootstrapping and search for best scoring ML tree" algorithm, using a random starting tree and the model GTR+G+I with 1000 bootstrap (BS) replicates. Bootstrap values higher than 75% were considered significantly supported (Felsenstein, 1985), whereas BS values above 70% were considered as nearly supported (Hillis and Bull, 1993). Consensus phylograms were generated in MrBayes and Geneious, and annotated and converted to graphics in FigTree v1.3.1 (Rambaut and Drummond, 2009). All figures were made in Inkscape 0.48.4 (Inkscape Team, 2013) and Gimp 2.10 (Mattis et al., 1995;Natterer et al., 2018).
The mitochondrial genes COI, 12S, and 16S showed good performance in consistently clustering genera and species, but the latter two genes failed in some cases to separate what seem to be recently diverged sister species (those with lower COI uncorrected p-distances between 6 and 8%). The nuclear 28S and H3 genes showed in general good resolution at genus level and were able to consistently cluster representatives of most species. However due to the lower phylogenetic signal of conserved nuclear genes, the analysis failed to tell apart some lineages (28S: Bakawan gen. nov.; Haloa aptei/H. cf. nigropunctatus and Haloa sp. 1 / Haloa sp. 2; H3: Lamprohaminoea cymbalum/L. cf. ovalis; Haloa sp. 1 / Haloa sp. 3). Tree resolution improved with concatenation and both BI and ML analyses resulted largely in the same topology (Figs. 1, 2, S12, S13).

Papawera gen. nov.
This new genus here described (see Discussion and Taxonomic section) is represented by two species, namely P. zelandiae (type species here designated) and P. maugeansis (PP = 1, BS = 100; Figs. 1 and 2) (see Taxonomic section and Discussion).

Smaragdinella A. Adams, 1848
Smaragdinella was highly supported (PP = 1, BS = 86; Figs. 1 and 2) and the phylogeny suggested two sub-clades, but this split could be an artefact of missing data since one of the species is mostly represented by specimens with only nuclear genes mined from GeneBank (GH5, GH6, GH7; Fig. 1, Table 1).
Habitat: Shallow waters on coral reefs, coral rubble, rocks covered in algae and algae beds (Cruz-Rivera and Paul, 2006;Gosliner et al., 2008Gosliner et al., , 2015Tibiriçá and Malaquias, 2017;Pittman and Fiene, 2018). Remarks: Lin (1997) is here regarded as the authority of Lamprohaminoea because she was the first to include a diagnosis for the genus and type species (see Oskars et al., 2019).
Etymology: The genus is named after the Tagalog Philippine name for mangroves, as three of the species are found in the Philippines.
Distribution: From the United Arab Emirates eastwards to Singapore and China, the Philippines, southwards along Queensland, Australia.
Etymology: The genus name is erected after Robert Burn renowned Australian malacologist. The name is composed by two parts: "papa" meaning to set fire in the Bunganditj or Boandik language of the indigenous people of Port MacDonnell, southern Australia (see Blake, 2003), and "wera" which means burn in the Maori language of the indigenous people of New Zealand.

Discussion
4.1. Tethyan vicariance, relictualism, and ecological specialization Oskars et al. (2019) showed the paraphyly of the genus Haminoea with all Atl + EP species (plus a temperate Indian Ocean species from South Africa) clustering together in a sister position to a clade containing the IWP and Australasia species of "Haminoea" (=Haloa sensu lato; Oskars et al., 2019) plus the genus Smaragdinella. The current study confirms this pattern and showed the occurrence of four distinct clades with unique morphological traits and ecologies among Haloa sensu lato.
The five sub-clades found within Haloa s. l. all have nearly distinct ecologies. Haloa sensu stricto includes dull-coloured species with cryptic colour patterns distributed across sub-tropical and tropical waters of the IWP. The species are seemingly generalist herbivores that feed on diatoms, dinoflagellates and green algae such as Ulva spp. (Usuki, 1966a(Usuki, , 1966bIto et al., 1996;Ito, 1997;Oskars and Malaquias, unpublished data).
Lamprohaminoea has a similar geographical span but only includes colourful species with flamboyant patterns associated with the presence of deterrent chemicals (Poiner et al., 1989;Fontana et al., 2001 for L. cymbalum;Mollo et al., 2008 for L. cyanomarginata), and species apparently feed on cyanobacteria (Cruz-Riviera and Paul, 2006). The development of aposematic patterns and diet specialization possibly played a role on the diversification of this clade. The genus Bakawan gen. nov. includes species restricted to mangrove habitats in the Indian ocean and western Pacific. The genus Papawera gen. nov. is restricted to temperate Australasian waters of New Zealand and southern Australia including Tasmania. On the other hand, the genus Smaragdinella is the only rock-dweller clade in the entire order Cephalaspidea.
This apparent ecological segregation suggests an important role of ecology, namely habitat selection and diet in shaping the phylogenetic structure of Haloa s. l. driving the diversification of its main lineages. The role of ecology in the formation of species, such as nutrient availability or productivity and temperature regimes, has been hypothesised to be a possible mechanism of diversification for example in cowries (Meyer, 2003), periwinkles (Williams and Reid, 2004), sponges (Duran and Rützler, 2006), and bubble-shells ) and we here stress its putative importance in also shaping the deep phylogenetic structure of major clades in Haminoeidae.
Another interesting aspect of the phylogeny is the occurrence of a putatively basal clade of southern Australasian species (=Papawera gen. nov.), which could represent a case of Tethyan relictualism. Nevertheless, caution is warranted, because some sister relationship between genera in our phylogeny, including Papawera gen. nov., are not fully resolved (Figs. 1 and 2). Genera that were once common in the European Eocene to Miocene and that now survive only in southern Australasia, have been interpreted as "living 'Tethyan relicts" (Houbrick, 1984a(Houbrick, , 1984bWilson and Allen, 1987;Hall, 1998;. However, until a calibrated phylogeny of Haloa s. l. and ideally better knowledge of the fossil record of these snails is available our hypothesis remains largely speculative.

Hidden diversity in Haloa sensu lato
Up to now, the taxonomy of the worldwide Haminoea s. l. snails was very much based on shells and to a lesser extent colour patterns, whereas anatomical features were only known for few species (e.g. Haminoea hydatis, H. navicula, Haloa japonica, Lamprohaminoea cymbalum, Papawera zelandiae). With few exceptions, the vast majority of species have similar shells that are difficult to tell apart and the colouration of the animals is complicated to use because of similar dull colour patterns. Moreover, whereas the anatomy of the Atlantic species is relatively well known (Malaquias and Cervera, 2006) few recent data is available for the IWP species (e.g. Er. Marcus and Burch, 1965;Rudman, 1971aRudman, , 1971bBharate et al., 2018).
Additionally, Oskars et al. (2019) pointed out to the possibility that Haloa s. l. could warrant further split because of the occurrence of three well supported sub-clades with unique features in the male reproductive system, parts of the digestive system, shells, and external colouration. The expanded dataset of IWP species of Haloa s. l. used in this work including all lineages known to us yielded a fourth sub-clade with only species of temperate affinities. Two of these clades have generic names available (Haloa and Lamprohaminoea) and two are here formally described (Bakawan gen. nov. and Papawera gen. nov.; see Results). Pilsbry (1921) Haloa was found to be monophyletic with 11 species; however, there are 16 nominal names available in the literature, but ongoing systematic work on this group revealed that several of those names are synonyms (Oskars and Malaquias, unpublished data).

The genus Haloa
All species in the genus are restricted to tropical and sub-tropical shallow waters of the IWP, except for H. japonica, which has a native temperate to sub-tropical distribution (Hanson et al., 2013a(Hanson et al., , 2013b. Most species of Haloa are externally similar and cryptically coloured and as a result several of them have been loosely assigned to the "common" and better known species H. natalensis (e.g. Gosliner, 1987, Table 2 Synopsis of the most useful morphological characters for diagnosis of genera.  Gosliner et al., 2008Gosliner et al., , 2015Tibiriçá and Malaquias, 2017;Johnson andJohnson, 2018a, 2018b;Michenet and Berberain, 2018;Pittman and Fiene, 2018). See theme 3.9. Taxonomic section and Table 2 for diagnostic features of Haloa.
Some species in this clade are difficult to tell apart as shells have similar shapes and colour patterns can depict both interspecific similarities and intraspecific variation. An ongoing systematic review of Lamprohaminoea species shows that apart of the DNA the most reliable characters to separate species are features of the reproductive system (Oskars and Malaquias unpublished data).
The origin of the colourful patterns in Lamprohaminoea has never been thoroughly studied, but seem to be a sign of distastefulness resulting from the presence of deterring secondary metabolites. It is however, unknown whether the chemical compounds are produced de novo or originate from their food (Poiner et al., 1989;Mollo et al., 2008). Even though the origin of these metabolites remains unknown, it seems that the acquisition of this trait, and the bright colours signalling their presence, was determinant in the evolution and radiation of Lamprohaminoea.

The genus Bakawan gen. nov.
All lineages ecologically restricted to mangrove habitats have clustered together and this included species like "Haloa" rotundata and "H." fusca plus two unidentified species that are presumably new. This clade is not only ecologically distinct, but species are also characterized by a unique morphology and colour pattern with animals possessing a uniformly pale to dark green or orange to reddish colouration in some cases with dark blotches (Riek, 2013(Riek, , 2014Mujino, 2016;Cobb, 2018;Yonow and Jensen, 2018). Therefore, we here introduce the new genus name Bakawan gen. nov. to reflect the phylogeny and the unique morphology and ecology of this clade. (see theme 3.9. Taxonomic section and Table 2 for diagnostic features of Bakawan gen. nov.).
All four species recognized in this clade are externally similar and only one species is often recognized in recent literature, often named as "Haminoea" fusca (e.g. Gosliner et al., 2008Gosliner et al., , 2015Atlas of Living Australia, 2018;Cobb, 2018). Species in this clade are restricted to mangrove habitats of the eastern Indian Ocean and western Pacific. Interestingly, there are no records of Bakawan gen. nov. species along the extensive mangrove systems of eastern Africa or in the central Pacific islands.

The genus Papawera gen. nov.
Australasian species from temperate waters were not included in the work by Oskars et al. (2019) and their inclusion in the current study revealed the occurrence of a fourth clade within Haloa s. l. to which no name was available. Therefore, we here introduce the new genus Papawera gen. nov., which besides its molecular distinctiveness is also characterized by unique morphological features, such as its large annulated prostate and nearly monocuspid rachidian, and a geographical range restricted to temperate waters of the northern parts of New Zealand, southern Australia and Tasmania (see theme 3.9. Taxonomic section and Table 2 for diagnostic features of Papawera gen. nov.).

Conclusions and a revised classification for Haminoea sensu lato
The results obtained with this study once again have clearly showed the importance of broad taxon sampling representing not only the traditional lineages, but also, ideally the entire morphological disparity and geographical span of the target group. In Cephalaspidea, this was previously demonstrated first with the work by  and recently even more emphatically with the works by Oskars et al. (2015Oskars et al. ( , 2019 and Bharate et al. (2018). In the current work it was of chief importance for example to include representatives of what we previously have identified as Haminoea natalensis, Haminoea cymbalum, Haminoea ovalis, and Haminoea fusca from across the entire distribution of these species. Otherwise, we would not have recognized cryptic diversity and even in some cases generic lineages.
Molecular phylogenetics including a complete or nearly complete species level diversity of Haminoea s. l. and Smaragdinella yielded a surprising result with on one hand a radiation including Haminoea s. s. with all Atlantic and eastern Pacific species plus a temperate lineage occurring in South Africa, and on the other a Indo-West Pacific plus Australasian radiation containing five genera, namely Haloa, Lamprohaminoea (here reinstated as valid), Smaragdinella, Bakawan gen. nov., and Papawera gen. nov. Additionally, these molecular groups are all supported by several morphological unique characters, particularly in the male reproductive system (see theme 3.9. Taxonomic section and Table 2).
Moreover, the phylogenetic pattern recovered strongly suggests an important role of Tethyan vicariance and ecological specialization in the diversification of these haminoid snails.