Carlgren’s hesitation allayed: redescription and systematics of Heteranthus verruculatus Klunzinger, 1877 (Cnidaria, Actiniaria), with a redefinition of Heteranthidae Carlgren, 1900 MrBayes 3.2: efficient Bayesian

Species boundaries delineating tropical sea anemones (Cnidaria, Actiniaria) of the zooxanthellate genus, Heteranthus Klunzinger, 1877, are unclear. There are currently two valid Heteranthus species: type species Heteranthus verruculatus Klunzinger, 1877, first reported from Koseir, Egypt, and H. insignis Carlgren, 1943, from Poulo Condore, Vietnam. In describing the latter from a single, poorly preserved specimen, zoologist Oskar Carlgren expressed apprehension with traits he had used to establish this species. Carlgren’s doubts persisted later in writing when he found a similar-looking sea anemone from the Great Barrier Reef. Crucial details to positively identify either species have since remained limited. Here, we re-diagnosed Heteranthus and re-described its type species based on observations of specimens we have obtained from Singapore and Pulau Ambon (Indonesia), and of museum material collected elsewhere across the Indo-West Pacific region (n > 180). Supported by molecular phylogenetic evidence, the family Heteranthidae Carlgren, 1900 was reinstated and re-diagnosed. Heteranthus verruculatus is encountered in the lower intertidal region amongst seagrass, in rocky crevices, or coral rubble. It occurs as solitary individuals or in clonal clusters, well-camouflaged against the substratum. Individuals were observed to frequently propagate by longitudinal fission, resulting in a varied appearance. Type material of H. verruculatus and H. insignis were re-examined and as we found no differences between them, the two were synonymised. We inferred that Carlgren probably misinterpreted cnidae and histological data in defining H. insignis as a distinct species. This revision clarifies the taxonomy and geographic range of H. verruculatus , an Indo-West Pacific species that is found from the Red Sea to subtropical Australia and Hawaii.


Introduction
As prolific as Swedish zoologist Oskar Carlgren was in describing hundreds of sea anemones (Cnidaria, Actiniaria) during his lifetime, uncertainties pervaded him as he struggled to distinguish between two species of the zooxanthellate genus Heteranthus Klunzinger, 1877. Currently, this genus comprises two valid species: Heteranthus verruculatus Klunzinger, 1877 and H. insignis Carlgren, 1943, the former being the type species of Heteranthus by monotypy (Daly & Fautin, 2020).
When Carlgren (1943: 32) first described H. insignis from a single, poorly preserved specimen collected from Puolo Condore [= Côn Sơn Island, Vung Tau] Vietnam, he stated, "I have with some hesitation referred the specimen to a new species, not to verriculatus [sic; = H. verruculatus]." Later, when he applied the name H. verruculatus to a specimen obtained from the Great Barrier Reef, his doubts further yap et al. lingered as he admitted (Carlgren, 1950: 440), "I have referred the single specimen with some hesitation to verruculatus …". This hesitation was also noted by Fautin et al. (2008: 52) in their entry of H. verruculatus from Moreton Bay, Australia. Clearly, Carlgren was doubtful of traits he had used to define species of Heteranthus, in particular, its type species H. verruculatus.
Heteranthus verruculatus is a sea anemone that is infrequently sighted because it is well camouflaged against the substratum it lives in, in addition to its relative small size (oral disc diameter <25 mm). We encountered individuals at the lower intertidal zone, each with its pedal disc attached to loose coral rubble or roots of seagrass buried underneath the sand, or sometimes within crevices of rocks. Populations of these sea anemones occur as randomly distributed solitary individuals or in densely packed, clonal clusters, consisting of tens of individuals. We have also observed that individuals frequently reproduce asexually by longitudinal fission, which appear to induce irregular symmetry and a variable appearance in resulting clones. In the aquarium trade, H. verruculatus are marketed and sold as 'sand anemones,' a name used by hobbyists that also encompasses species of Phymanthus (Fosså & Nilsen, 1998).
The objective of this study is to provide an updated and detailed characterization of H. verruculatus. This species is re-described, based on new data collected from its type material and more than 180 individuals assigned to this species. We also synonymise H. insignis with H. verruculatus after the holotype of the former was re-examined thoroughly, and since we found that evidence used to distinguish it as a new species agrees the sea anemone heteranthus verruculatus with what we obtained from the H. verruculatus type material (i.e., cnidae data and musculature; see Carlgren, 1943). Furthermore, we re-establish the family Heteranthidae solely for Heteranthus, based on molecular data from this study and others (Gonzalez-Muñoz et al., 2015;Brugler et al., 2018), which supports the polyphyly of Phymanthidae (Phymanthus + Heteranthus). As with many other sea anemone species throughout the world, data concerning H. verruculatus' biogeography and ecology are lacking. A detailed redescription of this widespread and common Indo-West Pacific sea anemone resulting from the present study allows further research on a firmer footing for a wide range of disciplines.

Specimen collection and processing
All living materials were collected from Singapore by hand, mainly by either digging individuals out of the substratum or extracting them from rock crevices using a chisel. Some individuals were observed and photographed in situ. Field sampling was also carried out between 2016 and 2019 in the Red Sea at Hurghada (Egypt), Eilat (Israel) and Thuwal (Saudi Arabia), these three sites being in the vicinity of H. verruculatus' type locality at Koseir (Egypt). Sea anemone surveys were further conducted at Zanzibar Island (Tanzania, Western Indian Ocean) and Pulau Langkawi (Malaysia, Eastern Indian Ocean). At all these localities apart from Singapore, no Heteranthus specimens were encountered (map in supplementary fig. S1).
Collected animals were brought back to the laboratory and kept alive for a week for further observation. For each sea anemone, details of its living appearance and any behaviour were noted. Tissues of tentacles or pedal discs of ten individuals were sampled and preserved in 100% molecular-grade ethanol for dna extraction. Afterwards, all sea anemone specimens were relaxed in 7.5% magnesium chloride in seawater and fixed in 10% formalin.

Morphological observations
To study the musculature of the sea anemone, serial transverse and longitudinal sections each 8 µm thick, were made from formalin-fixed specimens and stained with haematoxylin and eosin (Humason, 1967). Other specimens were dissected so that the internal morphology could be observed.
Unfired cnidae capsules were viewed and measured at 1000x magnification. These were extracted from tissues of the marginal and discal tentacle tips, marginal projections, mid-section column, actinopharynx, and mesenterial filaments. Conventionally, only measurements of unfired cnidae capsules are used in characterising the cnidom of a sea anemone species (Dunn, 1981). However, to verify Carlgren's (1943: 31) measurements of fired basitrich capsules from the holotype of H. insignis, we also measured these. Furthermore, we used fired capsules obtained from live individuals to confirm the identities of other cnidae types encountered in the tissues of the materials examined (see Yap et al., 2014). For some preserved specimens (including type material), we found it difficult to isolate and measure cnidae from the tissue sampled due to the state, age, and rigidity of the preserved material. In many instances the cnidae capsules visualized from these appeared degraded; we were unable to accurately identify or measure them. All identities of good cnidae capsules, which could be seen clearly and measured, were assigned following Mariscal (1974).

Molecular analyses
Genomic dna was isolated from column or pedal disc tissues of each sea anemone by standard ctab extraction (see Rodríguez et al., 2014, and references therein). Four molecular markers comprising three mitochondrial (i.e., partial 12S rDNA, 16S rDNA and cox3) and one nuclear (i.e. partial 28S rDNA) loci were targeted with polymerase chain reaction (pcr) using published cycling profiles (see Lauretta et al., 2013;Rodríguez et al., 2014 and references therein). We purified all pcr products with SureClean Plus (Bioline, Singapore). Cycle sequencing was carried out with BigDye Terminator v3.1 (Applied Biosystems, Foster City, California) following the manufacturer's protocol. Amplicons were sequenced on the abi 3130 XL Genetic Analyzer (Thermo Scientific).
All new sequences obtained in this study were assembled in Geneious v11.1.3 (Invitrogen Corporation) using default parameters. Contigs were searched against those available on GenBank via BLASTn, on the basis of sequence similarity, to affirm that they were of sea anemones. To determine the phylogenetic position of Heteranthus, the single published 12S sequence of Heteranthus sp. was included in our analyses (i.e., accession number: KC812147; see González-Muñoz et al., 2015), together with newly obtained and published sequences of taxa from both Actiniidae and Stichodactylidae Andres, 1883 (i.e., those published in Daly et al., 2017;Titus et al., 2019), and our own collections (supplementary table S1). To further clarify the relationship of Heteranthus and the family Phymanthidae, new sequences of Phymanthus loligo (Hemprich & Ehrenberg in Ehrenberg, 1834) collected recently from the Red Sea were also included (supplementary table S1).
Sequences for each marker were aligned using mafft v.7.313 under --auto setting. Aligned sequences of all mitochondrial and nuclear markers were also concatenated into a single matrix. We conducted separate analyses for alignments of each marker, and for the concatenated dataset. The concatenated dataset comprised 149 specimens with 6229 sites that were used for phylogeny reconstruction. Maximum likelihood (ml) analyses were conducted in RAxML v.8.2.11 (Stamatakis, 2014) with 100 random starting trees and 1000 the sea anemone heteranthus verruculatus bootstrap pseudo-replicates, under the gtrgamma model for both individual markers and the concatenated matrix. Pairwise genetic distances between samples for each molecular marker were calculated using Geneious v11.1.3.
Bayesian inference was carried out on the concatenated matrix using MrBayes v3.2.6 (Ronquist et al., 2012). Based on the Akaike information criterion (aic) implemented in jModelTest2 (Darriba et al., 2012), we selected GTR + I + G as the best-fit model for cox3, 12S, and 28S, whereas HKY + G was selected for 16S. We performed four Markov chain Monte Carlo (mcmc) runs for 12 million generations and sampled every 100th tree. To determine if the runs converged, we used Tracer v1.7 (Rambaut et al., 2018), and discarded the first 20,000 trees as burn-in. Overall, our four mcmc runs did not converge, as with observations from previous studies (see Titus et al., 2019). However, Bayesian inference (bi) probabilities were still included in our final results, on branches that were also recovered in the ml analyses.
All new H. verruculatus voucher specimens collected from Singapore were deposited in the Zoological Reference Collection (zrc), Lee Kong Chian Natural History Museum, National University of Singapore. All new sequences were deposited in GenBank (supplementary table S1). Sequence alignment for each molecular marker and pairwise genetic distances calculated were deposited in Zenodo (https://doi.org/10.5281/zenodo.4034666).
Remarks. Based on morphological similarity, Carlgren (1943) synonymised Heteranthidae and Phymanthidae, and classed Heteranthus in the latter. We resurrect Heteranthidae for Heteranthus as molecular phylogenetic results from this and previous studies (see González-Muñoz et al., 2015;Brugler et al., 2018) do not support Carlgren's (1943) conclusion. Following this line of evidence, we also revised the mention of higher-level classification in the original diagnosis. Among characters absent from Carlgren's (1900) diagnosis are cnidom and arrangement of the internal musculature. We included these here as they are a necessary part of contemporary sea anemone taxonomy.
Genus Heteranthus Klunzinger, 1877 Diagnosis modified after Carlgren (1949) and Fautin et al. (2008: 52). Substantial changes from our research are indicated in bold; all minor additions are in italics.
Diagnosis. Pedal disc well-developed. Column, apart from most proximal part, with verrucae, which are smaller and more numerous at the margin, and overhang the fosse, resembling papillae. Endodermal marginal sphincter muscle, restricted to circumscribed in form, may form a few folds. Tentacles as marginal and discal tentacles. Marginal tentacles conical, simple, without ramified protuberances, arranged in cycles, innermost cycle longer than outermost; discal tentacles shorter and papilliform, arranged as dense or sparse radial rows. Siphonoglyphs well-developed, when present. Number of directives pairs, and pairs of complete and incomplete mesenteries variable, gametogenic tissue may be present on both; retractor muscles diffuse to diffuse circumscript, well developed. Parietobasilar muscles weak to fairly strong. Frequently propagates by longitudinal fission. Cnidom: spirocysts, basitrichs, microbasic amastigophores, microbasic p-mastigophores Gender. Masculine Type species. Heteranthus verruculatus Klunzinger, 1877, by monotypy (see Fautin, 2016) Remarks. Carlgren (1949: 75) stated that the column is "provided with large verrucae," but we found this character to be variable in size and visibility, among preserved specimens. We adhere to the rewording by Fautin et al. (2008) in not specifying size of verrucae to allow for this variation. On marginal tentacles, Klunzinger (1877) reported that these were of equal length, but Fautin et al. (2008) found otherwise in studying live specimens. Likewise, we observed that the innermost cycle of marginal tentacles was longer than those at the outermost yap et al.
in living individuals. Note that in many preserved specimens, this feature may be less-pronounced and tentacles may contract. Specification of the number of siphonoglyphs present was removed as this varied among individuals or be absent entirely. Similarly, the number of directive pairs, and pairs of complete and incomplete mesenteries also varies among individuals. Like both Carlgren (1949) and Fautin et al. (2008), we too failed to find any gametogenic tissues in all but those specimens from Western Australia. Moreover, we observed that living polyps readily undergo longitudinal fission. The diagnosis is modified to indicate the presence of gametogenic tissue in some individuals and its reproductive mode.
We confirm the presence of microbasic amastigophores in live materials examined for cnidae; this cnidae type has not been reported in Heteranthus. Because unfired capsules of microbasic amastigophores strongly resemble those of microbasic p-mastigophores in preserved specimens, they may have been mistaken for the latter and were not reported (see Östman, 2000;Yap et al., 2014).
While Carlgren's (1900) publication dealt with material collected by Stuhlmann from East Africa, it also presented diagnoses of Heteranthidae and Heteranthus. Carlgren (1900: 92) did not report any species of the genus collected from that locality and explicitly states that, "… Heteranthus klunz. mit nur einer Spezies, H. verruculatus. Diese Spezies ist nicht in der STUHLMANN'schen Sammlung enthalten,". In other words, Carlgren did not find Heteranthus individuals from East Africa in Stuhlmann's collection. We infer that Carlgren's (1900) inclusion of the diagnoses served to facilitate discussion regarding East African Phymanthus specimens, which he described in detail in his publication.
Older labels within lot nrs 4861 provided further support that this fragment is part of the syntype kept in Berlin. Though faded, one label has the number "1852," written on it, which corresponds to the catalogue number present on the old label of the Berlin specimen ( fig. 2A). Other older labels, likely written by Carlgren himself, indicate that the fragment was from "Röda havet" [= Red Sea], and "Klunz orig!". In addition, the nrs fragment corresponds to the position of the missing slice of zmb 1852, though slightly smaller in width, with the oral disc and actinopharynx absent; presumably Carlgren had prepared histological slides from the nrs material (see Carlgren, 1950). This confirms that the H. verruculatus fragment numbered nrs 4861 is part of the syntype zmb 1852.
We hereby designate both zmb 1852 and nrs 4861, previously considered as separate syntypes but now found to be of the same syntype specimen, as the lectotype for the name Heteranthus verruculatus Klunzinger, 1877 (Code Article 74), to enhance its nomenclatural objectivity. While C.B. Klunzinger described the species from more than one individual that he had collected from Egypt (Klunzinger, 1877;Fautin, 2016), the first author was not able to locate any other syntypes at museums where Klunzinger's specimens are now kept, or in those that O. Carlgren  Heteranthus verruculatus -Lectotype. zmb 1852 ( fig. 2A, B), collected by C.B. Klunzinger, no later than 1877 from the Red Sea (Koseir). Two pieces of a single specimen, rigid and brittle, cut transversely across at mid-column ( fig. 2B). Distalmost column slit longitudinally, a slice of distalmost margin missing ( fig. 2B). An oily scum layer encrusts oral disc. Both pieces dark-greenish or grey. Total length 14 mm, mid-column 13 mm wide, distalmost and proximal end 15 mm wide, respectively; nrs 4861 ( fig. 2C) a single fragment originating from zmb 1852. A piece of distalmost end of column, also includes part of the mid-column, dark sandy-brown in colour, in good condition. Fragment 4 mm wide.
Heteranthus insignis -Holotype, nrs 4076, collected by D.C. Dawydoff from Vietnam and given to O. Carlgren ( fig. 2D). A single specimen, 17 mm in length, cut transversely into two pieces at mid-column, with a longitudinal cut along each. A slice of distalmost margin missing. Both pieces firm, in good condition. Both pieces cream-coloured in preservative.
Natural history. Encountered during low spring tides along shallow tropical coasts, with oral disc and marginal tentacles expanded ( Oral disc. When fully expanded, margin may be slightly undulate in solitary individuals; in smaller clustering ones, flat ( fig. 5A, B). Outline oval to round; elongated oval during onset of longitudinal fission ( fig. 4A). In life, light to dark-brown, may be speckled with white patches. Thin-walled, mesenterial insertions seen as radial dark lines extending from mouth to marginal tentacles in living animals, as white lines in preserved materials. Discal tentacles present, papilliform ( fig. 5C, D). Area immediately around mouth without discal tentacles ( fig. 5D), arranged in radial rows from mid-way extending towards marginal tentacles. Radial rows of discal tentacle both endocoelic and exocoelic, numerous. In solitary individuals, discal tentacles conspicuous and densely arranged, clustering type less conspicuous and less dense ( fig. 5D), becoming obscure when preserved. Central mouth flat, edges tinged brown, white or pink in life, cream-white in preserved specimens.
Column. Distalmost end flared slightly outwards when expanded. Distal margin with marginal projections, both endo/exocoelic.  6F). Longitudinal verrucae rows both endo/ exocoelic. Endocoelic rows longer, up to 20 verrucae per row; exocoelic: shorter, up to four in each row; shell fragments, coral rubble and/or sand grains may adhere to the verrucae. Column wall thin, mesenterial insertions can be seen through as faint light lines extending from distal to proximal end. From mid-column to proximal end of column smooth, without cinclides. Fosse present, shallow. In life, light-brown with greenishtinge to cream-white with reddish orange/ pink splotches, distalmost end dark grey ( fig. 6E-F). In preserved specimens, column entirely cream-white; distalmost end grey in recently preserved materials. Pedal disc. Adheres readily to surfaces when animal is alive. Outline: oval, may be flat or inflated. Limbus slightly scalloped. Thin-walled, mesenterial insertions seen as light lines. Diameter may be equal or greater than oral disc.

Discussion
Comparisons of type and voucher specimens. The animals from Singapore and Pulau Ambon, and of museum vouchers collected elsewhere (e.g., Oman and Australia), closely resemble H. verruculatus which Klunzinger had described and depicted (see Klunzinger, 1877: pl. 5, fig. 9). They also agree in habitats where Klunzinger collected his animals from (i.e., seagrass, rock clefts, small stones; see Klunzinger, 1877: 84 (1877) and what we have found. Klunzinger's (1877) original description did not include data on internal musculature form or cnidom, common in many sea anemone taxonomic descriptions published at that time. Overall, our data on the retractor and sphincter musculature of vouchers agree with Carlgren's (1943Carlgren's ( , 1950 observations of the H. verruculatus type material. While the sphincter muscle was circumscribed in most individuals (e.g., NHM 1954.6.28.23), in others we found that it may vary in shape and form; it can be restricted (e.g., MTQ-G59391) or multilobed ( fig. 8C, wam Z33600). In comparing new cnidae measurements made from the lectotype fragment of H. verruculatus (nrs 4861), and voucher materials we studied, the overlap is nearly complete. Our measurements also agreed with those made by Carlgren (1943 Carlgren (1943), for the type material (nrs 4861). Outliners of single capsules are indicated within parentheses, '?' denotes the unspecified number of capsules that Carlgren had measured. Abbreviations: N = number of specimens of which that type of cnida was found to total specimens examined; n = total number of undischarged capsules measured. Letters in parentheses following cnidae type refer to illustrations in fig. 9. the sea anemone heteranthus verruculatus Carlgren (1950) failed to find gametogenic tissue in H. verruculatus specimens he had studied and did not mention them when he described H. insignis (see Carlgren, 1943). Because of this, he was unclear about its distribution of these reproductive tissues as evident in his diagnosis for Heteranthus (Carlgren, 1949: 75). Similarly, on the H. verruculatus specimen he had collected from the Great Barrier Reef, he surmised that, "… the arrangement of gonads [= gametogenic tissue] … they are probably absent" (Carlgren, 1950: 440). Likewise, Fautin et al. (2008) did not find gametogenic tissue in specimens collected from the eastern coast of Australia. Like Carlgren (1950) and Fautin et al. (2008), the first author failed to find gametogenic tissue in the specimens examined by them and in all other vouchers studied (i.e., those from Singapore, Pulau Ambon). However, in the present study oocytes were found in the material collected from Rottnest Island in Western Australia (wam Z33600), which were in the early phase of oogenesis, attached to complete and incomplete mesenteries ( fig. 7C).
As stated at the outset, Carlgren (1943: 32) described H. insignis based on a single poorly preserved specimen. In studying the type specimens, we could not find differences between the holotype of H. insignis and the lectotype of H. verruculatus. The length of basitrich capsules from the holotype's column was one of the characters Carlgren (1943) used to distinguish H. insignis as a new Heteranthus species. He reported finding discharged column basitrichs as 33.8-38.1 × 2.5 µm from the H. insignis holotype, which are much larger than those present in the H. verruculatus syntype (Carlgren, 1943). Carlgren placed great emphasis on this character to differentiate between Heteranthus members, as evident from remarks in his later publication (see Carlgren, 1950: 440). Like him, we found large, discharged basitrichs in H. insignis holotype's column, and did not encounter any undischarged ones. However, we also found numerous smaller-sized, column basitrichs, figure 9 Cnidae of Heteranthus verruculatus Klunzinger, 1877 the sea anemone heteranthus verruculatus for which the size and distribution were consistent with those found in the H. verruculatus lectotype fragment (NRS4861), and all these we encountered were undischarged in both holotype and syntypes examined. These smaller column basitrichs were more transparent than the larger basitrichs in the H. insignis holotype and Carlgren likely may have missed them. All column basitrichs that Carlgren (1943) measured were discharged; we suspect that his data represent contamination from another tissue. In our experience, in fixed or preserved specimens, most cnidae capsules typically remain undischarged in the tissue from which it was produced. Carlgren's (1943) observations were probably based on contamination from the mesenterial filaments, as we found large undischarged basitrichs there, conforming in both size and form with Carlgren's (1943) column basitrichs. Carlgren (1943) also stated that sizes of actinopharynx cnidae present in H. insignis' holotype were different from those found in the H. verruculatus syntype, referring to his earlier work. He makes no mention of cnidae type present in this tissue, simply stating, "… the sizes of the nematocysts in the actinopharynx seem not to agree" (Carlgren, 1943: 32). We scoured the study (i.e., Carlgren, 1900) that Carlgren (1943) had cited but failed to find any Heteranthus cnidae data within it, even within its genus entry (see Carlgren, 1900: 92). We further reviewed all his pre-1943 publications (e.g. Carlgren, 1940) that might justify the claim but found none to support it. Presently, the well-preserved fragment of the H. verruculatus lectotype (nrs 4861) lacks an actinopharynx tissue, while the remaining material in Berlin (zmb 1852) is dried up, and not in a condition to yield any useful data from it.
Nevertheless we suggest that Carlgren (1943) may have been measuring basitrichs from the actinopharynx, based on data from his Great Barrier Reef specimen (i.e., NHM 1954.6.28.23). Basitrich sizes of the H. insignis holotype agree with these. Furthermore, Carlgren appears to abandon the use of this character to distinguish between the two Heteranthus species, as he made no reference to the difference of actinopharynx cnidae sizes between H. insignis and H. verruculatus in this latter publication (i.e., Carlgren, 1950). We also found microbasic p-mastigophores in the actinopharynx of the H. insignis holotype and NHM 1954.6.28.23, of which their size range overlaps well; Carlgren overlooked the presence of these cnidae in the actinopharynx as well.
Another character that Carlgren (1943) used to support his description of a new Heteranthus species was the form of the mesentery muscle. He reported that complete mesenteries of H. insignis "seem rarely to form humps," and that "the parietobasilar muscles form usually no distinct projection on the mesenteries; rarely is such a projection here …" (Carlgren, 1943: 32). We prepared histological cross-sections from the holotype, at the level which Carlgren (1943) did, and found otherwise: numerous 'humps' were present along the edges of complete mesenteries, and parietobasilar muscles extending as a pennon at the base of each mesentery were fairly common ( fig. 7A).
Although H. verruculatus and H. insignis were collected from two widely separated localities (the Red Sea and Vietnam, respectively), the overall morphology for both type specimens were virtually indistinguishable. In our re-examination of type material, we found that the characters used by Carlgren (1943) to separate the two species are questionable and inconsistent. All other characters (e.g., external morphology, sphincter muscle form, cnidom of remaining tissue, etc.) present in both type materials corresponded well with each other. We are therefore certain that the animal Carlgren (1943) had described as yap et al. H. insignis is H. verruculatus instead, and we thus synonymize the two species. England (1987: 207) reportedly found H. verruculatus from the intertidal mud in Aden. In examining England's Heteranthus specimens from Aden (NHM 1972.5.25.15 x2), we found that they were not this species. Based on the overall appearance in having only one tentacle type which are tightly packed in wedges with spaces in between each group, an absence of papillae-covered marginal projections and its cnidae type and sizes, these are actually Stichodactyla tapetum (Hemprich & Ehrenberg in Ehrenberg, 1834), following evidence presented in Carlgren (1943Carlgren ( , 1949Carlgren ( , 1950, Dunn (1981) and Fautin et al. (2008Fautin et al. ( , 2009. Comparisons with other species. Clonal clusters of H. verruculatus, as densely packed populations of similarly sized polyps, may be mistaken for dense mats of Zoanthus spp. (Cnidaria, Zoantharia); both occupy the same habitat ( fig. 3C). Zoanthus spp. are colonial invertebrates with their polyps interlinked by a network of connective tissue. Heteranthus verruculatus does not form colonies and exists as individual polyps. While in some instances H. verruculatus may appear as interconnecting polyps, this is temporary, and it is likely in the midst of longitudinal fission; the pair soon splits after.
In sharing the same habitat types and having longitudinal rows of conspicuous verrucae that extend distally from marginal projections, H. verruculatus superficially resembles species of Anthopleura or Oulactis. Furthermore, the distal margins of both H. verruculatus and Oulactis spp. appear frilly, due to the dense presence of overhanging verrucae. This is reflected in museum voucher specimens that were erroneously identified as either species of Anthopleura or Oulactis (e.g., MNHN-IK-2019-8, wam Z33601). When expanded, H. verruculatus is easily distinguishable due to the presence of discal tentacles arranged in radial rows on its oral disc; this feature is absent in the other two genera.
Stichodactyla tapetum (i.e., NHM 1972.5.25.15) is another species that H. verruculatus has been misidentified with. Individuals are similar in having an oral disc covered with tentacles, being small in size as well as being zooxanthellate. However, they are distinct from each other because: 1) Tentacles of Heteranthus verruculatus are clearly differentiated into two types: discal and marginal, whereas Stichodactyla tapetum only bears one tentacle type. 2) Tentacles present on the oral disc of S. tapetum are bulbous, may appear hexagonal in form, and are tightly arranged as dense radial wedges that extend from the mouth to the distal margin, while those of H. verruculatus are papillae-like and not arranged in dense radial wedges. Moreover, discal tentacles of H. verruculatus do not extend up until the distal margin, and are absent around the mouth. 3) Stichodactyla tapetum lacks both conspicuous verrucae on its column and a distal margin that overhangs with verrucae, whereas these are distinct and present on the column of H. verruculatus. 4) While both H. verruculatus and S. tapetum occur in similar habitats (e.g., rocky crevices, seagrass meadows), we have not encountered both species co-occurring in the field. With the presence of both marginal and discal tentacles, H. verruculatus may be confused for species of Phymanthus. Some specimens have been identified as such (e.g., ufm 4220), and Carlgren (1943) had thought both are closely related. Yap et al. (2019: 15) provided a detailed discussion on features distinguishing Heteranthus and Phymanthus.
Here we expand on that with the following the sea anemone heteranthus verruculatus observation: Heteranthus verruculatus reproduces asexually frequently, forming aggregating clonal populations, whereas individuals of Phymanthus tend to occur singly; so far asexual reproduction has not been observed in Phymanthus.
Phylogeny of Heteranthidae. Our phylogeny reconstruction of Heteranthus showed that it overall formed a group that was distinct from Phymanthus, corroborating González-Muñoz's et al. (2015) findings. Moreover, we recovered Phymanthus as polyphyletic, in agreement with Brugler et al. (2018: fig. 3) and Titus et al. (2019: fig. 2). As of writing, a revision of the genus Phymanthus is ongoing and its relationship with Heteranthus will be discussed in a separate manuscript. While support for the Heteranthidae clade in ml analyses was weak (i.e., bootstrap = 54; fig.  1), those of Bayesian analyses showed otherwise (posterior probability >0.8). Nonetheless, both optimality criteria agreed in the overall topologies (i.e., Heteranthus forms a distinct clade) for three of the four molecular markers (16S excluded) that we used as well for the concatenated dataset.
Phylogenetic relationships for most sea anemones remain poorly resolved at the species level (Daly et al., 2017;Titus et al., 2019). This could be attributed to two reasons: 1) current genetic markers are of limited utility to infer relationships at species level, and 2) insufficient taxon and genetic sampling (e.g., Rodríguez et al., 2014;González-Muñoz et al., 2015; this study). On the latter, we have increased the sample size of H. verruculatus and the number of genetic markers. Among our three mitochondrial markers, 16S differentiated some H. verruculatus individuals into different branches, although they were collected from the same population (i.e., ZRC.CNI.1230, supplementary table S1; fig.  1, supplementary figs. S2-S6). Despite this, we are certain that these individuals are of a single species (i.e., H. verruculatus); agreement in morphological features of these polyps and comparisons with others in the same population support this, despite the variation in some genetic markers used (i.e., 16S; https://doi.org/10.5281/zenodo.4034666). Nevertheless, the overall topologies of our ml and Bayesian trees are similar. We resurrected the family Heteranthidae because molecular data, including those obtained here, have repeatedly shown that Heteranthus and Phymanthus do not form a monophyletic group (see González-Muñoz et al., 2015;Brugler et al., 2018).

Summary and conclusion
In re-describing sea anemone species first reported in early sea anemone taxonomic works, we concur with Dunn's (1974: 177) view that the accompanying descriptions in those studies were, "… vague as to be virtually worthless for taxonomic purposes." Many species in those early works were mostly described from a single specimen, and species boundaries defined were unclear. We suspect that Carlgren's (1943Carlgren's ( , 1950 'hesitation' was in part because he had only examined three Heteranthus individuals during his lifetime, of which two were in poor condition (i.e. type specimens of H. verruculatus and H. insignis). Furthermore, like many of his peers, he was probably not thorough in examining the syntype of H. verruculatus. Carlgren (1950: 440) laments this in his later publication that, "… unfortunately I have no opportunity of examining the nematocysts of verruculatus in further detail." The animal's variable appearance also likely exacerbated Carlgren's uncertainty. Perhaps if he had re-examined the H. verruculatus syntype again and studied more individuals of H. verruculatus, he would have his hesitation allayed. yap et al.
Herein, we marshalled evidence from morphological, molecular, and historical data (i.e., specimen lot label details) from over 180 specimens collected across the Indo-West Pacific, to resolve the identity of H. verruculatus and define its species boundaries. For the first time since Klunzinger's (1877) description, we provide colour illustrations of H. verruculatus' living appearance, an updated cnidae dataset, an expanded description of the animal's internal musculature, and evidence of its reproductive behaviour. Moreover, in re-examining the holotype of H. insignis, we did not find any differences with the lectotype of H. verruculatus, and demonstrated that H. insignis is a junior synonym of H. verruculatus. As with another of Klunzinger's (1877) species that was re-described recently (Yap et al., 2019), we suspect that to resolve taxonomic confusion among many sea anemone species described in early accounts, this line of thorough treatment is required.