Lovenella gracilis Clarke, 1882

Calder, Dale R., 2019, On a collection of hydroids (Cnidaria, Hydrozoa) from the southwest coast of Florida, USA, Zootaxa 4689 (1), pp. 1-141 : 36-40

Lovenella gracilis Clarke, 1882

Figs. 9 a–g

Lovenella gracilis Clarke, 1882: 139 , pl. 9, figs. 25–39.— Shier, 1965: 42 (in part).

not Lovenella gracilis .— Joyce, 1961: 61, pl. 14, figs. 2, 3.— Shier, 1965: 42 (in part), pl. 23 [= Lovenella grandis Nutting, 1901 ].

Type locality. USA: Chesapeake Bay , 3–10 ftm (5–18 m) ( Clarke 1882: 135) .

Material examined. Sanibel Island, beach at Lighthouse Point, on detached Thalassia in water along shore, 13 December 2017, one colony, 4 mm high, with gonothecae, coll. D. Calder, ROMIZ B4351.— Fort Myers Beach , on dead sand dollars ( Mellita quinquiesperforata ) near low water, 19 January 2018, several colony fragments, up to 5 mm high, without gonothecae, coll. D. Calder, ROMIZ B4352 .— Sanibel Island , beach at Lighthouse Point, 26°26’55”N, 82°01’08”W, on detached Syringodium in water along shore, 21° C, 34.5‰, 19 March 2018, one colony, 3 mm high, without gonothecae, coll. D. Calder, ROMIZ B4353 .

Non-Florida material examined. USA: Virginia, York River entrance off Ellen Island , 37°15’N, 76°25’W, 06 August 1966, several colonies, with gonophores, coll. D. Calder, ROMIZ B1548 .— USA: South Carolina, Folly Beach , 18 November 1973, intertidal, on Mulinia lateralis and adhering barnacles, coll. D. Calder, ROMIZ B1551 .— USA: New Jersey, Avalon , 01 September 2001, on Donax , coll. J. Dougherty, ROMIZ B3504 .— USA: South Carolina, Bulls Bay , 32°55.9’N, 79°36.2’W, 5 m, 06 January 1976, coll. D. Calder, ROMIZ B1546 ( Lovenella sp.).

Remarks. Clarke (1882) described both hydroid and young medusa stages of this species from the Chesapeake Bay region, USA. While assigning the binomen Lovenella gracilis to it, he expressed uncertainty about its relationships and systematic position. Such questions still need attention. Given its resemblance to L. clausa ( Lovén, 1836) from northwest Europe, including habit of growth, turbinate hydrothecae, proboscis morphology, and number of tentacles and opercular facets, Clarke decided to refer his new species to Lovenella Hincks, 1868 [1869] rather than create a new genus for it. While recognizing certain similarities between it and L. clausa , he regarded the two as distinct. Clarke’s illustrations show that his hydroids differ in having: (1) a hydrocaulus that is segmented beyond the base by single nodes into a series of cylindrical internodes rather than being annulated or wrinkled; (2) hydrothecae that are relatively shallow and funnel-shaped rather than deep and nearly cylindrical; (3) an operculum that is not distinctly demarcated from the hydrotheca. From accounts of the medusa stage of L. clausa by Hincks (1871) and Russell (1936a, as Eucheilota hartlaubi ; 1936b, as E. clausa ; 1953, as L. clausa ), several differences are also immediately apparent between the planktonic stages of the two species: (1) gonads are developed at liberation in L. gracilis , but they do not appear until several days after release in L. clausa ; (2) interradial marginal cirri occur in L. clausa but not in L. gracilis ; (3) the umbrella is oval in cross-section in L. gracilis but circular in L. clausa . The two species are clearly different.

In his original description of L. gracilis, Clarke (1882) did not specify in detail where his specimens were collected. Although his work was done in Chesapeake Bay, Clarke spent research time there at both Crisfield, Maryland, and Fort Wool, Virginia. The type locality is therefore listed here simply as Chesapeake Bay.

Fraser (1912a, b) erroneously reported the European L. clausa from the east coast of the United States, and included L. gracilis as its synonym. In examining specimens from Bogue Sound, North Carolina, parts of his description and two of his figures ( Fraser 1912b, fig. 26B, 26C) were taken from accounts of L. clausa by Hincks (1868 [1869], 1871) and Hartlaub (1897). He later recognized that the two species were distinct ( Fraser 1944: 174), acknowledged that his report of L. clausa from eastern North America was wrong, and stated that the European species had not been observed in the western Atlantic. Regrettably, descriptions and illustrations actually based on those Carolinian hydroids by Fraser (1912b: 364, fig. 26A, 1944: 174, pl. 31, fig. 147b) conform more closely with characters of L. grandis Nutting, 1901 rather than L. gracilis . Hydrothecae of his specimens were much deeper than those of L. gracilis , and his drawings show distinct crease lines separating the opercular facets from the hydrothecal margin. Fraser’s record of the species from North Carolina is therefore assigned to L. grandis here. Specimens referred by Fraser (1912b) to L. clausa from Woods Hole, Massachusetts, were not illustrated and are of uncertain identity. Errors in characterization of L. gracilis in Fraser’s publications lead to confusion about the species in subsequent works on hydroids of this coast. For example, Defenbaugh & Hopkins (1973) considered specimens much like and probably conspecific with L. gracilis from Texas to be a new species, while a clearly different one resembling L. grandis was misassigned to it. Specimens referred to L. gracilis from Seahorse Key, Florida, by Joyce (1961), and part of those from the Cape San Blas area, Florida, by Shier (1965), also appear to have been misidentified. From descriptions of the operculum by Shier (“There are eight opercular sections; the basal hinge of these sections is often indistinct”), however, both species may have been present in her collections. Ones with “indistinct hinges” were likely assigned correctly to L. gracilis . Identifications of the hydroid of L. gracilis should be based on the original description of Clarke (1882), or on later accounts of specimens from the same general locality ( Calder 1971, this work), rather than on information in the guidebook of Fraser (1944). To further document the distinguishing characters of L. gracilis , topotypic specimens from southern Chesapeake Bay region (ROMIZ B1548) were examined during this study and are illustrated here ( Figs. 9e, f ).

Huvé (1952) compared the medusa stages of L. gracilis and Dipleuron parvum Brooks, 1882 from Beaufort, North Carolina, and concluded that they were identical. He thereupon applied the binomen Dipleuron gracilis to the species. His opinion that they were conspecific was shared by me ( Calder 1971), although I retained the name L. gracilis for it. The specific name gracilis of Clarke (1882), published in January of that year, has priority over parvum of Brooks (1882), published in March. In two later works (Calder 1990 [1991a]: 3; Calder & Stephens 1997: 31), Dipleuron Brooks, 1882 was accepted as distinct from Lovenella based on characters such as the unusual morphology of the stems of the hydroid, being segmented by variably spaced nodes rather than annulated, and the lack of a crease at the base of the operculum. Pires-Miranda et al. (2013), in a study of specimens identified as L. gracilis from Brazil, upheld the synonymy of the two genera. Although unresolved issues remain in the taxonomy and nomenclature of this and other lovenellids, current usage of the binomen L. gracilis for the species has been maintained here, following Cairns et al. (1991, 2002).

A long-held belief that Lovenella is closely related to Eucheilota McCrady, 1859 (e.g., Russell 1953; Kramp 1961) is supported, at least in part, by preliminary molecular studies ( Maronna et al. 2016). However, such results also suggest that both Lovenella and Eucheilota as presently constituted may be polyphyletic, and diagnoses of these genera likely need revision. Morphological and molecular studies of their type species ( L. clausa and Eucheilota ventricularis McCrady, 1859 ), based on topotypic material, are needed. Critically, no hydroid has yet been linked to E. ventricularis , a major limitation in definitive characterization of Eucheilota and in overall understanding of both genera. Another taxonomic question needing resolution is whether populations of hydroids assigned to L. gracilis from the open coast, especially those attached to shells of the bivalve Donax and other substrates along sandy ocean shores, are conspecific with those from estuarine areas such as the tributaries of Chesapeake Bay.

The medusa known as Dipleuron parvum (= L. gracilis ) has been assigned by some authors to the genus Eucheilota , and to the synonymy of E. duodecimalis L. Agassiz, 1862 ( Mayer, 1910b, as E. duodecimalis var. par- vum; Allwein 1967, as E. duodecimalis ). However, life cycle studies ( Calder 1971) have shown that the number of marginal vesicles in medusae of L. gracilis is indefinite, as currently stated in diagnoses of Lovenella , rather than fixed, as in Eucheilota ( Russell 1953; Kramp 1961; Bouillon et al. 2006). For that reason, conclusions that the species is fully conspecific with or constitutes only a variety of E. duodecimalis has not been adopted here. Allwein (1967: 127) reported finding both E. duodecimalis and E. duodecimalis var. parvum (= L. gracilis ) in the plankton at Beaufort, North Carolina.

A number of hydroids from other geographic regions resemble L. gracilis , particularly in the peculiar segmentation of the hydrocaulus. One of these, Clytia bakeri Torrey, 1904 , was found on bivalves in the surf zone at Pacific Beach near San Diego, California. Its medusa stage, described by Torrey (1909) as Phialium bakeri , is much like that of L. gracilis . No operculum was noted in the original description of the hydroid by Torrey (1904), but the species cannot be assigned to Clytia Lamouroux, 1812 or any other campanularioid genus given the characters of its medusa stage. Hydrothecal margins were said to have been damaged, and Huvé (1952) and Calder (1971) speculated that opercula may have been lost in Torrey’s material. The species was assigned to Eucheilota by Cairns et al. (1991), but returned to Lovenella by Cairns et al. (2002). It was discussed as Eucheilota bakeri in Mills et al. (2007) . Molecular studies reveal a particularly close genetic relationship of the species to hydroids identified as L. gracilis ( Maronna et al. 2016) , and the two are almost certainly congeneric. Another hydroid resembling L. gracilis was described as Gonothyraea (?) nodosa by Stechow (1914) from Rio de Janeiro, Brazil. Pires-Miranda et al. (2013) were uncertain of its taxonomic status, but Oliveira et al. (2016) included it in the synonymy of L. gracilis . Meanwhile, L. nodosa Fraser, 1938a , originally described from Ecuador ( Santa Elena Bay) and the Pacific coast of Mexico (off Morro de Petatlan, Tenacatita Point; Isabel Island; off Thurloe Point), differs from L. gracilis in having hydrocauli with fewer nodes and much deeper hydrothecae. Lovenella corrugata Thornely, 1908 , from 20 fathoms (37 m) off Khor Shin‘ab, along the Red Sea coast of Sudan, has a hydrocaulus with neither annulations nor nodes. Its hydrothecae are distinct in being relatively deep and cylindrical, with corrugated walls.

Several other species with hydroid stages currently or recently assigned to Lovenella differ from L. gracilis in having hydrocauli with annulations or sinuous constrictions rather than cylindrical internodes [ L. clausa ( Lovén, 1836) from NW Europe; L. briggsi Mulder & Trebilcock, 1915 from Australia; L. rugosa Fraser 1938b from the Pacific coast of Mexico; L. chiquitita Millard, 1957 from South Africa] or much deeper hydrothecae ( L. grandis Nutting, 1901 ). Two species of hydroids until recently included in Lovenella by some authors are now assigned to other genera after their polyp and medusa stages were linked in DNA barcoding studies by Schuchert et al. (2017). Firstly, the hydroid L. panicula (G.O. Sars, 1874) and the medusa Foersteria quadrata Hosia & Pagès, 2007 were found to be stages of the same species. The name Earleria quadrata ( Hosia & Pagès, 2007) was applied to the species by Schuchert et al., but the binomen Earleria panicula (G.O. Sars, 1874) has nomenclatural priority. Secondly, the hydroid Lovenella producta (G.O. Sars, 1874) and the medusa Cyclocanna welshi Bigelow, 1918 were shown to be conspecific, with Cyclocanna producta becoming the valid name of the species. Hydroids of both species differ significantly from those of L. gracilis in the morphology of their hydrothecae and hydrocauli. The taxonomy of hydroid and medusa species that have been assigned to Lovenella has been updated in WoRMS, and the characters of most have been summarized in Pires-Miranda et al. (2013).

The reported range of L. gracilis in the western North Atlantic extends from southern New England ( Calder 1975) to the southern Caribbean Sea ( Wedler 1975; Bandel & Wedler 1987), and includes the Gulf of Mexico ( Calder & Cairns 2009). It has also been identified as far south as Brazil in the western South Atlantic ( Oliveira et al. 2016). Hydroids identified as Lovenella gracilis from the Caribbean Sea by Wedler (2017b) are shown with distinct crease lines at the base of the opercula, and his specimens appear more like L. grandis in that character. Their identity is regarded here as doubtful, bringing those reported by Wedler (1975) and Bandel & Wedler (1987) into question as well.

Fewkes (1891) included L. gracilis in a guide to invertebrates of New England. No collection data were provided on the species in that work, and it is unclear whether specimens were ever collected by him in the New England region or elsewhere. The account has been excluded from the distribution records below. Lovenella gracilis was reported from a depth of 16 fathoms (29 m) off Cape Hatteras by Fraser (1944), but the identification is considered questionable. To the south in South Carolina (Calder, unpublished), hydroids largely indistinguishable from L. gracilis were found in the lower intertidal zone of sandy beaches on bivalve molluscs ( Mulinia lateralis ) and adhering barnacles (ROMIZ B1551). Farther north, specimens of the species were found on coquina clams ( Donax variabilis ) at Avalon, New Jersey ( Fig. 9g ). Other specimens (ROMIZ B1546) generally resembling the species, from Bulls Bay, South Carolina (32°55.9’N, 79°36.2’W), were unusual in being exceptionally large (up to 11 cm high) and profusely branched ( Fig. 9h ). The identity of the latter hydroids as a possible new species needs to be established. During studies of hydroids from the Chesapeake Bay region, the type locality of L. gracilis , colonies of the species were reported in shallow waters (1-6 m) from a variety of substrates including algae ( Agardhiella tenera ), eelgrass ( Zostera marina ), other hydroids ( Sertularia argentea ), oyster shells ( Crassostrea virginica ), and slipper shells ( Crepidula fornicata ) ( Calder 1971). In southwest Florida during this investigation, specimens were found on turtle grass ( Thalassia testudinum ), manatee grass ( Syringodium filiforme ) and dead sand dollars ( Mellita quinquiesperforata ). Hydroids of the species were found active in the Chesapeake region from late April through late October over a temperature range of 15-27° C, and medusa production was observed from July through October. Interactions between hydroids identified as L. gracilis and bivalve molluscs, and especially coquina clams ( Donax ) on sandy beaches, have been explored in studies such as those of Manning and Lindquist (2003) and Dougherty & Russell (2005).

Reported distribution. Gulf coast of Florida. Cape San Blas area ( Shier 1965: 42 (part)) .

Elsewhere in western North Atlantic. USA: Chesapeake Bay, 3–10 ftm (5–18 m) ( Clarke 1882).— USA: North Carolina, Beaufort (medusa) ( Brooks 1882: 140, as Dipleuron parvum ).— USA: North Carolina, off Cape Fear (medusa) ( Mayer 1910b: 284, as Eucheilota duodecimalis var. parvum ).—? USA: Massachusetts, Woods Hole, surface tow ( Fraser 1912a: 45, as Lovenella clausa ).— USA: Mississippi, off Horn Island, on clam shell ( Fincher 1955: 92).— USA: North Carolina, Beaufort area (medusa) ( Allwein 1967: 127, as Eucheilota duodecimalis (part)).— USA: Virginia, York River (Ellen Island, Perrin, Gloucester Point, Page’s Rock, Bell Rock) + James River (Hampton Flats) ( Calder 1971: 62).— USA: Delaware, Indian River Bay ( Watling & Maurer 1972: 648).— USA: Texas, off Galveston, 3–6 ftm (5–11 m) ( Defenbaugh & Hopkins 1973: 92, as Lovenella new species).— USA: Massachusetts, Cape Cod Bay, off Provincetown + Jeremy Point, Wellfleet ( Calder 1975: 298).—? Colombia: Santa Marta area, on gastropod shells, sandy substrates ( Wedler 1975: 336; Bandel & Wedler 1987: 41).— USA: Delaware, Indian River ( Maurer 1977: 597).— USA: South Carolina, exposed beaches (on shells of Donax and Mulinia ) + higher salinity regions of estuaries (on shells of Crassostrea ) ( Calder & Hester 1978: 90, as (?) Lovenella gracilis ).— USA: South Carolina, sandy beaches near Murrells Inlet, on Donax ( Knott et al. 1983: 585) .— USA: South Carolina, Huntington Beach ( Fox & Ruppert 1985: 38).— USA: Georgia: St. Catherines Island, Engineer Point + Middle Beach + North Beach + Picnic Point + Seaside Delta ( Prezant et al. 2002: 8).— USA: New Jersey, Seaside Park, sandy beach, on Donax ( Manning & Lindquist 2003: 416) .— USA: North Carolina, Pine Knoll Shores, sandy beaches, on Donax ( Manning & Lindquist 2003: 416) .— USA: New Jersey, sandy beaches, on Donax, Sea Isle City , 39°12.9’N, 74°70.7’W + Avalon, 39°07.3’N, 74°74.0’W + museum specimens: Atlantic City (from ~1886) + Wildwood (from 1900) + Sea Isle City (from 1901) + Avalon (from 1962) ( Dougherty & Russell 2005: 35, 37).— USA: Delaware, on Donax , museum specimens Indian River Inlet (from 1965) ( Dougherty & Russell 2005: 35, 37).— USA: New Jersey, Wildwood Crest ( Govindarajan et al. 2006: 824).—?Caribbean Sea ( Wedler 2017b: 119, figs. 114A–D).

Figures

FIGURE 9. a, Lovenella gracilis: part of colony with a gonotheca, Sanibel Island, ROMIZ B4351. Scale equals 0.2 mm. b, Lovenella gracilis: colony with two hydrothecae, Sanibel Island, ROMIZ B4353. Scale equals 0.3 mm. c, Lovenella gracilis: pedicel and hydrotheca, Sanibel Island, ROMIZ B4351. Scale equals 0.2 mm. d, Lovenella gracilis: hydrotheca and part of pedicel, Fort Myers Beach, ROMIZ B4352. Scale equals 0.2 mm. e, Lovenella gracilis: part of colony with a hydrotheca, York River, Virginia, USA, ROMIZ B1548. Scale equals 0.2 mm. f, Lovenella gracilis: part of colony with a gonotheca, York River, Virginia, USA, ROMIZ B1548. Scale equals 0.2 mm. g, Lovenella gracilis: part of colony with a hydrotheca, on Donax, New Jersey, USA, ROMIZ B3504. Scale equals 0.2 mm. h, Lovenella sp.: part of colony with a hydrotheca, Bulls Bay, South Carolina, USA, ROMIZ B1546. Scale equals 0.2 mm. i, Lovenella grandis: hydrotheca and distal end of pedicel, Fort Myers Beach, ROMIZ B4354. Scale equals 0.2 mm. j, Lovenella grandis: part of colony with a hydrotheca, Murrells Inlet, South Carolina, USA, ROMIZ B1549. Scale equals 0.2 mm. k, Lovenella grandis: hydrotheca and distal end of pedicel, Nantucket Sound, Massachusetts, USA, ROMIZ B3500. Scale equals 0.2 mm.