Hydroids ( Cnidaria , Hydrozoa ) from marine fouling assemblages in the Galápagos Islands , Ecuador

An account is given of hydroids collected in 2015 and 2016 from port and harbor fouling communities in the Galápagos Islands. Also included is the hydroid of Ectopleura media, discovered on the wreck of the tanker Jessica near Isla San Cristóbal in 2001. Among 20 species reported herein were six anthoathecates and 14 leptothecates. Most common in the samples were the kirchenpaueriid Ventromma halecioides and the halopteridid Halopteris alternata. Eight species (Bougainvillia muscus, Bimeria vestita, Clytia elongata, C. obliqua, C. thornelyi, Obelia oxydentata, Eucheilota sp., and Halecium labiatum) are reported for the first time from the Galápagos archipelago. Three of them (Clytia elongata, C. thornelyi, and Halecium labiatum) are also new to the eastern Pacific. Seven species treated here are considered introduced by shipping to the islands, bringing to eight the number of introduced hydroids. In addition, we treat four species as cryptogenic, bringing to five the total number of the latter in the Galápagos. The binomena Obelia thornelyi Nutting, 1927, Clytia stolonifera Blackburn, 1938, and C. latitheca Millard and Bouillon, 1973 are regarded as synonyms, with the first of these having nomenclatural priority. Meanwhile, the senior synonym O. thornelyi is reassigned to the genus Clytia Lamouroux, 1812 as C. thornelyi. Another species in the samples, Dynamena distans Lamouroux, 1816 (also widely known as Sertularia distans and Tridentata distans), is combined for the first time with the genus Amphisbetia L. Agassiz, 1862, as A. distans.


Introduction
Hydroids of the Galápagos Islands have received only modest attention.The most comprehensive contributions to knowledge of the group in the region are by Fraser (1938aFraser ( , b, 1948) ) on collections acquired in the archipelago during the Allan Hancock Pacific Expeditions of 1931-1932, 1933, 1934, and 1938. Earlier, Clarke (1907) reported three species collected during the 1904-1905 cruise of the United States Fish Commission Steamer Albatross.Hastings (1930) added a new species of zancleid (Zanclea protecta) from James Island (Isla Santiago) and from Pánama (type locality: Isla Taboga).Houvenaghel and Houvenaghel (1974) identified one hydroid (Dynamena crisioides Lamouroux, 1824) to species, and mentioned several others that were identified to genus only, in an investigation on vertical zonation of rocky intertidal biota at Isla Santa Cruz.Best known of all hydrozoans in the Galápagos is the family Stylasteridae Gray, 1847. Cairns (1986) reviewed earlier studies on stylasterids from the islands and provided accounts of 14 species, nine of them as new.Cairns (1991) added another new species, but the total number remained at 14, with one earlier species being placed in synonymy.Additions to the general hydroid fauna, and a checklist of species, were provided by Calder et al. (2003).Marshall et al. (2002) and Marshall and Edgar (2003) called attention to hydroids discovered on and adjacent to the wreck of the oil tanker Jessica off Isla San Cristóbal in 2001.A Galápagos field guide by Hickman (2008) included accounts of 12 of the more prominent hydroid species.In an overview of the local marine invertebrate fauna (Hickman 2009), bryozoans and hydroids were reported to be two of the more diverse groups.Including 14 stylasterids (Cairns 1991) and 96 others (Calder et al. 2003), 110 species have been recorded from the islands.
Most previous investigations of hydroids in the Galápagos have focused on species from natural habitats.The present study is based primarily on hydroids found as part of fouling communities, particularly those in ports and harbours.

Materials and methods
Hydroids from the Galápagos Islands examined here were obtained during surveys for invasive species in February 2015 and April 2016.Manual collecting from docks and pilings was undertaken in the vicinity of Puerto Ayora (Isla Santa Cruz) and at Isla Baltra (Figure 1).A single sample came from Isla Bartolomé.In addition, in April 2016, fouling panels [14 × 14 cm, 0.5 cm thick, grey polyvinyl chloride (PVC) plates, lightly sanded on the underside to optimize attachment conditions, and suspended horizontally at a depth of 1 m] that had been deployed 14 months or 3 months earlier in February 2015 and January 2016 respectively at (1) the Puerto Ayora main passenger docks, (2) a private dock in Franklin's Bay, Puerto Ayora, both on Santa Cruz Island, and (3) on a Navy floating dock on Baltra Island, were retrieved.Voucher samples of hydroids (and other taxa) were removed and preserved in 95% ethanol.An additional sample of hydroids was collected from the wreck of the oil tanker Jessica at Isla San Cristóbal in 2001 by G. Edgar and P. Marshall.

Results
Twenty species (six anthoathecates and 14 leptothecates) are reported here, eight of which are new to the islands (Table 1, Supplementary material Table S1).We note whether species may be cryptogenic or introduced, but otherwise we consider the species treated here as native (although whether they should bear the names of "cosmopolitan" species with type localities in other oceans often remains to be determined).
Unless otherwise stated, all vials referred to are SERC numbers.
Remarks.-The type locality of Ectopleura media Fraser, 1948, presently known only from the Galápagos, is Bindloe Island (Isla Marchena).It has also been reported from Punta Vicente Roca, Isla Isabela (Calder et al. 2003) and Isla San Cristóbal (Marshall et al. 2002;Marshall and Edgar 2003).
At the latter location, specimens were found on and adjacent to the wreck of the oil tanker Jessica, which grounded off Puerto Baquerizo Moreno on 16 January 2001.By May 2001, when assessments were undertaken by Marshall and colleagues, colonies of the hydroid were a conspicuous component of the fouling assemblage on and adjacent to the wreckage.Fraser (1948) believed that gonophores of Ectopleura media were medusa buds.However, from specimens observed previously in the Galápagos by us, they are medusoids that do not become released as free medusae.In that, they differ from those of the sympatric E. integra (Fraser, 1938a), whose type locality is Isla Baltra, Galápagos (Calder et al. 2009).Gonophores of E. crocea (L.Agassiz, 1862), reported from Isla Isabela by Fraser (1938a), are sporosacs, with those of the female having 6-8 crest-like processes distally (Schuchert 2010).Fraser's record of E. crocea from the archipelago has been considered doubtful (Calder et al. 2003(Calder et al. : 1205)), as it is a species more characteristic of cool temperate than tropical waters.Mills et al. (2007) and Carlton andEldredge (2009, 2015)   Order Filifera Kühn, 1913Family Oceaniidae Eschscholtz, 1829 ?Turritopsis sp.
Remarks.-Although these specimens lacked gonophores and appear to have been quite young, they are unmistakably oceaniids.The colonies are similar to stolonal hydroids from Fiji that Gibbons and Ryland (1989) identified as Tubiclava sp.They also resemble Tubiclava triserialis Fraser, 1938a, a species inquirenda originally described from the Islas Revillagigedo of Mexico in the warm eastern Pacific.As for Tubiclava Allman, 1863, Schuchert (2004) justifiably regarded it as a genus of doubtful identity that should no longer be recognized as valid.Another hydroid of somewhat similar morphology is that of Oceania armata Kölliker, in Gegenbaur et al., 1853.While the medusa stage of that species is widely distributed and wellknown, its hydroid had until recently been known only as juveniles from laboratory cultures of Mediterranean material (Metschnikoff 1886;Schuchert 2004).However, the medusa of O. armata has now been linked by Schuchert (2016) to the hydroids of Turritopsis chevalense (Thornely, 1904) and T. fascicularis Fraser, 1943 via DNA barcoding.Hydroids of both nominal species are relatively large and polysiphonic, and distinct from the miniscule ones observed here.There is no record of this hydromedusa, or of the hydroids above, in the tropical eastern Pacific.
In having clavate hydranths with scattered filiform tentacles, a prominent hypostome, and an exoskeleton comprising non-nested tubes of perisarc, we have assigned our specimens with question to the oceaniid genus Turritopsis McCrady, 1857.Further identification to species level on the basis of morphology is inadvisable based on the available material.The oceaniid T. nutricula McCrady, 1857 has been recorded earlier in the Galápagos from Isla Isabela (Fraser 1938a) and Isla Wolf (Calder et al. 2003).If our hydroid is indeed T. nutricula, a species originally described from Charleston, South Carolina, USA, and reported worldwide in temperate and tropical waters, shipping would be a likely means of introduction.
As for Turritopsis nutricula, molecular studies now indicate the probable existence of a species complex hidden under that name.According to Miglietta et al. (2007) and Miglietta and Lessios (2009), the hydrozoan originally assigned that binomen may be restricted in distribution to the western Atlantic Ocean.Related species originally described from the Indo-Pacific region, and their type localities, include T. rubra (Farquhar, 1895) from Wellington Harbour, New Zealand, T. lata von Lendenfeld, 1885 from Port Jackson, Australia, T. pacifica Maas, 1909 from Sagami Bay, Japan, T. chevalense from Sri Lanka (Gulf of Manaar), and T. minor Nutting, 1905 from Maui, Hawaii.Of these, molecular sequencing has been undertaken to date on T. rubra, T. lata, and T. chevalense (=Oceania armata).Species limits and distributions within the genus remain obscure, and our sterile material is identified here simply as ?Turritopsis sp.
Remarks.-These hydroids, although sterile, were assigned to Bougainvillia muscus (Allman, 1863) based on their similarity to colonies studied earlier in Hawaii (Calder 2010).The latter specimens had been identified after examination of both hydroid and medusa stages.As noted earlier, B. muscus is presumably a eurytopic Atlantic species (type locality: Devon, England) that has also been reported in temperate and tropical waters of the Pacific and Indian oceans.Molecular data revealing low divergence values between populations from Europe and New Zealand (Schuchert 2007) support the conclusion that the species is widely distributed.Its hydroid is sometimes part of the fouling community in ports and canals (Billard 1926, as B. ramosa muscus;Millard 1959, as B. ramosa;Schuchert 1996;Calder 2010), as in this study, suggesting that long-range transport has been effected by shipping.Reports of the species from several other remote oceanic islands, including Bermuda (Calder 1988), New Zealand (Schuchert 1996), and Hawaii (Calder 2010), also provide evidence of its capacity for humanmediated dispersal.The medusa stage of B. muscus is also known to have a wide geographic range.While most accounts of it are from the eastern North Atlantic and Mediterranean Sea (Kramp 1961, as B. ramosa), records also exist from the Pacific and Indian oceans (Bouillon 1980, as B. ramosa;Navas-Pereira and Vannucci 1991, as B. ramosa;Schuchert 1996;Santhakumari and Nair 1999, as B. rumosa;Xu et al. 2014a).
Information on the biology of Bougainvillia muscus has been summarized by Schuchert (2007).This constitutes the first record of the species, a likely introduction, from the Galápagos Islands.
?Bougainvillia sp.  that in life it had a "tall hypostome".In preserved material, the hypostome was dome-shaped.The colony was stolonal, and hydranth pedicels tapered very gradually from distal to proximal end.In lacking gonophores, identification to species is impossible based on morphology.

Figure 2f
Bimeria vestita Wright 1859: 109, pl. 8, fig. 4 Wright, 1859. Earlier, Fraser (1938a, b) reported the species from the coast of Ecuador, as well as from the Pacific coasts of Mexico and Pánama.This is the first record of the species, or one like it, from the Galápagos.A distinct variety of the species (B.vestita forma nana) was described from Sri Lanka by Leloup (1932), and he identified it again from Vietnam (Leloup 1937).Confirmation is needed whether populations identified as B. vestita from warm waters of the Pacific are the same as Wright's (1859) species from the Firth of Forth, Scotland.Worldwide distribution records are summarized by Schuchert (2007), with reports of the species from temperate and tropical waters of the Atlantic, Pacific, and Indian oceans.We consider B. vestita to be cryptogenic in the Galápagos.
Knowledge of the gonophores, and especially those of female colonies, is now considered critical in the characterization of species of Eudendrium Ehrenberg, 1834.So too is knowledge of the cnidome.Eudendrium breve constitutes a species inquirenda because type material was sterile and its nematocyst complement has yet to be unequivocally established.Large isorhizas in addition to microbasic euryteles were found in specimens assigned with question to E. breve by Cooke (1975) from Enewetak Atoll.Such nematocysts were not observed in material examined here.Re-examination and redescription of type material (SBMNH 345383, SBMNH 345384, SBMNH 346372) of the species, in collections at the Santa Barbara Museum of Natural History, Santa Barbara, California, is needed.Our specimens are assigned to E. breve with question given the inadequate knowledge of the species, and with nothing but sterile material available in the present collection.
In addition to records from the Galápagos and Enewetak Atoll, Eudendrium breve has also been reported from Baja California (Fraser 1948).
Remarks.-The hydroid known as Egmundella amirantensis Millard and Bouillon, 1973 was shown, through life cycle studies by Migotto and Cabral (2005), to be conspecific with the medusa Cirrholovenia tetranema Kramp, 1959.The latter name has nomenclatural priority and is accepted as valid in WoRMS (Schuchert 2018).
The medusa of Cirrholovenia tetranema, originally described by Kramp (1959) from the Solomon Islands (holotype) and from several other localities in the Indo-Pacific, is now known from coastal and shelf waters of the Atlantic, Pacific, and Indian oceans (Migotto and Cabral 2005).Based on the known distribution of its medusa, Kramp (1968: 190) suspected that C. tetranema was native to the Indo-Malayan region.The hydroid stage, first described as E. amirantensis from Amirante, Seychelles (Millard and Bouillon 1973), is also taken to be circumglobal in tropical and warm temperate waters (Migotto and Cabral 2005).Other hydroids are a common substrate of the species, although colonies have also been found on bryozoans, macroalgae, and rocks.
Hydroids of this cryptogenic species have been reported earlier from the Galápagos at Isla Daphne Chica, Isla Española, Isla Santa Cruz, and Isla Isabela (Calder et al. 2003, as Lafoeina amirantensis).
Lesueur, 1810 collapse as hydranths develop, leaving a wrinkled skirt of perisarc hanging from the hydrothecal bases.The presence of punctae above a basal diaphragm reinforces the impression of a haleciid.Our hydroid may be referable to Eucheilota comata (Bigelow, 1909), the medusa stage of which occurs in the tropical eastern Pacific, but definitive identification is impossible in the absence of life cycle or genetic information.Reduction of the hydrotheca accompanying growth of the hydranth also occurs in several other species of leptothecate genera including Eutonina Hartlaub, 1897, Eirene Eschscholtz, 1829, andEutima McCrady, 1859.In some works, Eucheilota and Eucheilotidae Bouillon, 1984 have been included in Lovenellidae Russell, 1953. Bouillon (1984) and Cornelius (1995a), among others, are followed here in recognizing Eucheilotidae as a distinct family.The hydroid stages of eucheilotids, where known, are much different morphologically from those of lovenellids (see Cornelius 1995a), evidence that these families are distinct.Molecular studies (e.g.Leclère et al. 2009;Maronna et al. 2016) that have explicitly or implicitly sunk Eucheilotidae in the synonymy of Lovenellidae have misinterpreted relationships of the two because certain included species were assigned to the wrong genus.For presumed relationships of these families to be based on sound evidence, genetic comparisons of the type species of their type genera, Eucheilota and Lovenella Hincks, 1868 (Eucheilota ventricularis McCrady, 1859 and Campanularia clausa Lovén, 1836, respectively), are warranted.
No hydroids resembling Eucheilota sp. have been reported before from the Galápagos.
Clytia elongata, a probable introduction to the tropical eastern Pacific, is previously known only from the western Pacific.Originally described from Auckland, New Zealand, it has been reported elsewhere from New South Wales, Australia (Ritchie 1911) and Cook Strait, New Zealand (Vervoort and Watson 2003).Rees and Thursfield (1965) re-examined Ritchie's Australian specimens and found that they "…agree well with the figure given by Marktanner-Turneretscher…".By contrast, Vervoort and Watson expressed uncertainty that specimens examined by them from Cook Strait were the same species.In their illustration of the hydrotheca, marginal cusps are more deeply incised than shown by Marktanner-Turneretscher, and the hydrothecae are much larger.Those characters also differ from specimens examined here.Clytia elongata Warren, 1908, a name applied to a different species from Natal, South Africa, is an invalid junior primary homonym.That binomen has been replaced by the name C. warreni Stechow, 1919.Hydroids of Clytia elongata from the Galápagos were found on Amathia verticillata, a ctenostome bryozoan that is also non-native to the archipelago (McCann et al. 2015).
Clytia linearis, taken to be circumglobal in tropical and subtropical waters (Medel and Vervoort 2000), was reported earlier from the Galápagos by Calder et al. (2003).Included in the synonymy of the species by Calder et al. (2009) were C. acutidentata Fraser, 1938a from the Galápagos (Isla Floreana) and the Pacific coast of Mexico (Barra de Potosi), C. carinadentata Fraser, 1938a from the Galápagos (Isla Isabela), and Gonothyraea serialis Fraser, 1938a from the Pacific coast of Colombia (Puerto Utria).Also included in the synonymy of C. linearis is Obelia striata Clarke, 1907 from two locations in the tropical eastern Pacific (Rees and Vervoort 1987;Calder 1991a).Although one of these records was based on a colony from a trawl sample made at a depth of 2320 fm (4243 m), the specimen was found on a pteropod likely picked up in much shallower water.Another record of the species in the warm eastern Pacific is that of Calder (1996) from Rocas Alijos, west of Baja California, Mexico.
While Clytia linearis is a substrate generalist (Calder 1991a, b), it is also known to be frequent on shells of certain species of pteropods (Rees and Vervoort 1987).Such an association might account, at least in part, for the wide distribution of this hydroid.Alternatively, a circumglobal distribution may indicate existence of a species complex under the name.

Figure 3e
Campanularia (?) obliqua Clarke, 1907: 9, pl. 5, figs. 1-4.Material.-IslaSanta Cruz, Puerto Ayora, Puerto Ayora dock, 25 April 2016, on stem of the hydroid Halopteris alternata, 1 colony, 2 mm high, without gonophores, coll.J.J. Mallinson, JJM Hyd. 2. Remarks.-In the morphology of its trophosome, the hydroid examined here corresponds with species from the Indo-Pacific region that have been identified as Clytia gracilis (M.Sars, 1850) (e.g.Gibbons and Ryland 1989;Schuchert 2003), and as C. warreni Stechow, 1919(e.g. Millard 1975;Watson 2000).The type localities of those two species are Lofoten, Norway, and Algoa Bay, South Africa, respectively.More likely, hydroids from the Galápagos are conspecific with the morphologically similar Campanularia (?) obliqua Clarke, 1907, originally described from Isla Perico, Pánama.Clarke's species is referable to Clytia, as first suggested by Fraser (1936) and repeated later by Hirohito (1995).In an earlier work (Calder 1991a), Clytia obliqua was included in the synonymy of C. linearis (Thornely, 1900), but we regard them as distinct because of differences in the morphology of their marginal cusps.In C. obliqua, there are no keel-like thickenings of perisarc on the inner edge of each cusp as in C. linearis.A degree of uncertainty in our identification arises because species of the genus Clytia Lamouroux, 1812 are often inadequately distinguished and in need of revision.Identification of our specimens is also hindered by the absence of gonothecae.
Clytia obliqua has been reported from Japan (e.g.Fraser 1936; Yamada 1959; Hirohito 1995) and from California (Fraser 1948) in addition to the type locality of Pánama.Hirohito expressed uncertainty whether hydroids from Japan were actually conspecific with those reported by Fraser from the eastern Pacific.

Figure 3f
Obelia thornelyi Nutting 1927: 202.Material.-IslaBaltra, navy dock, collected April 2016 from fouling plate deployed 15 January 2016, 10 colony fragments, up to 20 mm high, with gonothecae, coll.K. Larson, vial 233083.-Isla Santa Cruz, Puerto Ayora, Franklin's Bay, fouling community on a pontoon, 27 April 2016, 6 colony fragments, up to 5 mm high, without gonophores, coll.J. Geller, JJM Hyd.11.Remarks.-Thesespecimens corresponded with accounts of Clytia latitheca Millard and Bouillon, 1973, a distinctive clytiid hydroid originally described from the Seychelles.The species is unusual in having hydrothecae with strongly flaring hydrothecal walls, resembling a shallow funnel.While hydrothecae somewhat resemble those of the Atlantic C. noliformis (McCrady, 1859), they differ in having a thin diaphragm and in lacking a distinct subhydrothecal spherule.Well-developed, fertile colonies were found on fouling plates immersed at the Baltra navy dock (SERC 233083).Gonothecae were paddle-shaped in lateral view, with smooth walls and a truncate distal end.
As for Clytia latitheca, it has been considered identical with a hydroid from Papua New Guinea identified by Thornely (1900) as Obelia serrulata (Bale, 1888) (see Kirkendale and Calder 2003).Working on specimens from the Philippines, Nutting (1927) justifiably argued that Thornely's identification was incorrect and recognized her hydroid as a new species, O. thornelyi.Clytia latitheca has also been regarded as a junior synonym of C. stolonifera Blackburn, 1938, an Australian species (Watson 2005).The senior synonym and valid name of the species is therefore O. thornelyi, here assigned to Clytia.While the specific name should have been spelled thornelyae, in recognition of Laura Roscoe Thornely (1860Thornely ( -1951)), the original spelling is to be maintained (ICZN Art. 32.3;Brandon-Jones et al. 2007).
Clytia thornelyi has not been reported before from the Galápagos.Although still a relatively obscure species, it appears to be widespread, with

Figure 3g
Sertularia dichotoma Linnaeus 1758: 812.Obelia dichotoma.-Fraser 1938a: 36;1938b: 133. -Calder et al. 2003 Remarks.-Itseems almost certain that a species complex exists under the binomen Obelia dichotoma (Linnaeus, 1758).With a type locality on the coast of southwest England (Cornelius 1975), the species has long been considered abundant in many areas and virtually cosmopolitan in distribution.It has been reported across the tropical and temperate Indo-Pacific region (e.g.Fraser 1937;Millard 1975;Cooke 1977;Hirohito 1995;Vervoort and Watson 2003;Watson 2005;Xu et al. 2014b), including the Galápagos (Fraser 1938a, b;Calder et al. 2003;Hickman 2008).Affinities of local populations remain to be determined, along with their origin.Of note, Obelia dichotoma is considered a common component of marine fouling assemblages (Woods Hole Oceanographic Institution 1952;Millard 1959;Standing 1976;Carlton and Eldredge 2009;Karlson and Osman 2012), having been reported on various substrates including test panels, buoys, water pipes, and hulls of ships.Long-range transport by shipping is therefore likely.It may thus be that open ocean, sublittoral populations of what are now called O. dichotoma in the tropical eastern Pacific (including the Galápagos) will prove to be an endemic species, whereas O. dichotoma found in ports and harbors in the same region will be found to be members of a global "harbor clade", whose origin may, after centuries of global shipping, be difficult to determine.Pending global genetic dissection of this species, we regard O. dichotoma in the islands as cryptogenic.Stechow, 1914 (type locality: St. Thomas, Virgin Islands), have been reported before from the Galápagos Islands.Elsewhere in the eastern Pacific, hydroids similar to these two species have been reported to the east from Isla Perico, Gulf of Panama (Clarke, 1907, as Obelia (?) sp.), to the north in San Francisco Bay, California (Fraser 1937, as O. bicuspidata;Mills et al. 2007, as O. bidentata), and to the south in the fjords region of southern Chile (Galea et al. 2009, as O. bidentata).Gibbons and Ryland (1989) remarked that hydroids from Fiji resembled O. oxydentata (as illustrated by Hirohito 1969), but assigned them instead to O. bidentata following Cornelius (1975).Gravier-Bonnet (1999) applied the binomen O. oxydentata to similar hydroids from seagrass beds in Madagascar.Specimens from the Galápagos are much the same as those from Fiji and Madagascar, although gonothecae differ somewhat in tapering distally rather than having a distinct terminal collar.

Obelia oxydentata Stechow, 1914
Over the last four decades, Obelia oxydentata has generally been taken to be a synonym of O. bidentata (e.g.Cornelius 1975;Calder 1991a; WoRMS), a supposedly cosmopolitan species (Medel and Vervoort 2000;Peña Cantero and Gili 2006).Under that broad concept, however, O. bidentata likely comprises a species complex (see Calder 2013Calder , 2017)).If hydroids identified as O. bidentata across the warm Indo-Pacific region are indeed identical with those of Clark (1875) from the cool-temperate western North Atlantic, genetic confirmation is needed.Morphologically, Clark's robust specimens were as much as 15 cm high and had polysiphonic stems, unlike the diminutive (< 1 cm high), monosiphonic colonies examined here.Instead, our small specimens closely resemble accounts of O. oxydentata from warm waters of the Caribbean (Stechow 1914) and Florida (Calder 2013).Stechow (1914)  Unusual even in an era when global distributions of species were widely accepted, Stechow was struck by finding the same hydroid species on both sides of the Isthmus of Panama: "Die Auffindung dieser pacifischen Art nunmehr hier auf der atlantischen Seite Amerikas ist außerordentlich bemerkenswert" (The discovery of this Pacific species now, here on the Atlantic side of America, is extraordinarily remarkable).
Obelia oxydentata is therefore known from the warm western Atlantic (Stechow 1914;Calder 2013), the Indian Ocean (Gravier-Bonnet 1999) and, very likely, Fiji (Gibbons and Ryland 1989).Until now, the sole eastern Pacific record consisted of Clarke's 1904Clarke's -1905 material from Isla Perico, an important Pacific Pánama port of the era.Once clearly distinguished morphologically from O. bidentata, records of O. oxydentata are likely to expand, perhaps significantly.At least some records of O. bidentata or another of its many synonyms, O. bicuspidata Clark, 1875, from tropical and warm temperate regions of the world, based on small, monosiphonic hydroids (e.g.Millard and Bouillon 1973;Gibbons and Ryland 1989;Calder 1991a, b;Migotto 1996;Kirkendale and Calder 2003;Calder and Kirkendale 2005;Carlton and Eldredge 2009;Galea 2010), may represent O. oxydentata.
While Obelia oxydentata has almost certainly been introduced to the Galápagos, the source of eastern Pacific populations is unclear, as the species may originate in either the tropical Indo-west Pacific or in the tropical western Atlantic.Once many Obelia species can be assessed phylogenetically, we predict that O. oxydentata will group with species endemic to one of these two regions.
Order Macrocolonia Leclère, Schuchert, Cruaud, Couloux andManuel, 2009 Family Haleciidae Hincks, 1868 Halecium labiatum Billard, 1933 Figure 4a, b Halecium labiatum Billard, 1933: 21, figs. 8K-M.Material.-Specimenscollected April 2016 from fouling plates deployed at the Isla Baltra navy dock, 15 January 2016, coll.K. Larson: 3 colony fragments, up to 4 mm high, without gonophores, vial 232885; 7 colony fragments, up to 4 mm high, with female gonophores, vial 234225; 4 colony fragments, up to 7 mm high, without gonophores, specimens dry at some point, vial 234190; 1 colony, 6 mm high, without gonophores, vial 232895; 6 colony fragments, up to 5 mm high, with female gonophores, vial 232882; 2 colony fragments, up to 6.5 mm high, without gonophores, vial 232875; 8 colony fragments, up to 9 mm high, with female gonophores, vial 232848; 3 colony fragments, up to 6 mm high, with female gonophores, vial 234201; 6 colony fragments, up to 5 mm high, without gonophores, vial 234222.Remarks.-Inbeing unfamiliar to us at first, these tiny specimens were compared with accounts of all 140 species currently recognized in the genus Halecium Oken, 1815.They are closest morphologically to a group of about 40 species, including the type species H. halecinum (Linnaeus, 1758), having female gonothecae of a fundamentally similar form (symmetry bilateral; outline irregularly obovate in lateral view; aperture lateral rather than terminal, with one or more gonophoral hydranths).Within that group, characters of both trophosome and gonosome conformed closest with those of H. labiatum Billard, 1933.Halecium labiatum is very poorly known.The species was originally described by Billard (1933) from material collected at a depth of 36-55 m in the northern Gulf of Suez, Egypt.Subsequent published reports have been from the Dahlak Archipelago in the southern Red Sea (Vervoort 1967), the Gulf of Aden (Rees and Vervoort 1987), and (to where it may also be introduced) Martinique, French Lesser Antilles (Galea and Ferry 2015).Although seldom reported, it seems probable that this small hydroid has on occasion been overlooked, misidentified, identified only to genus, or simply ignored.At least some records of H. lankesterii (Bourne, 1890) are likely to have been based on H. labiatum (in particular, see Millard 1968Millard , 1975)).Misidentifications of Bourne's species have been noted earlier (Peña Cantero and García Carrascosa 2002), although some reports of it from the eastern Atlantic region were likely based on a species other than H. labiatum.Halecium lankesterii is notable in being one of only a few species of the genus having symbiotic zooxanthellae (Peña Cantero and García Carrascosa 2002).Such symbionts are absent in H. labiatum.
Colonies of H. labiatum are small, with few distinguishing characters beyond the singular shape of their female gonothecae.As described best by Galea and Ferry (2015), these structures are saccular in form, with a bulging convex wall, a slightly sigmoid concave wall, and there provided with a lateral aperture with a pair of adjoining hydrothecae towards the distal end, a prominent lip protruding over the aperture, and two defensive hydranths.Gonothecae often occur in pairs, as was usually the case in our specimens.Material examined by Galea and Ferry had 1-2 large eggs per gonotheca, while the usual number in ours was three.Curiously, all fertile colonies examined here were female.
This species is recorded for the first time anywhere in the Pacific Ocean.Specimens were quite frequent, and well-developed, on fouling plates immersed at the Baltra navy dock.Transport to the Galápagos by shipping therefore seems probable, given the location, but the source of these populations is obscure.
The species has been reported elsewhere in both the Indo-west Pacific region (e.g.Pennycuik 1959, as Halecium sessile Norman, 1867;Gravier-Bonnet and Migotto 2000;Kirkendale and Calder 2003;Galea et al. 2012) and the warm western Atlantic (e.g.Calder 1991a;Migotto 1996;Galea 2008;Oliveira et al. 2016).A second species of Nemalecium Bouillon, 1986, N. gracile, was recently described from the Caribbean region (Galea et al. 2012).It is uncertain how many records of N. lighti from the Atlantic might be referable to N. gracile instead.Meanwhile, Gravier-Bonnet and Migotto (2000) and Galea et al. (2012) have noted that additional species of Nemalecium may exist in the Indo-Pacific.Colonies examined here were particularly noteworthy for the great length and slenderness of their gonothecae.Some of them measured more than 2.5 mm from base to apex, while being less than 0.5 mm in diameter (Figure 4d).The largest ones included in a table by Galea et al. 2012 (those of N. cf.lighti from Brazil, described by Migotto 1996), were 1.9 mm long and 0.8 mm in diameter.Additional taxonomic investigations on the population in the Galápagos are warranted.We tentatively assign an introduced status to N. lighti, with possible origins in either the Indo-west Pacific or the western Atlantic.
Although traditionally assigned to Haleciidae, and so classified here, Nemalecium lighti now appears misplaced in that family.Evidence from molecular data indicates that the species is not a haleciid, and Maronna et al. (2016) included it in a new leptothecate suborder, Plumupheniida.As noted by them, research is needed to better resolve its systematic position.
Originally reported from pelagic Sargassum and other floating objects in the Atlantic Ocean (Lamouroux 1816), Amphisbetia distans is believed to be circumglobal in distribution.A species with distinctively tiny and slender hydrothecae, it appears to be common in the Galápagos.To date, A. distans has been reported from the following islands in the archipelago: Wolf (Fraser 1938a, as Sertularia stookeyi;Calder et al. 2003, as Tridentata distans), Isabela (Fraser 1938a, as S. stookeyi), Seymour Norte (Fraser 1948, as S. stookeyi), San Cristóbal (Fraser 1948, as S. stookeyi), and Bartolomé (this study).Moura et al. (2011) suggested that cryptic species may exist within Sertularia distans (=A.distans), at least in the Atlantic Ocean.
Originally discovered in "Australasia" (Lamouroux 1816), T. turbinata is now considered circumglobal in tropical and subtropical waters.The species has been identified previously from the Galápagos at Isla Tortuga (Calder et al. 2003).Hickman (2008) also reported it from the islands, but did not specify a location.Family Kirchenpaueriidae Stechow, 1921 Ventromma halecioides (Alder, 1859)

Figure 7a, b
Plumularia halecioides Alder 1859: 353, pl. 12, figs. 1-5. Plumularia inermis.-Fraser 1938a: 64, p. 15, figs. 74a, b. Ventromma halecioides.-Calder et al. 2003 1859) from the northeast coast of England, opinions currently differ over the generic assignment of this species.Although no longer considered a species of Plumularia Lamarck, 1816, it has been combined in some works with Kirchenpaueria Jickeli, 1883, and in others with Ventromma Stechow, 1923.Morphological differences between the two genera are minor, with nematophores being naked in Kirchenpaueria and protected by nematothecae in Ventromma.More conclusively, molecular studies (Leclère et al. 2007;Moura et al. 2008;Peña Cantero et al. 2010;Maronna et al. 2016) have demonstrated that P. halecioides, the type species of Ventromma, is highly divergent genetically from Kirchenpaueria pinnata (Linnaeus, 1758), type species of Kirchenpaueria.Support for recognition of the genus Ventromma is most recently provided by 16S sequence data showing Kirchenpaueria halecioides forming a sister clade to other taxa of Kirchenpaueriidae (Moura et al. 2018).We therefore adopt Ventromma halecioides as the valid name of the species.
Although confirmation is needed, hydroids of V. halecioides are currently believed to be circumglobal in warm-temperate and tropical waters (Calder 2013).The species has been reported earlier from the Galápagos at Isla Fernandina (Calder et al. 2003).Another record, from Isla Floreana (Fraser 1938a, as Plumularia inermis Nutting, 1900), is likely based on a misidentification.While P. inermis is a junior subjective synonym of V. halecioides, Fraser's material differs in having homomerously segmented hydrocladia.Moreover, gonothecae were elongated and obovate with smooth walls and with a rounded distal end instead of being deep cone-shaped with distinct transverse ridges and with a truncate distal end, as in V. halecioides.Clarke (1907) identified a hydroid from bathyal waters (1063 m) off the Pacific coast of Pánama as Plumularia helleri Hincks, 1872, a name now included in the synonymy of Kirchenpaueria pinnata (Linnaeus, 1758).Clarke's illustration shows that his species, a kirchenpaueriid with nematothecae, is referable instead to Ventromma, but not to V. halecioides or any other known species of the genus.While its hydrocladia are heteromerously segmented, the adcauline wall of the hydrotheca is about one-third free (Clarke 1907: pl. 15, figs. 1-6).This deep-water species, deemed herein to be new, warrants redescription and naming.Material to serve as a holotype exists at the NMNH (USNM 29620, as Plumularia helleri, R/V Albatross Station Number 4622, United States Fish Commission Eastern Pacific Expedition, centroid latitude 6.5167, centroid longitude −81.7333, south of Coiba Island, Pánama, 21 October 1904).
Ventromma halecioides, cryptogenic in the islands, is frequent in areas of quiet, shallow water (Calder 2013), as with the dock areas sampled in Puerto Ayora.It is known to be a component of certain fouling communities, occurring on ships and buoys (Woods Hole Oceanographic Institution 1952, as Plumularia halecioides) and in ports (Cornelius 1992b).
Halopteris alternata was reported in the Galápagos from Isla Wolf and Isla Floreana by Fraser (1938a, as Plumularia alternata), but Schuchert (1997) believed that those records were based on a different, undetermined species.Unlike the spindle-shaped gonothecae of H. alternata, as described by Schuchert, those of colonies described by Fraser were in the shape of curved cornucopiae.
Originally described from the Bahamas (Nutting 1900), the range of H. alternata extends from North Carolina and Bermuda to Brazil in the western Atlantic, and from the vicinity of Madeira, the Canary Islands, and the Cape Verde Islands in the eastern Atlantic (Ansín Agís et al. 2001).It is reported here as a probable introduction for the first time from the Galápagos Islands.Earlier records of the species from various locations across the Pacific Ocean are considered doubtful (Schuchert 1997).H. alternata is widespread in the Caribbean (Calder and Kirkendale 2005)

Discussion
Constraining the detection of introduced hydroids in regions such as the Galápagos Islands is the reported natural cosmopolitan, and especially tropicopolitan, distribution of a great many species of shallow-water hydroids, most or all of which, if not transported around the world by centuries of shipping, may prove to be species complexes.This said, for many marine taxa, the Central American Land Barrier has served to isolate Caribbean and eastern Pacific biotas for millions of years (James 1991;Hickman 2009), making it unlikely that the same species now naturally occur on either side of the Isthmus of Panama.Global shipping since the 1500s has, however, served to disperse many marine invertebrates, algae, and fishes across and between oceans (Carlton 2009) such that the appearance of, for example, Caribbean or western Pacific hydroid species in the Galápagos Islands would not be unexpected.
Regionally, the Panama Canal may have served as an important corridor since 1914 for Caribbean hydroids entering the tropical eastern Pacific Ocean (Cohen 2006).Although the freshwater Gatun Lake would appear to be a barrier to the dispersal of marine species, shipping has nevertheless likely transported many hydroid species from the Atlantic to the Pacific (and vice-versa).Dormant stages exist in their life cycles that could survive temporarily unfavourable environmental conditions encountered over the short distance and travel time required in crossing through Pánama.Perhaps more important, however, is the presence of both medusae and polyp stages of hydrozoans in ships' ballast water (Carlton and Geller 1993;Smith et al. 1999, J. T. Carlton, personal observations), such that species transported inside vessels would pass imperviously through the Panama Canal and be released on the Central American Pacific coast.
We conservatively recognize here eight introduced and five cryptogenic species of hydroids (Table 1).The introduced species are recognized through their remarkable disjunct distributions, making natural distribution and dispersal unlikely.The cryptogenic species are primarily so designated because they may be members of world-wide species complexes, within which endemic tropical eastern Pacific species may reside.To the species treated here, we add one introduced species reported earlier (Calder et al. 2003) of either Caribbean or Indo-west Pacific origin, Clytia hummelincki (Leloup, 1935), and one cryptogenic species of unknown origin, Pennaria disticha Goldfuss, 1820.Calder et al. (2003) recorded additional species of hydroids in the open ocean, sublittoral Galápagos fauna that were previously known either from the Atlantic or the western Pacific, but whose biogeography and taxonomy require further investigation.
Of the introduced species, two (Bougainvillia muscus and Halopteris alternata) may be from the Atlantic Ocean, and two from the broad Indo-Pacific theater (Halecium labiatum and Clytia elongata).The remaining four (Clytia thornelyi, C. hummelincki, Obelia oxydentata, and Nemalecium lighti) may originate from either the warm waters of the western Atlantic or the western Pacific.Genetic analyses seeking centers of haplotype diversity, phylogenetic analyses grouping sibling species, or both, may eventually point to the origins of these species.No introduced species known only from the tropical eastern Atlantic have yet been detected in the Galápagos.
The 2016 dates of first collection of seven of the hydroid species newly added here as introductions or cryptogens to the Galápagos marine fauna do not likely represent new invasions; that is, these records should not be interpreted as part of a new wave of invasions now occurring in the archipelago.Rather, there have been few historical investigations of the biofouling communities in Galápagos ports and harbors (Carlton et al. 2019).While the present investigations (Carlton et al. 2019) thus form a baseline of the diversity of these communities, we do not know when these species first arrived.Given that vessels have been visiting the Galápagos since the 1500s, we have little doubt that with further study, and with resolution of the biogeography of many other species, the list of non-native hydroids will grow considerably.
Finally, we note that the present 2015 and 2016 collections took place during a strong El Niño episode (Stramma et al. 2016) in the equatorial eastern Pacific.With no seasonal or long-term data available on the diversity and distribution of hydroids in the Galápagos Archipelago, it is difficult to interpret what effect these unusually warm seawater temperatures may have had on the fauna.It may be that future studies will be able to place the current collections in a broader temporal framework.
scattered records from the tropical and subtropical western Atlantic (e.g.Calder and Kirkendale 2005, as C. stolonifera; Oliveira et al. 2016, as  C. stolonifera), the west and central Pacific (e.g.Thornely 1900, as Obelia serrulata;Nutting 1927, as O. thornelyi; Kirkendale and Calder 2003, as  C. latitheca; Di Camillo et al. 2008, as C. latitheca; Carlton and Eldredge 2009,  as C. latitheca)  and Indian (e.g.Millard and Bouillon 1973, as C. latitheca)   oceans.We suggest that this species is a likely introduction to the Galápagos from either the western Pacific or the western Atlantic Ocean.

Figure
Figure 3h, iObelia oxydentataStechow 1914: 131, fig. 7.    Material.-IslaSanta Cruz, Tortuga Bay, epizoic on bryozoan Amathia verticillata, 21 February 2015, 3 colony fragments, up to 6 cm high, with gonothecae, coll.J. T. Carlton.-Isla Santa Cruz, Puerto Ayora, Franklin's Bay, fouling community on floats, 1 colony, 9 mm high, with gonothecae, 27 February 2015, coll.J. T. Carlton.-Isla Santa Cruz, Puerto Ayora, Franklin's Bay, 30 April 2016, 1 colony fragment, 5 mm high, without gonophores, coll.K. Collins, JJM Hyd.19.Remarks.-Nohydroids such as these, resembling the widely reported Obelia bidentata Clark, 1875 (type locality: Long Island Sound, USA) and especially the much less familiar O. oxydentataStechow, 1914 (type locality: St. Thomas, Virgin Islands), have been reported before from the Galápagos Islands.Elsewhere in the eastern Pacific, hydroids similar to these two species have been reported to the east from Isla Perico, Gulf of Panama(Clarke, 1907, as Obelia (?) sp.), to the north in San Francisco Bay, California(Fraser 1937, as O. bicuspidata; Mills et al. 2007, as O. bidentata), and to the south in the fjords region of southern Chile(Galea et al. 2009, as  O. bidentata).Gibbons and Ryland (1989) remarked that hydroids from Fiji resembled O. oxydentata (as illustrated by Hirohito 1969), but assigned them instead to O. bidentata followingCornelius (1975).Gravier-Bonnet (1999) applied the binomen O. oxydentata to similar hydroids from seagrass beds in Madagascar.Specimens from the Galápagos are much the same as those from Fiji and Madagascar, although gonothecae differ somewhat in tapering distally rather than having a distinct terminal collar.Over the last four decades, Obelia oxydentata has generally been taken to be a synonym of O. bidentata (e.g.Cornelius 1975;Calder 1991a; WoRMS), a supposedly cosmopolitan species(Medel and Vervoort 2000;Peña Cantero and Gili 2006).Under that broad concept, however, O. bidentata likely comprises a species complex (seeCalder 2013Calder , 2017)).If hydroids identified as O. bidentata across the warm Indo-Pacific region are indeed identical with those of Clark (1875) from the cool-temperate western North Atlantic, genetic confirmation is needed.Morphologically, Clark's robust specimens were as much as 15 cm high and had polysiphonic stems, unlike the diminutive (< 1 cm high), monosiphonic colonies examined here.Instead, our small specimens closely resemble accounts of O. oxydentata from warm waters of the Caribbean(Stechow 1914) and Florida(Calder 2013).Stechow (1914) included Clarke's (1907) Obelia (?) sp.collected in 1904-1905 (before the 1914 opening of the Panama Canal) from the Gulf of Panama, tropical eastern Pacific, in the synonymy of his new species O. oxydentata.Unusual even in an era when global distributions of species were widely accepted, Stechow was struck by finding the same hydroid species on both included Clarke's (1907) Obelia (?) sp.collected in 1904-1905 (before the 1914 opening of the Panama Canal) from the Gulf of Panama, tropical eastern Pacific, in the synonymy of his new species O. oxydentata.
and may have been transported into the eastern Pacific from the Caribbean region via the Panama Canal.
regard it as an introduction to the northeast Pacific Ocean and to the Hawaiian (Gonzalez-Duarte et al. 2016)s demonstrating an invasion propensity(Gonzalez-Duarte et al. 2016), having arrived in the Mediterranean in the 1990s; records since the 1980s from the Indo-west Pacific (summarized in Gonzalez-Duarte et al.