Two New Brittle Star Species of the Genus Ophiothrix

Home , Comactinia, Ophiocoma echinata, Ophionereis reticulata, Ophiothrix, Ophiothrix angulata, Ophiothrix suensoni

Two New Brittle Star Species of the Genus Ophiothrix (Echinodermata: Ophiuroidea: Ophiotrichidae) from Coral Reefs in the Southern Caribbean Sea, with Notes on Their Biology

GORDON HENDLER

Natural History Museum of Los Angeles County, 900 Exposition Boulevard, Los Angeles, California 90007, U.S.A. [email protected]

ABSTRACT.—Two new species, Ophiothrix stri and Ophiothrix cimar, inhabit shallow reef-platforms and slopes in the Southern Caribbean, and occur together at localities in Costa Rica and Panama, nearly to Colombia. What appears to be an undescribed species resembling O. cimar has been reported from eastern Venezuela. In recent years, reefs where the species were previously observed have deteriorated because of environmental degradation. As a consequence, populations of the new species may have been reduced or eradicated. The new species have previously been mistaken for O. angulata, O. brachyactis, and O. lineata. Ophiothrix lineata, O. stri, and O. cimar have in common a suite of morphological features pointing to their systematic affinity, and a similar pigmentation pattern consisting of a thin, dark, medial arm stripe flanked by two pale stripes. Ophiothrix lineata is similar to Indo-Pacific members of the subgenus Placophiothrix and closely resembles Ophiothrix stri. The latter is extremely similar to O. synoecina, from Colombia, and both can live in association with the rock-boring echinoid Echinometra lucunter. Although O. synoecina is a protandric hermaphrodite that reportedly broods its young externally, the new species are gonochoric and do not brood. Their eggs are of a size indicative of abbreviated larval development, and the gametes of O. stri produce demersal embryos that develop within a large, adhesive fertilization envelope. The new species and related forms may belong to one of several clades containing shallow-water Caribbean Ophiothrix species with abbreviated development and relatively restricted larval dispersal.

KEYWORDS.—Abbreviated demersal development, brooding, Comactinia, Echinometra, Placophiothrix

INTRODUCTION drawn the attention of underwater photog- raphers. Therefore, frustration created by The brittle stars of the family Ophiotri- the unsettled taxonomy of ophiotrichids is chidae are readily distinguished by the felt not only by systematists. stumps, spines, and spinelets on their The confusion engendered by ophiotri- disks, their long, thorny arm spines, and an chid taxonomy is a long-standing problem oral armament consisting solely of dental in the Caribbean region where the common papillae. However, the distinctions among Ophiothrix species are extremely variable, genera within the family are problematical, and the variability of the very rare species and the identification of ophiotrichids has has not been characterized (Hendler et al. been a continuing “source of considerable 1995). Under the circumstances, it is not trouble to systematists owing to the great surprising that the occurrence of some variability of many of the species” (A. M. Ophiothrix species, even accessible wide- Clark 1966:637). Due to their abundance spread species, has gone undetected. In the and ubiquity in shallow, warm-water ma- present contribution, two such new species rine habitats, ophiotrichids in the genus of Ophiothrix are described, which live on Ophiothrix have been studied in nature and coral reefs from Costa Rica to Colombia. the laboratory by biologists in various dis- They were most recently observed and col- ciplines. The brilliant colors of some spe- lected in Bocas del Toro Province, Panama, cies, and their association with attractive at sites that are relatively unaffected by the corals and sponges, have consistently regional deterioration of coral reefs 583 584 GORDON HENDLER

(Guzmán and Guevara 1998a, b, 1999). survey sponsored by the STRI’s Bocas Re- However, it is not know if both species still search Station, and in 1986 and 1987 at survive at severely impacted sites where nearby Costa Rican localities during a they occurred in previous decades. Their study of Cahuita National Park. The results populations may have been reduced or are also derived from specimens of the Al- eradicated during the decline of coral reefs lan Hancock Foundation Pacific Expedi- that is manifested throughout the Carib- tions deposited in the Natural History Mu- bean, largely as a consequence of human- seum of Los Angeles County (LACM). caused environmental change (Gardner et Type specimens of related species from the al. 2003; Birkeland 2004). echinoderm collections of the Museum of In the vicinity of Panamanian reefs near Comparative Zoology of Harvard Univer- Galeta, and in the San Blas Islands where I sity (MCZ) and the National Museum of found both of the new species in the 1970s, Natural History of the Smithsonian Institu- there have been several destructive oil tion (USNM) were compared as well. spills, and declines in coral cover attributed Living and preserved specimens were to coral bleaching, coral mining, and un- examined with a stereomicroscope and planned coastal development (Birkeland et measured with a calibrated ocular mi- al. 1976; Guzmán et al. 1994; Guzmán 2003). crometer, calipers, or millimeter ruler, as At Cahuita National Park, Costa Rica, necessary. Animals were anesthetized us- where I collected them in the 1980s, the reef ing magnesium chloride or Epsom salts so- was subsequently damaged by coastal up- lutions prior to preservation in ethanol, and lift triggered by an earthquake, and by the some specimens were dried. Descriptive impact of siltation and other anthropogenic statistics are based exclusively on alcohol- stressors (Cortés and Risk 1985; Cortésetal. preserved animals. To prepare skeleton 1993; Fonseca 2003). The situation at Cale- structures for scanning electron micros- donia Bay, where the species were first col- copy, soft tissue was removed from alco- lected in the 1930s, has not since been re- hol-preserved material using dilute sodium ported (Garth 1945), but the condition of hypochlorite solution, and the specimen the reefs has probably deteriorated there, as was washed in water and air-dried prior to it has elsewhere in Panama (Guzmán 2003). sputter coating. The terminology employed The similarity of the new species to each for morphological characters follows Hen- other, and to congeners including the dler et al. (1995). Color patterns running the highly variable Ophiothrix angulata (Say, length of the arm are referred to as stripes; 1825), has hampered their recognition and those running across the arm are called study. Therefore, my objective is to facili- bands. Abbreviations used are dd (disk di- tate their identification. For the same rea- ameter) and AL (arm length). Type speci- son, I also refer to what may be another mens of the new species are deposited at similar, undescribed species from Venezu- the LACM, USNM, and Museo de Zoología ela (Zoppi de Roa 1967). A further aim is to of the Universidad de Costa Rica (UCR). report on some significant aspects of their reproduction and ecology, and their rela- SYSTEMATIC ACCOUNT tionship to other species of Ophiothrix. Family Ophiotrichidae Ljungman, 1867 Genus Ophiothrix Müller and Troschel, MATERIALS AND METHODS 1840 Ophiothrix cimar, new species Initial studies and collections were done FIGURE 1A-F in the vicinity of the Galeta Marine Labo- ratory of the Smithsonian Tropical Re- Holotype.—COSTA RICA, LIMON search Institute (STRI), from 1971 to 1975. PROV.: (LACM 1987-68.3), Sta. CR 87-31, 24 However, the present contribution is May 1987, 9°44.97Ј N, 82°49.3Ј W, NE of largely based on observations and collec- Sloth River mouth, 0-1 m, coll. G. Hendler tions gathered in 2003 during a regional and R. W. Peck. NEW SOUTHERN CARIBBEAN BRITTLE STARS 585

Paratypes.—COSTA RICA, LIMON 82°48.7Ј W, Punta Cahuita Reef, 1-2 m, coll. PROV.: (LACM 1986-106.2), 8 alc, Sta. CRA G. Hendler and R. W. Peck; (LACM 1987- 86-1, 26 Jul. 1986, 9°44.18Ј N, 82°48.6Ј W, 68.2), 1 alc, Sta. CR 87-31, 24 May 1987, Punta Cahuita Reef, 1-3 m, coll. G. Hendler; 9°44.97Ј N, 82°49.3Ј W, NE of Sloth River (LACM 1986-115.1), 1 alc, Sta. CRA 86-8, 27 mouth, 0-1 m, coll. G. Hendler and R. W. Jul. 1986, 9°37.9Ј N, 82°36.97Ј W, off Punta Peck. PANAMA, BOCAS DEL TORO Mona (Punta Carreta), 6-15 m, coll. G. Hen- PROV.: (LACM 2003-51.6), 1 alc, Sta. BDT dler; (LACM 1986-137.1), 1 alc, Sta. CRA 03-4, 4 Aug. 2003, 9°20.66Ј N, 82 ° 10.33Ј, 86-23, 28 Jul. 1986, 9° 44.18Ј N, 82° 48.6Ј W, Wild Cane Key, N. Bastimentos Id., 0.9-4.2 Punta Cahuita Reef, 2-9 m, coll. G. Hendler; m, coll. G. Hendler et al.; (LACM 2003- (LACM 1986-141.2), 1 alc, Sta. CRA 86-27, 64.3), 1 alc, Sta. BDT 03-16, 7 Aug. 2003, Ј Ј 29 Jul. 1986, 9°44.18 N, 82°48.6 W, Punta 9°27.2Ј N, 82°18.02Ј W, S. Swan Key, 3 m, Cahuita Reef, 2-6 m, coll. G. Hendler; coll. G. Hendler et al.; (LACM 2003-65.3), 1 (LACM 1986-154.2), 3 alc, Sta. CRA 86-37, alc, Sta. BDT 03-17, 7 Aug. 2003, 9°25.6Ј N, Ј Ј 31 Jul. 1986, 9°39.7 N, 82°45.53 W, off Pu- 82°19.5Ј W, Boca del Drago, N. W. Colon erto Viejo, 1-4 m, coll. G. Hendler; (LACM Id., 4.5 m, coll. G. Hendler et al. PANAMA, 1986-197.2), 1 alc, 29 Oct. 1986, 9°44.18Ј N, Ј COLON PROV.: (LACM 1939-181.3), 1 dry, 82°48.7 W, Punta Cahuita Reef, 1-5 m, coll. Sta. AHF A8-39, R/V Velero III, 4 Apr. R. C. Brusca and P. M. Delaney; (LACM 1939, 8°53.72Ј N, 77°41.4Ј W, Bahía Caledo- 1987-40.1), 8 alc, Sta. CR 87-1, 19 May 1987, Ј Ј nia, off San Fulgencio Pt., shore; (LACM 9°44.18 N, 82°48.6 W, Punta Cahuita Reef, 1939-217.4), 2 dry, Sta. AHF A50-39, R/V 1-2 m, coll. G. Hendler and R. W. Peck; Velero III, 26 Apr. 1939, 8°53.63Ј N, 77°41.4Ј (LACM 1987-43.1), 6 alc, Sta. CR 87-4, 19 W, Bahía Caledonia, off San Fulgencio Pt., May 1987, 9°44.18Ј N, 82°48.6Ј W, Punta shore; (LACM 1939-222.3), 4 dry, Sta. AHF Cahuita Reef, 1-3 m, coll. G. Hendler and A57-39, R/V Velero III, 27 Apr. 1939, R. W. Peck; (LACM 1987-67.6), 1 dry, Sta. Ј Ј CR 87-4, 19 May 1987, 9°44.18Ј N, 82°48.6Ј 8°53.72 N, 77°41.5 W, Bahía Caledonia, off W, Punta Cahuita Reef, 1-3 m, coll. G. Hen- San Fulgencio Pt., shore; (LACM 1971- 671.1), 2 dry, 3 Apr. 1971, 9°24.3Ј N, dler and R. W. Peck; (LACM 1987-43.3), 2 Ј SEM, Sta. CR 87-4, 19 May 1987, 9°44.18Ј N, 79°51.9 W, Punta Galeta, 0.3-0.6 m, coll. G. Ј Hendler; (LACM 1971-672.1), 2 dry, 1-8 82°48.6 W, Punta Cahuita Reef, 1-3 m, coll. Ј Ј G. Hendler and R. W. Peck; (LACM 1987- Apr. 1971, 9°24.3 N, 79°51.9 W, Punta 50.2), 1 alc, Sta. CR 87-12, 20 May 1987, Galeta, coll. G. Hendler; (LACM 1974- Ј 9°44.53Ј N, 82°48.55Ј W, Punta Cahuita 248.1), 2 dry, 30 Oct. 1974, 9°24.05 N, Ј Reef, 2-3 m, coll. G. Hendler and R. W. 79°51.1 W, Largo Remo Key, 0.3-0.6 m, Peck; (USNM 1073475), 1 alc, Sta. CR 87-13, coll. G. Hendler; (LACM 1974-250.1), 20 May 1987, 9°44.37Ј N, 82°48.7Ј W, Punta 4 dry, 30 Nov. 1974, 9°24.3Ј N, 79°51.9Ј W, Cahuita Reef, 1-3 m, coll. G. Hendler and Punta Galeta, 6-10.6 m, coll. G. Hendler; R. W. Peck; (UCR-467), 1 alc, Sta. CR 87-13, (LACM 1974-252.1), 5 dry, 24 Aug. 1974, 20 May 1987, 9°44.37Ј N, 82°48.7Ј W, Punta 9°24.3Ј N, 79°51.9Ј W, Punta Galeta, 0-1 m, Cahuita Reef, 1-3 m, coll. G. Hendler and coll. G. Hendler; (LACM 1974-253.1), 1 R. W. Peck; (LACM 1987-63.2), 6 alc, Sta. dry, 14 Dec. 1974, 9°24.3Ј N, 79°51.9Ј W, CR 87-25, 23 May 1987, 9°43.95Ј N, 82° 48Ј Punta Galeta, coll. G. Hendler; (LACM W, Punta Cahuita Reef, 3-15 m, coll. G. 1975-704.1), 1 dry, 26 Jan. 1975, 9°33.7Ј N, Hendler and R. W. Peck; (LACM 1987- 79°41Ј W, Iron Castle Pt., Portobello, 66.1), 1 alc, Sta. CR 87-28, 23 May 1987, 0.9-3 m, coll. G. Hendler; (LACM 1977- 9°44.7Ј N, 82°48.68Ј W, Punta Cahuita Reef, 305.1), 12 alc, 1-30 Nov. 1977, 9°24.3Ј 1 m, coll. G. Hendler, R. W. Peck; (LACM N, 79°51.9Ј W, Punta Galeta, coll. K. 1987-67.1), 8 alc, Sta. CR 87-29, 87-30, 23 Dunlap. May 1987, 9°44.37Ј N, 82°48.7Ј W, Punta Etymology.—The species was studied, in Cahuita Reef, 1-2 m, coll. G. Hendler and part, under the auspices of the Centro de R. W. Peck; (LACM 1987-67.5), 2 alc, Sta. Investigaciones en Ciencias del Mar y Lim- CR 87-29, 87-30, 23 May 1987, 9°44.37Ј N, nología, of the Universidad de Costa Rica. 586 GORDON HENDLER

The specific name is based on the acronym edge, due to presence of longest arm spines for that organization, CIMAR. at that position; thereafter tapering gradu- Description.—Disk diameter of holotype ally toward arm tip. Dorsal arm plates 9.3 mm, length of longest arm 58 mm. rhombic, somewhat convex or keeled; Specimens examined ranged from 3.3-11.3 wider than long, with concave proximal mm dd, and from 18-65 mm AL. AL/dd edges, convex distal edges; distal edge ratios ranged from 4.4-8.5 (X = 5.9, SE = markedly thickened, upper surface of plate 0.12, n = 49). with small, slightly inflated proximal and Disk typically pentagonal, somewhat in- distal medial protrusions (Fig. 1D). Adja- flated, interradii protruding between arms; cent dorsal plates in contact or nearly so; spent individuals with little gonadal tissue several plates nearest disk markedly re- have round disks (Fig. 1A, B). Scales mod- duced in size, barely contiguous, not in erately large, rounded, tessellate, partially contact with disk; basal lateral arm plates obscured by integument. Spines on disk and spines on opposite sides of arm in close evenly spaced, somewhat tapering, rugose, contact. Surface of dorsal arm plate of dried short (approximately 1 mm tall), largest be- specimens appearing granulose under low tween radial shields; generally no more magnification. than 1 spine borne on a scale; spines dis- Lateral arm plate with prominent ridge tributed in 1-2 irregular rows between ra- bearing up to 10 slender arm spines; spine dial shields, in approximately 5 irregular length increasing markedly from dorsal- interradial rows; usually a few spinules on most to fourth spine, then decreasing step- dorso-medial portion of interradius and ra- wise to ventral-most spine. Length of long- dial shield. Radial shields flat, length ap- est spine approximately 2.5-3 times width proximately one-third dd; subtriangular, of dorsal arm plate. Dorsal spines at base of with thickened, concave adradial edge, arm rounded in cross-section, rugose, slen- abradial edge almost straight, distal edge der, gradually tapering from base; shaft of concave, sometimes strongly so; corners more distal dorsal spines dorso-ventrally forming rounded lobes; adradial corners in compressed (Fig. 1C, D). Ventral spines, contact, extending beyond disk and over particularly longest spine, dorso-ventrally arm base (Fig. 1C). compressed, with slight distal bend; lateral Jaw with small, peg-shaped dental papil- edges aciculate, distal tip somewhat flared lae on ventral apex, larger papillae on lat- (Fig. 1E). Dorsal-most spines directed eral edges, surrounding several shorter, somewhat dorsally; spines beneath disk thicker papillae, separated from teeth by short, directed posteriorly; ventral-most several pairs of larger, bar-shaped, ob- spines on distal one-third to one-half of liquely set dental papillae (Fig. 1F). Oral arm very small, with hyaline terminal hook shield subtriangular, centrally depressed, and several subterminal teeth. Tentacle with slightly concave lateral sides, long scale single, embedded in integument, convex distal edge, rounded corners. minute, flat to granular, with distal point Madreporite larger than oral shield, more (Fig. 1E). irregularly proportioned, sometimes with First ventral arm plate in mouth angle pore at distal edge. Adoral shield proximal with medial slot separating 2 proximally to oral shield and half its size, subtriangu- directed points; 2-3 most basal plates cen- lar, somewhat inflated. Ventral interradius trally depressed, overlapping, longer than naked, with long bursal slits buttressed by wide, centrally depressed (Fig. 1F). Re- pair of small, flared ossicles distal to oral maining plates approximately twice as shield and large, flared genital scales. wide as long, nearly touching, narrow, con- Arms somewhat dorso-ventrally com- stricted medially; lateral edges convex, con- pressed in appearance due to long, hori- cave proximal and distal edges of adjacent zontally projecting lateral arm spines; in plates bridged by band of soft tissue ap- cross section dorsal surface somewhat con- proximately one-half as long as plate (nar- vex, ventral surface flat. Arm breadth flar- rower in dry specimens), obscuring proxi- ing at about one-half to one dd beyond disk mal portion of plate. NEW SOUTHERN CARIBBEAN BRITTLE STARS 587

FIG.1. Ophiothrix cimar. A-B, photomicrographs of dried paratype, disk diameter = 7.7 mm (LACM 1987-67.6); C-F, scanning electron micrographs of paratypes (LACM 1987-43.3), treated to remove integument. A, entire, dorsal view; B, disk and base of arms, dorsal view; C, Disk edge showing few spines on radial shields, gap between disk and arm, and rugose, cylindrical spines on dorsal surface of arms and disk; D, arm, dorsal view; E, arm, ventral view, with tentacle scale indicated by arrowhead; F, Detail of oral frame, ventral view, showing structures associated with jaw including deeply incised 1st ventral arm plate (arrowhead).

Living individuals brown, pinkish- ently colored, frequently paler, rarely brown, reddish-brown, grayish-brown, darker, than dorsal surface of arm. Disk blackish, or yellowish. Arm color usually spines often whitish, yellowish or reddish, contrasts with disk, and arm spines differ- usually contrasting in coloration with 588 GORDON HENDLER darker radial shields. Interradial coloration with maximum AL of 58 mm and up to 8.5 generally contrasts with dorsal surface of times the dd, compared with the 120 mm disk; pale colored gonads sometimes vis- arms, 10 times the dd, found in O. lineata.It ible through naked interradial integument. is distinguished by spines on the disk, Arms have thin, dark, medial dorsal stripe rather than sparse, minute stumps. More- flanked by paler stripes; arms also banded over, it is cryptic and lives in hard sub- dorsally and ventrally, with about every strate, whereas O. lineata is a symbiont of fourth segment more darkly pigmented tube sponges (Hendler 1984). The dorso- than the others. Ventral arm plates and oral ventrally compressed body shape of O. frame with dark brown flecks and pale cimar somewhat resembles that of O. stri, O. grayish-brown and reddish-brown blotches brachyactis H. L. Clark, 1915, and O. hart- that may cover a portion or almost entire fordi A. H. Clark, 1939, but it is reliably dis- plate, surrounded by pale pigmentation. tinguished by the spinose disk. Tube feet yellowish; oral tentacles pale yel- An Ophiothrix species from eastern Ven- lowish-brown. ezuela, identified by Zoppi de Roa (1967) as Preserved specimens with disk grayish, Ophiothrix brachyactis, has a spinose disk purplish-gray, purplish-brown, or brown- like that of O. cimar rather than a granulose ish-gray; large ossicles such as radial disk like the holotype of O. brachyactis. Its shields and ventral arm plates often disk morphology, banded arms, and later- marked with a combination of dark, mod- ally directed arm spines also resemble the erately dark and pale colors; patches of features of O. cimar. However, the arms white at the base of some disk spines and at lack a dark medial stripe and are only 2-3 edges of radial shields; black patches on times as long as the dd (see Zoppi de Roa some scales; some specimens with micro- 1967: fig. 20). Moreover, it lives in masses of scopic network of black and pale markings mussels and algae, and thus appear to oc- between disk scales. Disk spines pale cupy a different microhabitat than O. cimar. brown. Ventral interradius dark brown to The available evidence indicates a similar- black sometimes mottled with yellowish- ity of the Venezuelan specimens to O. brown. Arms retain pattern of the stripes cimar, but suggests that it represents an un- and bands visible in living individuals. described species. Oral tentacles white with black flecks, or Distribution.—At present, known from grayish. Cahuita, Costa Rica to Caledonia Bay, Variations.—Juveniles differ from adults Panama, and from the intertidal to a depth in several respects. An individual 3.3 mm of approximately 10 m, generally in less dd has typical disk spines, but bare radial than 3 m. shields. The arm spines are slender, some- Biology.—Individuals are cryptic, and what dorso-ventrally compressed, lacking were found beneath coral rubble, including flared tip; the largest are slightly rugose. large slabs of Acropora palmata, and at least Its primary plates are irregular, but in a sometimes in close proximity to encrusting 4.4 mm individual a central and 5 radial sponge. When disturbed, they initiated a plates are evident. vigorous escape response. Individuals typi- Comparisons.—Ophiothrix cimar is notably cally accumulated clumps of flocculent, similar to Ophiothrix lineata Lyman, 1860. mucus-bound silt on the disk and arm Both have a somewhat dorso-ventrally spines, unlike sympatric O. angulata and O. compressed body shape, large fairly bare stri. The species is restricted to shallow wa- radial shields, and a similar pattern of arm ter near fringing coral reef, and generally stripes. However, continuation of the dark occupies rubble-strewn back reef habitats and pale arm stripes onto the disk between with varying degrees of coral cover and the radial shields, which is marked in adult three-dimensional relief. Some individuals O. lineata, is only noticeable in very small were found on fairly high-energy reef plat- individuals of O. cimar. Ventral arm bands form, as well as under much more pro- are more noticeable in O. cimar than in O. tected conditions beneath coral rubble in a lineata. Ophiothrix cimar is a smaller species, Thalassia bed. NEW SOUTHERN CARIBBEAN BRITTLE STARS 589

All of the specimens of O. cimar were ex- 9°44.53Ј N, 82°48.55Ј W, Punta Cahuita amined externally for evidence of brood- Reef, 1-2 m, coll. G. Hendler, and R. W. ing, and embryos were not seen in the bur- Peck; (LACM 1987-50.1), 1 alc, Sta. CR 87- sal slits, or in the bursae of dissected 12, 20 May 1987, 9°44.53Ј N, 82°48.55Ј W, females. Of the 20 specimens that were dis- Punta Cahuita Reef, 2-3 m, coll. G. Hendler, sected, 12 were male (X = 7.89 mm dd, SE = R. W. Peck; (LACM 1987-51.1), 1 alc, Sta. 0.51, range = 4.79-10.51), and eight were fe- CR 87-13, 20 May 1987, 9°44.37Ј N, 82°48.7Ј male (X = 9.11 mm dd, SE = 0.47, range = W, Punta Cahuita Reef, 1-3 m, coll. G. Hen- 6.92-11.31). The overlapping sizes of males dler and R. W. Peck; (LACM 1987-51.4), 6 and females, and the lack of hermaphro- alc, Sta. CR 87-13, 20 May 1987, 9°44.37Ј N, ditic specimens, indicate that the species is 82°48.7Ј W, Punta Cahuita Reef, 1-3 m, coll. gonochoric. The eggs in gonads of 5 of the G. Hendler and R. W. Peck; (LACM 1987- larger alcohol-preserved females (X = 9.87 63.1), 3 alc, Sta. CR 87-25, 23 May 1987, mm dd, SE = 0.43, range = 9.04-11.31) were 9°43.95Ј N, 82°48Ј W, Punta Cahuita Reef, whitish, and had a mean diameter of 0.16 3-15 m, coll. G. Hendler, and R. W. Peck; mm (SE = 0.002, n = 50, range = 0.128- (LACM 1987-67.4), 11 alc, Sta. CR 87-29, 87- 0.208). 30, 23 May 1987, 9°44.37Ј N, 82°48.7Ј W, Punta Cahuita Reef, 1-2 m, coll. G. Hendler, Ophiothrix stri, new species and R. W. Peck; (LACM 1987-68.1), 1 alc, Sta. CR 87-31, 24 May 1987, 9°44.97Ј N, Ј FIGURE 2A-F. 82°49.3 W, NE of Sloth River mouth, 0-1 m, coll. G. Hendler and R. W. Peck. Holotype.—PANAMA, BOCAS DEL PANAMA, BOCAS DEL TORO PROV.: TORO PROV.: (LACM 2003-51.2), Sta. BDT (LACM 2003-105.1), 2 alc, 12 Aug. 2003, 03-4, 4 Aug. 2003, 9°20.66Ј N, 82° 0.33Ј W, 9°27.17Ј N, 82°18.01Ј W, Swan Cay, 1.5 m, Wild Cane Key, N. Bastimento Id., 0.9-4.2 coll. K. Fitzhugh and W. G. Keel; (LACM m, coll. G. Hendler et al. 2003-51.1), 18 alc, Sta. BDT 03-4, 4 Aug. Paratypes.—COSTA RICA, LIMON 2003, 9°20.66Ј N, 82°10.33Ј W, Wild Cane PROV.: (LACM 1986-106.1), 3 alc, Sta. CRA Key, N. Bastimento Id., 0.9-4.2 m, coll. G. 86-1, 26 Jul. 1986, 9°44.18Ј N, 82°48.6Ј W, Hendler et al.; (LACM 2003-51.4), 2 alc, Sta. Punta Cahuita Reef, 1-3 m, coll. G. Hendler; BDT 03-4, 4 Aug. 2003, 9°20.66Ј N, 82° (LACM 1986-121.1), 1 alc, Sta. CRA 86-12, 10.33Ј W, Wild Cane Key, N. Bastimento 27 Jul. 1986, 9°40.8Ј N, 82°45.53Ј W,Nof Id., 0.9-4.2 m, coll. G. Hendler et al.; (LACM Puerto Viejo, 1-3 m, coll. G. Hendler; 2003-51.5), 25 alc, Sta. BDT 03-4, 4 Aug. (LACM 1986-127.1), 1 alc, Sta. CRA 86-18, 2003, 9°20.66Ј N, 82° 10.33Ј W, Wild Cane 27 Jul. 1986, 9°40.8Ј N, 82°45.53Ј W,Nof Key, N. Bastimento Id., 0.9-4.2 m, coll. G. Puerto Viejo, 1-3 m, coll. G. Hendler; Hendler et al.; (USNM 1073476), 2 alc, Sta. (LACM 1986-136.1), 6 alc, Sta. CRA 86-22, BDT 03-4, 4 Aug. 2003, 9°20.66Ј N, 82°10.33Ј 28 Jul. 1986, 9°44.18Ј N, 82°48.6Ј W, Punta W, Wild Cane Key, N. Bastimento Id., 0.9- Cahuita Reef, 2-9 m, coll. G. Hendler; 4.2 m, coll. G. Hendler et al.; (LACM 2003- (LACM 1986-139.1), 1 alc, Sta. CRA 86-25, 51.8), 2 SEM, Sta. BDT 03-4, 4 Aug. 2003, 28 Jul. 1986, 9°44.18Ј N, 82°48.6Ј W, Punta 9°20.66Ј N, 82°10.33Ј W, Wild Cane Key, N. Cahuita Reef, 2-9 m, coll. G. Hendler; Bastimento Id., 0.9-4.2 m, coll. G. Hendler (LACM 1986-141.1), 3 alc, Sta. CRA 86-27, et al.; (LACM 2003-51.7), 1 dry, Sta. BDT 29 Jul. 1986, 9°44.18Ј N, 82°48.6Ј W, Punta 03-4, 4 Aug. 2003, 9°20.66Ј N, 82°10.33Ј W, Cahuita Reef, 2-6 m, coll. G. Hendler; Wild Cane Key, N. Bastimento Id., 0.9-4.2 (LACM 1986-154.1), 5 alc, Sta. CRA 86-37, m, coll. G. Hendler et al.; (LACM 2003- 31 Jul. 1986, 9°39.7Ј N, 82°45.53Ј W, off 64.1), 3 alc, Sta. BDT 03-16, 7 Aug. 2003, Puerto Viejo, 1-4 m, coll. G. Hendler; (UCR- 9°27.2Ј N, 82°18.02Ј W, S. Swan Key, 3 m, 468), 3 alc, 29 Oct. 1986, 9°44.18Ј N, 82°48.7Ј coll. G. Hendler et al.; (LACM 2003-64.2), 3 W, Punta Cahuita Reef, 1-5 m, coll. R. C. alc, Sta. BDT 03-16, 7 Aug. 2003, 9°27.2Ј N, Brusca, and P. M. Delaney; (LACM 1987- 82°18.02Ј W, S. Swan Key, 3 m, coll. G. Hen- 49.1), 4 alc, Sta. CR 87-11, 20 May 1987, dler et al.; (LACM 2003-64.6), 1 alc, Sta. 590 GORDON HENDLER

BDT 03-16, 7 Aug. 2003, 9°27.2Ј N, 82°18.02Ј hemispherical tip bearing approximately 20 W, S. Swan Key, 3 m, coll. G. Hendler et al.; microscopic thorns; generally only 1 stump (LACM 2003-65.2), 2 alc, Sta. BDT 03-17, 7 borne on a scale; stumps distributed in 1-2 Aug. 2003, 9°25.6Ј N, 82°19.5Ј W, Boca del irregular radial rows, 4-5 irregular interra- Drago, N.W. Colon Id., 4.5 m, coll. G. Hen- dial rows; interradial rows protruding dler et al. PANAMA, COLON PROV: slightly between arms; a dense group of (LACM 1939-174.2), 1 dry, Sta. AHF A1-39, stumps at disk edge capping distal lobe of R/V Velero III, 3 Apr. 1939, 8°53.7Ј N, radial shield; only rarely are a few granules 77°40.5Ј W, Bahia Caledonia, islet south of on lateral edge of interradius. Radial Isla de Oro, shore; (LACM 1939-222.4), 1 shields subtriangular, approximately one- dry, Sta. AHF A57-39, R/V Velero III, 27 third dd in length, with slightly concave Apr. 1939, 8°53.72Ј N, 77°41.5Ј W, Bahia adradial edge, concave distal edge, straight Caledonia, off San Fulgencio Pt., shore; to slightly convex abradial edge; corners (LACM 1974-249.1), 3 dry, 15 Dec. 1974, ca. forming rounded lobes; central portion of 9°33.7Ј N, 79°41Ј W, Iron Castle Point, Por- plate slightly inflated; bearing several small tobello, 0-3 m, coll. G. Hendler; (LACM granules, generally near abradial edge; 1974-250.2), 4 dry, 30 Nov. 1974, 9°24.3Ј N, adradial corners separated by several small 79°51.9Ј W, Punta Galeta, 6-10.6 m, coll. G. ossicles in series with dorsal arm plates, Hendler; (LACM 1974-251.1), 1 dry, 2 Sep. bridging gap between arm and disk edge. 1974, 9°24.3Ј N, 79°51.9Ј W, Punta Galeta, Surface of radial shields and dorsal arm 0-10.6 m, coll. G. Hendler; (LACM 1974- plates of dried specimens appearing granu- 254.1), 1 dry, 25 Nov. 1974, ca. 9°33.7Ј N, lose under low magnification (Fig. 2C). 79°41Ј W, Iron Castle Pt., Portobello, 0-1.5 Jaw with short, peg-shaped dental papil- m, coll. G. Hendler; (LACM 1975-705.1), 1 lae on ventral apex, larger papillae on lat- dry, 18 Apr. 1975, ca. 9°33.7Ј N, 79°41Ј W, eral edges surrounding inner series of Iron Castle Pt., Portobello, 0.9-3 m, coll. G. shorter papillae, separated from teeth by Hendler; (LACM 1975-706.1), 1 dry, 25 Mar. several chevrons of large, bar-shaped, ob- 1975, ca. 9°32.75Ј N, 78° 54.5Ј W, Media liquely set dental papillae (Fig. 2F). Oral Tupo, San Blas, 0.9-1.5 m, coll. G. Hendler; shield subtriangular to subpentagonal, cen- (LACM 1975-707.1), 4 dry, 25 Mar. 1975, ca. trally depressed, with straight or somewhat 9°30.75Ј N, 78°47.83Ј W, Salar Salar, 6 m, concave lateral edges, long convex distal coll. G. Hendler; (LACM 1975-708.1), 4 dry, edge, rounded corners. Madreporite 26 Mar. 1975, ca. 9°30.75Ј N, 78°47.83Ј W, larger, more rounded or lobate than oral Salar Salar, 1.5 m, coll. G. Hendler. shield, sometimes with 1-2 microscopic Etymology.—The species was studied, in pores near concave distal edge. Adoral part, under the auspices of the Smithsonian shield proximal to oral shield and one-half Tropical Research Institute. The specific its size, subtriangular, somewhat inflated. name is based on the acronym for that or- Ventral interradius naked, with long bursal ganization, STRI. slits buttressed by pair of ossicles distal to Description.—Disk diameter of holotype oral shield, and with large, flared genital 9.8 mm, length of longest arm 42 mm. scales. Specimens examined range from 2.3-10.6 Arm compressed in appearance due to mm dd, and from 8.6-54 mm AL. AL/dd horizontal orientation of arm spines; in ratios ranged from 2.8-7.2 (X = 5.0, SE = cross section dorsal surface slightly convex, 0.08, n = 91). ventral surface flat. Arm breadth flaring at Disk generally round, not inflated, dorsal less than 1 dd beyond disk edge, due to margin sharply demarcated from naked presence of longest arm spines and widest ventral interradius; interradii protruding in dorsal arm plates at that position; thereafter ripe individuals (Fig. 2A, B). Scales moder- tapering gradually toward arm tip. Dorsal ately large, rounded, tessellate, partially arm plates thick, generally flat or slightly obscured by integument, bearing promi- rounded, not keeled; successive plates nent stumps. Stumps numerous, nearly as closely abutting, in contact for about one- wide as tall (approximately 0.2 mm), with third width of plate (Fig. 2D). Plates sub- NEW SOUTHERN CARIBBEAN BRITTLE STARS 591

FIG.2. Ophiothrix stri. A-B, photomicrographs of dried paratype, disk diameter = 8.0 mm (LACM 2003-51.7); C-F, scanning electron micrographs of paratypes (LACM 2003-51.8), treated to remove integument. A, entire, dorsal view; B, disk and base of arms, dorsal view; C, Disk edge showing few rounded stumps on radial shields, series of small ossicles connecting disk and arm, and rugose, compressed, proximally directed spines on base of arm; D, arm, dorsal view; E, arm, ventral view, with tentacle scale indicated by arrowhead; F, Detail of oral frame, ventral view, showing structures associated with jaw including deeply incised 1st ventral arm plate (arrowhead). hexagonal, twice as wide as long; proximal shape. Small, basal dorsal arm plates differ and distal edges straight, proximo-lateral in having medial protrusion on distal edge; edges concave; 3 disto-lateral edges convex, first plate overlaps the second and is over- but sometimes merging into single convex lapped by lateral arm plates. 592 GORDON HENDLER

Lateral arm plates with prominent ridge individuals; 2-3 pale arm joints alternate bearing up to ten arm spines; dorsal and with each darker joint. Ventral surface of ventral spines conspicuously different. arms whitish, pale purple or pinkish, with Spine length increasing markedly from broad brown medial stripe which is dorsal-most to fourth spine, then decreas- nearly imperceptible in deeply pigmented ing gradually to ventral-most spine. Length individuals; oral frame whitish with flecks of longest spine approximately 1.3 times of darker color; ventral interradii deep width of dorsal arm plate. Dorsal-most purple or brown, partially whitish in ripe spine only 1-4 times longer than wide, individuals; oral tentacle and first tube feet ovoid. Dorsal spines on first free joint di- pale yellowish; tube feet gray with black rected toward disk, those on distal joints flecks, oral tentacles yellowish with brown directed laterally in horizontal plane of arm tip. (Fig. 2C). On basal joints dorsal-most three In darkly-pigmented, preserved speci- spines thick, rugose; shaft distinctly ovoid mens, disk dark purplish-gray and black; in cross section; broad, distal portion some- arms with thin, blackish medial stripe what dorso-ventrally compressed, tapering flanked by pale stripes. Arm spines gray to blunt point (Fig. 2C, D). Lower spines with black flecks, contrasting with arms. narrow, with slight distal bend, shaft Lateral arm plates and ventral arm plates dorso-ventrally compressed, lateral edges with mottled black pattern forming micro- aciculate (Fig. 2E). Space between spines on scopic white bands near distal edge of lat- adjacent basal joints constricted due to eral arm ridge. Oral region whitish to large diameter of dorsal spine; distance be- pale grayish-violet, with black patches and tween slender spines on distal joints ex- flecks on major ossicles. Ventral interradius ceeding spine diameter. Ventral-most dark blackish-purple and brown. Ventral spines on distal one-half to two-thirds of arm surface pale brownish-gray or violet- arm very small, with hyaline terminal hook and 1-3 subterminal teeth. Tentacle scale gray, with microscopic black flecks and single, minute, embedded in integument, broad dark medial stripe. Tube feet black granular to flat, triangular with distal point with pale papillae. Small individuals rela- (Fig. 2E). tively pale. Yellow individuals with arm First ventral arm plate in mouth angle spines and median arm stripe brown. with medial slot separating 2 proximally Variations.—A juvenile individual directed points; following plate longer than 2.3 mm dd, has a central plate covering 25% wide, centrally depressed, overlapping 3rd of the dd, and the radial plates are very plate, which is also centrally depressed small and separated by intercalary scales. (Fig. 2F). Remaining plates very short, Disk armament consists of multifid broad, octagonal; 5 proximal edges nearly spinules with a thick shaft; 0-1 per scale straight, 3 distal edges straight or concave; and several per radial shield. The arm successive plates separated by straight, spines are compressed, and basal spines broad ribbons of soft tissue nearly one- are wide, with strongly aciculate lateral third width of plate. edges. The central plate covers 10% of the Living individuals purplish-red, pur- dd in a 3.1 mm dd individual, 6% in a 6.7 plish-brown, pale grayish purple-brown, mm individual, and is often indiscernible ochre, or yellow; ventral surface pinkish. in large specimens. The 3.1 mm dd animal Disk may be grayish brown or purple with has spine-capped stumps on the disk, black interradii, arm spines grayish-brown; which developed from multifid spinules, arms brownish-gray, brown, or grayish- and broad, dorso-ventrally compressed, ru- purple with black, dark brown, orange, or gose arm spines. A 3.7 mm dd specimen reddish medial stripe flanked by tan or has stripes on radial shields resembling the salmon-colored stripes. Stripes indistin- pattern in O. lineata. Only the first dorsal guishable in intensely dark individuals, ex- arm plate is wide in the 2.3-3.1 mm speci- cept near tips of arms. Arm bands particu- mens, but an individual 4.7 mm dd has larly evident in small or pale-colored mostly adult features. NEW SOUTHERN CARIBBEAN BRITTLE STARS 593

Comparison.—Ophiothrix stri is strikingly grass. Individuals beneath rocks cling to similar to O. synoecina Schoppe, 1996. Both the rocks, and do not lie on the sediment. have a somewhat dorso-ventrally com- The species is restricted to shallow water, pressed body shape, large fairly bare radial near fringing coral reef. The most densely shields, smooth, cylindrical basal arm populated collecting sites were exposed spines projecting over the disk, and other reef platforms, although individuals also shared morphological, biological and be- occurred on rubble-strewn back reef habi- havioral traits. Among the features distin- tats and in the shallow subtidal. In the San guishing O. stri are a greater number of col- Blas Islands, where the shoreline is sandy, umns (4-5 vs. 2-3) of interradial stumps, the shallowest individuals were found un- dorsal arm plates lacking a keel, fewer arm der slabs of coral rubble at depths to 2 m, spines (maximum of 10 vs. 11), and rela- and in stands of Agaricia and Porites spp. at tively few stumps on the lateral interradial depths to 6 m, on the reef slope. surface of the disk. Furthermore, the At Bocas del Toro relatively large num- striped arm pattern of O. stri, visible bers of individuals were found at two high- throughout the arms of pale-colored indi- energy hard-bottom localities where the viduals and at the arm tips of deeply pig- substrate is riddled by the boring echinoid mented individuals, differs from O. synoe- Echinometra lucunter (Linnaeus, 1758). Some cina, which has a bluish-gray and whitish individuals were found inside the echi- or reddish disk, arms with whitish, blue noid’s burrows, but the association was not and reddish dorsal bands, or uniform blu- obligatory for either species. The ophiu- ish-black dorsal coloration, and a purely roids were absent from many boreholes, cream-colored ventral arm surface and regularly occurred in rock crevices (Schoppe 1996). without echinoids, and under large rocks Other Caribbean congeners with which that sheltered other ophiuroids including O. stri might be confused all share a some- Ophiocoma and Ophioderma species. Rela- what dorso-ventrally compressed body tively few individuals were found at pro- shape, and large rather bare radial shields. tected localities. Under those conditions These include O. lineata and O. cimar, which other species of Ophiothrix predominated in also have a dark medial arm stripe set the substrates occupied by O. stri, such as at off by pale flanking stripes, and O. brachy- Cahuita where 58 ophiuroids in a sample of actis, O. hartfordi, O. platyactis H. L. Clark, the calcareous alga, Halimeda spp., were 1939, and O. lineata which also have sparse, composed of 4 individuals of O. stri and 42 granulose disk armament. However, O. angulata. Similarly, in a sample of coral rugose, ovoidal, proximally-project- and rubble collected at a low-energy habi- ing basal arm spines, and comparatively tat in the San Blas Islands, 4 O. stri,1O. large, rounded disk stumps distinguish O. angulata, and 28 O. orstedii Lütken, 1856, stri. were the only ophiuroids found. Distribution.—At present, known from The association between O. stri and Echi- Cahuita, Costa Rica to Caledonia Bay, nometra lucunter seen at Bocas del Toro, Panama, and from the intertidal to a depth Panama was also noted at Puerto Viejo, of approximately 10 m, generally in less Costa Rica, but not at Cahuita or Galeta. At than 3 m. the latter two localities its presence in echi- Biology.—Individuals are cryptic, lethar- noid boreholes may have been overlooked, gic, and relatively unresponsive to distur- since exhaustive surveys for the species bance. Depending on the locality, they live were not conducted. Furthermore, in at under coral rubble such as large slabs of least two instances at Galeta, O. stri was Acropora palmata or rocks, but in one case found with the nocturnal comatulid cri- were found in clumps of the calcareous noid, Comactinia echinoptera (Müller, 1840). alga Halimeda sp., and in several instances The ophiuroids were seen in contact with were found clinging to encrusting sponge. the cirri of crinoids during night dives on They generally live in elevated structures the reef slope, at depths of approximate- rather than in proximity to sand or sea- ly 8 m. 594 GORDON HENDLER

In addition to its biological associations, 9.87 mm dd, SE = 0.32) had a mean diam- the coloration of O. stri varies within and eter of 0.16 mm (SE = 0.002, n = 50, range = between localities. Although the species is 0.11-0.19). However, formalin-preserved typically a dark purplish-brown or pur- eggs that were spawned and fertilized were plish-gray, yellow individuals were occa- somewhat larger (X = 0.23 mm, SE = 0.002, sionally found in Panama, but not else- range = 0.21-0.24, n = 30), and were sur- where. In one instance at Galeta, a large rounded by a fertilization envelope 30% number of yellow individuals were found larger than the egg. The 30% greater diam- in the shallow subtidal, under slabs of eter of the eggs preserved in formalin than coral. However, at the same locality indi- those in alcohol is probably due to the de- viduals collected in the intertidal had such hydration and shrinkage of the alcoholic an intense purple coloration that arm material. stripes were evident only near the tips of the arms. At Bocas del Toro, darkly pigmented individuals prevailed, but some DISCUSSION animals were a distinctly pale purplish- gray or ochre. In contrast, the pigmentation Distribution and habitat of animals at Cahuita was less intense than in the Panamanian animals, and stripes and banding patterning was evident on the full The westernmost populations of both length of their arms. There is also a degree new species were found at Cahuita Na- of intraspecific variability in the color of O. tional Park in Costa Rica (see Cortés and cimar and O. synoecina, but it is not clear if Risk 1985; Fonseca, 2003 for descriptions of the pigmentation is a genetically deter- the locality). Populations were also found mined trait, or under the control of an en- in western Panama at Bocas del Toro (see vironmental influence such as diet. Guzmán and Guevara 1998a, b, 1999), from A group of individuals of O. stri held in Galeta to Portobello in the central region of a bowl of standing seawater spawned late Panama (see MacIntyre and Glynn 1976; in the evening of the day they were col- Cubit et al. 1989), and as far east as the San lected at Bocas del Toro. The eggs are whit- Blas Islands and at Caledonia Bay near the ish in color, and each fertilized egg was en- Colombian border (see Garth 1945; Robert- closed in a spacious fertilization envelope. son and Glynn 1977). Despite differences in Most of the developing embryos reached a reef structure among the localities, all the two cell stage several hours after fertiliza- animals were collected on shallow fringing tion. Eggs and embryos were readily trans- coral reefs exposed to strong currents, large ported by water movement, but were waves, and varying amounts of coastal run- heavier than seawater. Many adhered off. tightly to the surface of the glass container, The individuals in Central Panama and and could be freed only with a strong cur- the San Blas Islands occupied reef plat- rent of water. forms and shallow subtidal slopes. Those at All of the specimens of O. stri were ex- Galeta were abundant intertidally, despite amined externally for evidence of brood- the stress from emersion, heating, wave im- ing, and embryos were not seen in the bur- pact, and predation that cause periodic sal slits, or in the bursae of dissected mass mortality in local echinoid popula- females. Of the 15 specimens that were dis- tions (Hendler 1977). At Bocas del Toro sected, six were male (X = 6.10 mm dd, SE they were found near the periphery of the = 0.31, range = 4.66-6.78), and 9 were fe- archipelago in relatively high-energy habi- male (X = 7.00 mm dd, SE = 0.47, range = tats, and at Cahuita they occurred in shal- 4.66-8.65). The overlapping sizes of males low reef directly behind the outer reef crest. and females, and the lack of hermaphro- It seems likely that at the mouth of Caledo- ditic specimens, indicates that the species is nia Harbor, Panama, they were also found gonochoric. The eggs in gonads of five of on exposed coastline, based on the informa- the larger alcohol-preserved females (X = tion in Garth (1945). NEW SOUTHERN CARIBBEAN BRITTLE STARS 595

Ophiothrix cimar was relatively more In contrast with O. stri, O. synoecina was common than O. stri in backreef microhabi- described as an obligate associate of E. tats, but neither species occurred in nearby lucunter (Schoppe 1996; Schoppe and lagoonal and mangrove habitats. A pos- Werding 1996). At the study site in Colom- sible preference of O. stri for high energy bia, where large numbers of Echinometra sites is borne out by the roughly equal riddled the reef substrate, up to 95% of the number of both species collected in the boreholes were occupied by O. synoecina, backreef at Cahuita, in contrast to the pre- and up to 10 ophiuroids were found per dominance of O. stri at the considerably borehole, reportedly nestled beneath the more exposed habitats in Bocas del Toro. echinoid. Schoppe and Werding (1996) also Even in the fairly sheltered environment at indicated that the Colombian echinoids do Cahuita, most individuals were found in not leave their cavities, which is surprising less than one meter of water, very close to since that is not the case elsewhere in the the forereef crest. At least in certain cases, Caribbean (Hendler et al. 1995). Although both species were found in sponge- observations on O. synoecina led to the con- encrusted microhabitats. Other ophiuroids clusion that it is “found exclusively in as- occurring in close proximity to the new sociation with Echinometra” (Schoppe and species included common shallow-water Werding 1996:184), it is not clear whether Caribbean species such as Ophiothrix angu- the ophiuroid was entirely absent from lata, O. orstedii, Ophioderma appressum (Say, other substrates. That point should be re- 1825), O. brevicaudum Lütken, 1856, O. ci- evaluated, since the observations on O. stri nereum Müller and Troschel, 1842, O. rubi- suggest that O. synoecina may be less host- cundum Lütken, 1856, Ophiocoma echinata specific than has been suggested. (Lamarck, 1816), O. pumila Lütken, 1859, O. The observation that O. stri and O. cimar wendtii Müller and Troschel, 1842, Ophio- may live in sponge-encrusted niches is of lepis impressa Lütken, 1859, Ophiomyxa flac- interest, considering the habits of O. lineata. cida (Say, 1825), Ophionereis reticulata (Say, The latter species is an obligate symbiont 1825), and O. olivacea H. L. Clark, 1901. on tube sponges. It occurs on sponges with Ophiothrix angulata and O. orstedii. How- Biological associations ever, it has not been found in association with O. cimar or O. stri, despite its close The occurrence of O. stri in the cavities morphological similarities, particularly to excavated and occupied by Echinometra the latter species. The nature of the relation- lucunter is noteworthy, as a similar relation- ship of the new species with encrusting ship was previously reported for O. synoe- sponges is yet to be determined, and more cina Schoppe and Werding 1996). The asso- extensive, detailed observations on their as- ciation is not obligatory for O. stri, which sociations are desirable. frequently occurs in crevices without echinoids, and also associates with a crinoid, Reproductive biology Comactinia echinoptera. Thus, it is likely that the echinoid burrows and crinoid crevices Two Caribbean Ophiothrix species are offer a suitable microhabitat for the ophiu- known to have planktotrophic larvae, O. roid, but the presence of O. stri together angulata and Ophiothrix suensonii Lütken, with echinoids and crinoids is incidental. 1856. They have small oocytes that are 0.10- The association of O. stri and E. lucunter 0.13 mm in diameter, and only slightly was observed in Bocas del Toro, Panama smaller than the fertilization envelope. and Puerto Viejo, Costa Rica, and the asso- Ophiothrix suensonii develops from an ciation with C. echinoptera was seen at 8-armed ophiopluteus larva, which is typi- Galeta, but not at other localities where the cal of the ophiotrichids that have been species co-occurred. However, extensive reared (Stancyk 1973; Mladenov 1985). In surveys were not carried out to assess the contrast, O. orstedii, another Caribbean spe- habitat preferences and requirements of O. cies, produces eggs 0.25 mm diameter, en- stri. closed in a very spacious fertilization enve- 596 GORDON HENDLER lope, and has a rapidly developing, conditions in vitro were substantially dif- 2-armed larva (Mladenov 1979). Although ferent than under natural conditions, the it is the only ophiotrichid known to have laboratory observations suggest that were abbreviated development, lecithotrophic O. stri to spawn in a semi-enclosed space, larvae are the norm for ophiuroid species like an Echinometra burrow, its fertilized with moderately yolky eggs (Hendler eggs would not be “carried” by the females, 1991). although they could adhere to the walls of The three Caribbean Ophiothrix species the cavity and develop near adult ophiu- with known development have adults with roids. Thus, there are similarities in the re- separate sexes, and although their eggs are productive behaviors of O. synoecina and O. heavier than seawater they are readily sus- stri in terms of their yolky eggs and de- pended because the fertilization envelope mersal development, however the latter is not adhesive. They differ markedly from species is gonochoric and does not appear O. synoecina, which is protandric, with a to brood or to carry associated juveniles. small male, and a transitional stage with ovotestes that develops into a large female, Systematic affinities which is thought to brood embryos externally (Schoppe and Holl 1994). Its large There are similarities between O. lineata yolky eggs, up to 0.30 mm in diameter, are and members of the subgenus Placophio- reportedly carried in “clusters…in front of thrix that may have significance for the the mouth” of females, and the young classification of the new species. A. M. ophiuroids were described as clinging to Clark (1966) restricted the genus Placophio- the females for two weeks (Schoppe and thrix H. L. Clark, 1938 to subgeneric rank, Holl 1994:472). Its combination of repro- because the distinctions among Placophio- ductive traits is unusual, as O. synoecina is thrix and the other ophiotrichid subgenera the only protandric ophiotrichid that is were, and remain, elusive (Hoggett 1991). known, and only the second ophiuroid spe- She characterized Placophiothrix by arms cies that is suggested to brood externally; usually 6-9 times the dd, radial shields bare the embryos of other brooding ophiuroids or with few stumps, moderately large disk develop in the parent’s bursae. The expres- scales bearing spaced spines, spinelets, or sion of protandry in the species is of addi- stumps, and opaque arm spines (Clark and tional significance since, among the ophiu- Rowe 1971). Despite the resemblance of O. roids, hermaphroditism has only been lineata to the Placophiothrix species, all of observed in brooding species (Hendler which are Indo-West Pacific, A. M. Clark 1991). In regard to the issue of brooding in (1966) referred the Caribbean species to the Caribbean Ophiothrix species, it is notable subgenus Ophiothrix. that there are minute, juvenile ophiuroids Ophiothrix stri shares a number of impor- clinging to the very small type specimens tant features with O. lineata including: of O. brachyactis and O. hartfordi. It remains large, separated radial shields bearing few to be determined if the association is in- stumps, granulose disk armament almost dicative of external or internal brooding re- entirely restricted to the dorsal surface, production or was coincidental. borne on moderately large scales; wide, Schoppe and Holl (1994) did not describe flat, hexagonal dorsal arm plates, broadly how the eggs of O. synoecina come to ag- in contact; slotted first ventral arm plate gregate in “clusters” and how they are held and overlapping 2nd and 3rd ventral arm by the females. If the eggs of O. synoecina plates; broad, octagonal ventral arm plate; have an adhesive fertilization envelope, dorso-ventrally compressed, rugose lateral like those of O. stri, it is possible that they arm spines; striped arms. They also differ could stick to one another or to the female, in important respects, particularly the more facilitating their manipulation by the fe- elongated, slender compressed basal arm male. The eggs of O. stri, however, did not spines, and smaller, multiple stumps on the form clusters, nor were they gathered or disk scales of O. lineata. However, O. stri is clasped by the adult individuals. Although most similar to O. synoecina, the only note- NEW SOUTHERN CARIBBEAN BRITTLE STARS 597 worthy differences being in the shape and small, about 0.13 mm in diameter, and give microstructure of the dorsal arm plate, rise to a planktotrophic, 8-armed ophioplu- proximity of the 2nd and 3rd ventral arm teus. The morphological similarities among plates, structure of the lowest arm spine, the first four species suggest that lecitho- and superficial differences in pigmentation. trophy is expressed in members of one, or The three species resemble each another, possibly two clades of Caribbean Ophiotri- considerably more than they resemble O. chidae. The particularly close similarity be- cimar, but have in common with the latter: tween O. stri and O. synoecina suggests that disk armament borne on moderately large they are sister species, probably with a scales, wide dorsal arm plates, rugose, common ancestor that was a gonochoric compressed arm spines, slotted 1st ventral broadcast-spawner, and they appear to arm plate, and long, narrow 2nd ventral have a greater systematic affinity to O. lin- arm plate overlapping the 3rd ventral arm eata than to O. cimar. Furthermore, those plate. They resemble to a lesser degree sev- four species are considerably more similar eral small, poorly studied Caribbean con- to one another than to O. orstedii, which geners with broad dorsal arm plates and also has lecithotrophic development, but small disk stumps, Ophiothrix brachyactis, has a radically different dorsal arm plate, Ophiothrix hartfordi, and Ophiothrix platyac- 1st ventral arm plate, and spine morphol- tis. The latter three species are represented ogy. A phylogenetic analysis of use in in museums by only a few, small speci- evaluating those hypotheses might also mens, which might in fact be immature in- clarify the evolutionary connection be- dividuals. Each differs in a significant way tween hermaphroditism and brooding, from O. cimar, O. stri, O. synoecina, and O. particularly if the full range of Caribbean lineata. In addition to considerable differ- Ophiothrix could be considered. At present, ences in arm spine morphology, and the carrying out the analysis would be a daunt- prevalence of multiple stumps on disk ing challenge, since several key species are scales, O. platyactis has a scaled ventral in- known only from one or two, old, dried terradius, O. brachyactis has square ventral type specimens, and it is likely that other arm plates, and O. hartfordi has relatively species still remain to be described. large scales at the center of the disk. The eggs of O. stri and O. cimar are ap- Acknowledgments.—This contribution is proximately 0.16 mm diameter, slightly dedicated to Capt. Fred C. Ziesenhenne, smaller than those of O. lineata (X = 0.19 who was the first scientist to collect O. mm, SE = 0.002, n = 50). The egg size of the cimar and O. stri, and who established the latter species was estimated from measure- Allan Hancock Foundation echinoderm ments of oocytes in the gonads of alcohol- collection, which is now part of the Natural preserved females (X = 11.76 mm dd, SE = History Museum of Los Angeles County. 1.09, n = 5), as were the size of O. cimar and Limitations of my memory and of page O. stri eggs. However, the larger size of for- space preclude a full accounting of those malin preserved oocytes of O. stri suggests who assisted in a study pursued episodi- that all three species have egg diameters cally for 35 years. Among the most signifi- somewhat in excess of 0.2 mm. In any case, cant contributors at the Smithsonian Tropi- they fall within the size range (0.13-0.35 cal Research Institute are: R. Collin, G. mm diameter) of eggs produced by ophiu- Jacome and staff of the Bocas Research Sta- roids that have abbreviated development tion; C. Birkeland, D. Meyer, K. Meyer, J. (Hendler 1991). Stames and other colleagues at the Galeta Even if O. lineata and similar species are Marine Laboratory; and the captain and related to Placophiothrix, the characteristics crew of the R/V Stenella. I owe the success of their eggs sets them apart them from of my research in Costa Rica to M. Murillo Ophiothrix spongicola Stimpson, 1855, the and the staff of the Centro de Investiga- only member of Placophiothrix whose devel- ciones en Ciencias del Mar y Limnología opment has been studied (Selvakuma- (CIMAR) of the Universidad de Costa Rica, raswamy and Byrne 2000). Its eggs are the Costa Rican Servicio de Parques Na- 598 GORDON HENDLER cionales, and my colleagues R. Brusca, P. American coral reefs, ed. J. Cortés, 241-274. Amster- Delaney, R. Peck, and R. Wetzer. The Kuna dam: Elsevier. nation and the government of Panama per- Guzmán, H. M, K. A. Burns, and J. B. C. Jackson. 1994. Injury, regeneration and growth of Caribbean reef mitted work in the San Blas. I am also corals after a major oil spill in Panama. Mar. Ecol. grateful to R. Woollacott and F. Collier Prog. Ser. 105:231-241. (MCZ), and D. Pawson and C. Ahearn Guzmán, H. M., and C. A. Guevara. 1998a. Arrecifes (USNM), for loans of specimens, to F. Bayer coralinos de Bocas del Toro, Panamá: I. Distribu- for nomenclatorial guidance, and to F. ción, estructura y estado de conservacióndelos Nishida, D. Pawson and an anonymous re- arrecifes continentales de la Laguna de Chiriquí y la Bahía Almirante. Rev. Biol. Trop. 46:601-623. viewer for comments offered on the manu- Guzmán, H. M., and C. A. Guevara. 1998b. Arrecifes script. Support for this study was provided coralinos de Bocas del Toro, Panamá: II. Distribu- by a Walter Rathbone Bacon Scholarship, ción, estructura y estado de conservacióndelos travel funds from the Smithsonian Institu- arrecifes de las Islas Bastimentos, Solarte, Carenero tion, and a grant from the Weiler Founda- y Colón. Rev. Biol. Trop. 46:889-912. tion. Assistance at LACM was provided by Guzmán, H. M., and C. A. Guevara. 1999. Arrecifes C. Groves, G.-A. Kung, and M. Schwengel, coralinos de Bocas del Toro, Panamá: III. Distribu- ción, estructura, diversidad y estado de conser- and by the National Science Foundation’s vación de los arrecifes de las islas Pastores, Cris- MRI grant DBI-0216506 for the museum’s tóbal, Popa y Cayo Agua. Rev. Biol. Trop. 47:659- SEM facility. 676. Hendler, G. 1977. The differential effects of seasonal stress and predation on the stability of reef-flat LITERATURE CITED echinoid populations. Proc. Third International Coral Reef Symp. I: 217-223. Birkeland, C. 2004. Ratcheting down the coral reefs. Hendler, G. 1984. The association of Ophiothrix lineata BioScience 54:1021-1027. and Callyspongia vaginalis: a brittlestar-sponge Birkeland, C., A. A. Reimer, and J. R. Young. 1976. cleaning symbiosis? P.S.Z.N.I: Marine Ecology 5:9- Survey of marine communities in Panama and ex- 27. periments with oil. EPA-600/3-76-028, Natl. Tech. Hendler, G. 1991. Echinodermata: Ophiuroidea. In Re- Info. Serv., Springfield, VA. production of marine invertebrates, Vol. 6, eds. A. C. Clark, A. M. 1966. Notes on the Family Ophiotrichidae Giese, J. S. Pearse, and V. B. Pearse, 355-511. Pacific (Ophiuroidea). Ann. Mag. Nat. Hist. Ser. 13, 9:637- Grove, California: Boxwood Press. 655, pls. 10-11. Hendler, G., J. E. Miller, D. L. Pawson, and P. M. Kier. Clark, A. M., and F. W. E. Rowe. 1971. Monograph of 1995. Sea stars, sea urchins, and allies: Echinoderms of shallow-water Indo-West Pacific echinoderms. London: Florida and the Caribbean. Washington: Smithsonian British Museum (Natural History). Institution Press. Cortés, J., and M. J. Risk. 1985. A reef under siltation Hoggett, A. K. 1991. The Genus Macrophiothrix stress: Cahuita, Costa Rica. Bull. Mar. Sci. 36:339- (Ophiuroidea: Ophiotrichidae) in Australian wa- 356. ters. Invert. Taxon. 4:1077-1146. Macintyre, I. G., and P. W. Glynn. 1976. Evolution of a Cortés, J., R. Soto, C. Jiménez, and A. Astorga. 1993. modern Caribbean fringing reef, Galeta Point, Pan- Earthquake associated mortality of intertidal and amá. Am. Assoc. Pet. Geol. Bull. 60:1054-1072. coral reef organisms (Caribbean of Costa Rica). Proc.7th International Coral Reef Symp.1:235-240. Mladenov, P. V. 1979. Unusual lecithotrophic development of the Caribbean brittle star Ophiothrix oer- Cubit, J. D., H. M. Caffey, R. C. Thompson, and D. M. stedi. Mar. Biol. 55:55-62. Windsor. 1989. Meteorology and hydrography of a Mladenov, P. V. 1985. Observations on reproduction shoaling reef flat on the Caribbean coast of and development of the Caribbean brittle star Panama. Coral Reefs 8:59-66. Ophiothrix suensoni (Echinodermata: Ophiuroidea). Fonseca, A. C. 2003. A rapid assessment at Cahuita Bull. Mar. Sci. 36:384-388. National Park, Costa Rica, 1999 (Part 1: Stony cor- Robertson, D. R., and P. W. Glynn. 1977. Field guide- als and algae). Atoll Res. Bull. 496:249-257. book to the reefs of San Blas Islands, Panama. Miami: Gardner, T. A., I. M.Côté, J. A. Gill, A. Grant, and A. R. Atlantic Reef Committee, University of Miami. Watkinson. 2003. Long-term region-wide declines Schoppe, S. 1996. Ophiothrix synoecina new species in Caribbean corals. Science 301:958-960. (Echinodermata: Ophiuroidea) from the Caribbean Garth, J. S. 1945. Geographical account and station rec- coast of Colombia. Bull. Mar. Sci. 58:429-437. ords of Velero III in Atlantic waters in 1939. Rept. Schoppe, S., and A. Holl. 1994. Ophiothrix n. sp. Allan Hancock Atlant. Exped. 1:1-106. (Ophiuroidea: Ophiotrichidae) from Colombia, a Guzmán, H. M. 2003. Caribbean coral reefs of Panama: protandric hermaphrodite that broods its young. present status and future perspectives. In Latin In Echinoderms through time, eds. B. David, A. NEW SOUTHERN CARIBBEAN BRITTLE STARS 599

Guille, J. P. Féral, and M. Roux, 471-475. Rotter- roids from Australia and New Zealand. Invert. Biol. dam: Balkema. 119:394-402. Schoppe, S., and B. Werding. 1996. The boreholes of Stancyk, S. E. 1973. Development of Ophiolepis elegans the sea urchin genus Echinometra (Echinodermata: (Echinodermata: Ophiuroidea) and its implications Echinoidea: Echinometridae) as a microhabitat in in the estuarine environment. Mar. Biol. 21:7-12. tropical South America. P.S.Z.N.I: Mar. Ecol. 17: 181-186. Zoppi de Roa, E. 1967. Contribución al estudio de los Selvakumaraswamy, P., and M. Byrne. 2000. Repro- equinodermos de Venezuela. Acta Biol. Venez. 5: duction, spawning, and development of 5 ophiu- 267-333.