A new species of the peracaridan order Bochusacea, Thetispelecaris yurikago, is described from a submarine cave on Grand Cayman Island, the Caribbean Sea. The new species is the fourth species of the order and family, and the second of the genus. Recent studies have strongly suggested a close phylogenetic affinity between cave-dwelling and deep-sea taxa in the Bochusacea as recognized in other cavernicolous/deep-sea crustaceans such as amphipods and copepods. The morphology of the gut and female reproductive system is observed for the first time in the Bochusacea: the stomach is complex with structures such as ridges, processes, spinules, and hairs in the lumen; paired gonopores are located near the base of the fifth pereiopods on the sternite.
The enigmatic peracaridan order Mictacea accommodated only three species: Hirsutia bathyalis Sanders, Hessler and Garner, 1985 from a muddy bottom at a depth of 1000 m off the tropical Atlantic, H. sandersetalia Just and Poore, 1988 from a mud and silt bottom at a depth of 1500 m off southeastern Australia, and Mictocaris halope Bowman and Iliffe, 1985 from a marine cave on Bermuda, each genus of which constituted its own family, the Hirsutiidae and the Mictocarididae. The Mictacea was distinguished from the other peracaridan orders by the combination of the following features (Bowman et al., 1985): (1) the cephalon fused with the first thoracic somite, which is covered with the cephalothoracic shield; (2) lack of the posterior carapace fold; (3) eyestalks present or absent; (4) both rami of uropod multisegmented; (5) antennules with 3-segmented peduncle and 2 multisegmented flagella; (6) antennae with 5-segmented peduncle bearing scale; (7) mandibles with incisor widely separated from molar process; (8) maxillules lacking palp; (9) maxillipeds without an epipod; (10) pereio-pods I-V or II-VI with natatory exopods, without branchiae; (11) coxa-basis articulation of pereiopods monocondylic; (12) limb plane of the endopods of pereiopods bending only at merus-carpus articulation; (13) oostegites originating from the coxa of pereiopods I-V or II-VI, armed or unarmed; (14) pleopods I-V rudimentary, unisegmented, separate from or fused to the body; (15) male pleopod II modified; (16) the presence of manca larva. Phylogenetic relationships between peracaridan orders have been intensively debated since 1904 when Calman established the superorder Peracarida, and are as yet incompletely solved, partly because of the lack of a strict homologous comparison between peracaridan orders (Hessler and Watling, 1999; Watling, 1999).
Recently Gutu and Iliffe (1998) described a new hirsutiid, Thetispelecaris remex Gutu and Iliffe, 1998, from anchialine and submarine caves of Bahamas, and established a new order, Bochusacea to accommodate the family Hirsutiidae. Simultaneously Gutu (1998) erected a new order, Cosinzeneacea to contain the family Mictocarididae together with the Spelaeogriphacea. Consequently the order Mictacea proposed by Bowman et al. (1985) was ranked as a suborder of the Cosinzeneacea (Gutu, 1998).
During a faunal survey in 2000 of submarine caves on Grand Cayman Island in the Caribbean Sea, the third author discovered a bochusacean and a thermosbaenacean crustaceans. The present paper describes the new species of the Bochusacea and comments on phylogenetic relation-ships between cavernicolous and deep-sea crustaceans. The morphology of the gut and the female reproductive system are also described.
MATERIALS AND METHODS
Bochusaceans were collected in a submarine cave named by local SCUBA divers as “The Mouse Trap” (see Davis, 1999). The cave is located in a coral reef approximately 400 m offshore on the west side of the Grand Cayman Island (19° 19′ 14.8″ N, 81° 23′ 22.8″ W). The cave opens at depth of about 10 m with an entrance ca. 1 m high and ca. 2 m wide, and extends toward the island. The cave consists at least of two parts: the first part is an almost horizontal, straight tube ca. 15 m long and the second part is a crevicular tube of unknown length. The two are connected by a narrow passage that spirals down to the bottom about 34 m deep. The cave was filled with normal seawater at the time of collecting, but Davis (1999) mentioned that there is a slight outflow of green fresh water during the rainy summer months. Bochusaceans inhabited the middle to the end of the first part, where the floors were covered with a substantial amount of mud and it was totally dark. They were collected with hand-held fine-mesh nets (mouth diameter 30 cm; mesh size 0.35 mm) by brushing the under-surface of rocks on the bottom sediments. The wall and ceiling of the cave were covered densely with tube-dwelling annelids. Other animals were scarce, except for some acroid bivalves.
The bodies and appendages of the new bochusaceans were observed in lactophenol and gum-chloral media, respectively, with a differential interference/phase contrast microscope (Nikon Optiphoto). Type specimens are deposited at the National Science Museum, Tokyo (NSMT-Cr). Four partly damaged adult females from the same locality were rinsed with distilled water, dehydrated by ethanol series, critical point dried, sputter-coated with gold, and then observed with a scanning electron microscope (Jeol T-20). Body lengths were measured from anterior tip of rostrum to posterior tip of telson excluding setae. Terminology follows Bowman and Iliffe (1985) and Sanders et al. (1985). The internal structures of the stomach are tentatively identified by reference to those of isopods (Schmitz, 1992), mysids (Nath and Pillai, 1973) and euphausiids (Suh and Nemoto, 1988). Reproductive systems are referred to those of Thermosbaenacea (Monod and Cals, 1999).
Order Bochusacea Gutu and Iliffe, 1998
Remarks.-Gutu and Iliffe (1998) have pointed out differences in the structures of pereiopods and uropods between the two families, Hirsutiidae and Mictocarididae in the Mictacea. This resulted in the recognition of a new order Bochusacea to accommodate the former family. We follow Gutu and Iliffe (1998), emphasizing the following features: (1) the displacement of oostegites (= epipods) bearing plumose marginal setae, as well as the setal appearance in early stages, (2) pereiopod I modified into a feeding appendage; (3) the presence of long palpiform elongations of the paragnath. However, the validity of these two orders should be rigorously reconsidered, since phylogenetic relationships among all peracaridan orders are still incompletely revealed (cf. Hessler and Watling, 1999; Watling, 1999).
Family Hirsutiidae Sanders, Hessler and Garner, 1985
Type genus.-Hirsutia Sanders, Hessler and Garner, S. Ohtsuka et al. 1985.
The identity of rami originating from the posterior surface of coxae of pereiopods II-VI has been disputed since the discovery of the Hirusutiidae (Sanders et al., 1985; Just and Poore, 1988; Gutu and Iliffe, 1998): are these epipods or oostegites? Sanders et al. (1985) pointed out a unique (not medial but posterior) configuration of the rami, and tentatively regarded these as oostegites in accordance with Bowman and Iliffe (1985). Just and Poore (1988) suggested a similar configuration between “oostegites” of hirsutiids and corophiid amphipods, both of which are characterized by the cylindrical shape of the body, and supposed that the uniqueness may be due to the body shape. However their opinions were not yet conclusive, because of no observation of juvenile stages. Gutu and Iliffe (1998) first found that juveniles of Thetispelecaris remex also bear these rami, and considered that these could play roles both in respiration and brooding. In addition they thought that eggs/embryos might be retained at most by plumose setae of the rami without formation of a well developed marsupium as in other peracarids. Our observation strongly supports their idea, and we use epipods throughout this paper.
Genus Thetispelecaris Gutu and Iliffe, 1998
Type species.-Thetispelecaris remex Gutu and Iliffe, 1998.
Other species.-Thetispelecaris yurikago n. sp.
Diagnosis (emend.).-Cephalon with sharply pointed rostrum. Antennule with 3-segmented peduncle and 3- or 4-segmented flagella. Antenna bearing 2-segmented protopod, unisegmented exopodal scale, and multiseqmented endopod comprising 3-segmented peduncle and 5- or 6-segmented flagellum. Palpiform extensions of paragnath covered with long hairs. Pereiopod I with exopod but lacking epipod; basis and ischium of pereiopods II-VI incompletely or almost completely fused. Pleopods IV and V unisegmented, articulated with body. Uropodal protopod with single inner spine.
Remarks.-The genus Thetispelecaris was briefly defined by Gutu and Iliffe (1998), but the discovery of the new species described below strengthens the generic validity. Thetispelecaris differs from Hirsutia in: (1) pereiopod I with a 2-segmented exopod (absent in Hirsutia); (2) the numbers of exopodal segments of pereiopods II-VI are 3 (4, 3), 4 (4, ?), 4 or 5 (4, ?), 4 or 5 (4, 4), and 4 (3, ?), respectively (numbers in parenthesis in H. bathyalis and H. sandersetalia, respectively); (3) the reduced number of terminal setae/spines on carpus and propodus of pereiopod II (4 and 7 in Thetispelecaris vs 6 and 12 in Hirsutia, respectively); (4) unisegmented pleopods IV and V, distinctly articulated with the body (completely incorporated into the body in Hirsutia); (5) the reduced numbers of elements on the uropodal protopod and endopod (single spine on each segment in Thetispelecaris vs 2 or 3 spines on protopod and 6 inner spines on the first endopodal segment in Hirsutia); (6) a palpiform lobe of the paragnath with numerous hairs along the entire length (only setose in the basal part in Hirsutia). The body size may also be diagnostic because Thetispelecaris (T. remex 1.2–1.6 mm; the new species described below 1.30 −1.78 mm) is much smaller than Hirsutia (H. bathyalis 2.7 mm; H. sandersetalia 3.3 mm). On the other hand, there are several differences between two congeners of Hirsutia: (1) a rostrum is present in H. sandersetalia but absent in H. bathyalis; (2) epipods are furnished with many long plumose setae in H. sandersetalia where in H. bathyalis fewer spiniform setae. However the epipods of H. bathyalis may not be fully developed because the only one individual examined (holotype) of the species is a “preparatory female” (Sanders et al., 1985).
The habitats of these two genera are also different. Hirsutia is a deep-sea taxon, while Thetispelecaris is cavernicolous.
Thetispelecaris yurikago n. sp.
Material examined.-Twenty adult females, collected from a submarine cave named “The Mouse Trap,” the Grand Cayman Island, November 15 and 19, 2000.
Types.-Holotype: ♀, whole specimen, NSMT-Cr 14298. Paratypes: 2 ♀ ♀, dissected and mounted on glass slides, NSMT-Cr 14299; 13 ♀ ♀, whole specimens, NSMT-Cr 14300.
Body length.-Range 1.30–1.78 mm (mean±standard deviation=1.50±0.13 mm, N=16); Holotype 1.50 mm.
Description based on holotype and paratypes dissected. Female. Body (Figs. 1A, 6A) nearly cylindrical, but somewhat depressed; cephalothorax slightly longer than wide, subrectangular; rostrum (Fig. 1B) produced anteroventrally or anteriorly, acutely pointed at tip (see Fig. 6B); lateral carapace folds poorly developed, only covering bases of postmandibular appendages (Fig. 6C); dorsal car-apace fold absent. Pereion shorter than pleon; pereionite I small; pereionites II and III equal in length; pereionites IV-VI longer than pereionites I and II; pereionite VII slightly shorter than preceding pereionite. Pleonites (Fig. 1A, D, H) gradually increasing in length posteriorly; pleonite VI expanded posterolaterally; telson (Fig. 1G) tongue-like with concavity at apex, one-fifth wider than long, bearing 4 lateral pairs of simple spiniform setae, 1 terminal pair of setae, and 1 dorsal pair of slender setae; anus (Fig. 1D) located ventrodistally, longitudinal slit-like, with lateral valves (Fig. 6D).
Antennule (Fig. 2A, B) anteriorly directed, with 3-segmented peduncle and 4-segmented flagella of subequal length; peduncular segment 1 longest, bearing 2 proximal setae directed posteriorly, segment 2 with 6 terminal setae of unequal length, segment 3 with 4 long subterminal setae and 3 minute setules on terminal rectangular projection (Fig. 2B); setal formula of inner flagellum 1, 5, 1, 4; outer flagellum with setal formula of 0, 1, 1, 4; aesthetasc present on segments 2 and 3 of outer flagellum.
Antenna (Fig. 2C) with 2-segmented protopod (peduncle), unisegmented scaphocerite (exopodal scale), and 9-segmented endopod comprising 3-segmented peduncle and 6-segmented flagellum; exopodal scale with 6 marginal setae; setal formula of endopod 1, 6, 12, 0, 4, 2, 3, 2, 5.
Labrum (Fig. 2H: slightly depressed in preparation) with numerous spinular rows on anterior surface; terminal portion truncate, slightly concave midway, ornamented with fine setules along distal margin.
Mandibular gnathobase (Fig. 2D, F, G) complex in structure; incisor and molar processes widely separate; incisors multicusped; left gnathobase with lacinia mobilis bearing spinules subterminally; 3 subequal stout setae near lacinia mobilis; 12 curved setae originating from round protuberance; right gnathobase with 1 stout spinulose spine and 15 long setae along inner margin; molar process (Fig. 2D, G) protruded from gnathobase, heavily chitinized only terminally, with several lamellar plates and grinding surface. Mandibular palp (Fig. 2E) 3-segmented, slender; segment 1 small, unarmed; segments 2 and 3 almost equal in length; segment 2 unarmed; segment 3 with 1 naked and 2 spinu-lose setae at tip.
Paragnath (Fig. 2I) bilobed; each lobe tapering distally, terminating in elongate hirsute projection (Fig. 8A); each lobe with inner spinulose round projection and 3 stout spines of unequal length midway.
Maxillule (Figs. 2J, 8C) bilobed, with numerous rows of minute spinules; inner lobe with 4 comb-like setae and 1 spinulose spine fused to lobe at base along inner margin; outer lobe curved inward at midlength, with 2 surface setae and 2 groups of marginal elements, proximal group of which consisting of 7 chitinized spines, distal group of 4 terminally curved spines and 6 spinulose long setae.
Maxilla (Fig. 3A–C) with quadrate protopod bearing 21 “pushing setae (with bulbous base)” (setal nomenclature following Fryer, 1965) and 9 spiniform setae along proximal inner margin; distal inner corner of protopod having 9 stout spines of various ornamentations and 8 setae; ventral surface almost entirely covered with short hairs and bearing irregularly longitudinal row of short spinules; distal margin of protopod near common base of basal endites with stout ser-rate spine; basal endite 1 with about 23 comb-like setae (Fig. 8B); basal endite 2 with 3 rows of spatulate setae (approximately 11, 9 and 8 setae, respectively) and longitudinal row of fine spinules.
Maxilliped (Fig. 3D, E) with elongate protopod; proximal part of protopod with 2 proximal, 1 middle spiniform and 1 plumose setae on surface and 1 minute spine and 9 well developed setae along inner margin; penultimate inner seta with serration at expanded tip; basal endite protruded distally, 7 hooked and 7 serrate spines and 1 spinulose and 1 plumose seta (Fig. 8D); palp 5-segmented, longer than basis, ischium small, with 1 seta and row of spinules along outer distal margin; merus about twice as long as ischium, with single seta; carpus as long as proximal 2 segments combined, bearing 2 middle, 2 subterminal and 1 minute distal setae; propodus inserted to preceding segment at angle of about 120°, with 6 distal setae; dactylus directed inwards, with 3 spiniform and 2 fine setae terminally.
Configuration of pereiopods shown in Fig. 4A; pereiopod I directed anterolaterally, modified into mouthpart appendage; pereiopods II-VI with origins of exopod, endopod and epipod aligned along anterior-posterior direction paralleling long axis of body. Pereiopods bearing short coxa and long basis, furnished with numerous spinular rows (Figs. 7B–D, 8E); pereiopod I (Fig. 4A, B) biramous, lacking epipod; pereiopods II-VI (Figs. 4A, C–E, 5A–C) biramous, each with epipod; all pereiopods each with 5-segmented endopod (ischium, merus, carpus, propodus, dactylus) and 2-(I), 3-(II), 4-(III, VI) or 5-(IV, V)segmented exopod; ischium incompletely (pereiopod II) or almost completely (pereio-pods III-VI) coalescent to basis; pereiopod VII (Figs. 4A, 5D) with basis separate from ischium, uniramous, lacking exopod and epipod. Within brooding chamber surrounded by both epipods of pereiopods II-VI only one droplet-like embryo remaining in holotypic female (Fig. 4A); embryo approximately 0.06 mm in width.
Pereiopod I (Fig. 4B) with basis bearing 1 anterior and 3 posterior setae; ischium having 1 posterior and 2 anterior setae; merus short, with 1 short posterior and 2 anterior setae; carpus elongate, bearing 3 anterior setae increasing in length distally; propodus as long as carpus, with 3 subterminal setae; dactylus short, with 1 stout spine terminally and 2 spiniform and 5 setae subterminally; exopod short, 2-segmented, setal formula 1+1, 2.
Pereiopod II (Fig. 4C, D) longest and thickest of pereio-pods (see Fig. 4A), covered entirely with spinular rows, much more than in other pereiopods; epipod relatively short, bearing 4-5 marginal plumose setae; basis bearing 1 anterior and 4 posterior setae distally; ischium with 2 nearly equal setae terminally; merus slightly longer than ischium, bearing 2 posterior and 4 anterior setae distally; carpus longest, with 1 anterior and 2 posterior stout spines in addition to distal anterior seta; propodus with 2 posterior and 2 anterior long, stout spines distally and 1 short serrate spine and 2 setae at anterodistal corner; dactylus with 2 stout spines of unequal length and 1 spiniform and 1 slender seta terminally (see Fig. 8E); exopod 3-segmented, setal formula 1, 1+1, 2 (see Fig. 7C).
Pereiopod III (Fig. 4A, E) shorter than preceding leg; epipod more developed than preceding one, with 8-9 setae in total; basis with 2 proximal and 4 terminal setae along posterior margin; ischium with 4 setae of unequal length terminally; merus bearing 3 terminal setae; carpus with 4 posterior marginal and 4 anterior terminal setae and 5 serrate surface spines increasing distally in length; propodus bearing 3 posterior and 2 surface serrate spines and 6 terminal setae; dactylus with 1 fine anterior middle, 1 short subterminal and 1 long terminal seta; exopod 4-segmented, setal formula 2, 1+1, 1+1, 2.
Pereiopod IV (Figs. 4A, 5A) similar to preceding leg; epipod developed as in preceding one, with 12 setae in total; basis with 2 proximal, 1 surface and 3 terminal setae; ischium with 4 setae terminally; merus with 2 posterior setae and 1 anterior short spiniform seta; carpus bearing 4 posterior marginal and 3 anterior terminal setae and 5 serrate spines increasing gradually in length distally; propodus with 3 posterior marginal and 1 surface serrate spines and 5 terminal setae; dactylus with 1 fine anterior seta and 1 short and 1 long seta terminally; exopod 5-segmented, setal formula 2, 1, 1+1, 1+1, 2.
Pereiopod V (Figs. 4A, 5B) similar to pereiopod IV, but different in armature of endopod: ischium bearing only 3 setae terminally; merus lacking anterior spiniform seta; carpus with 5 posterior setae terminally in spite of 3 anterior setae; propodus having only 3 spines along posterior margin and 4 setae terminally.
Pereiopod VI (Figs. 4A, 5C; 7D also) with epipod bearing 10–13 marginal setae; basis incompletely fused to ischium with suture more clearly visible than in preceding legs, bearing 4 middle and 2 terminal setae; ischium, merus and carpus bearing 2, 2 and 4 distal setae, respectively; propodus longest, bearing 5–6 serrate inner spines and 1 spini-form middle and 2 distal setae along outer margin; dactylus with 2 unequal spiniform setae and 2 fine setules terminally and 1 short seta subterminally.
Pereiopod VII (Figs. 4A, 5D; 7D also) with basis bearing 3 anterior and 2 posterior setae; ischium with 1 posterior terminal seta; merus as long as ischium, bearing 2 setae terminally; carpus with 2 posterior and 2 anterior setae; propodus having 3 serrate spines along posterior margin and 1 middle and 2 terminal setae along anterior margin; dactylus with 2 spiniform and 2 short setae terminally, serrated along posterior margin.
Pleopods I-III (Figs. 1D, E) rudimentary, represented by triangular lobe with 1 terminal and 1 outer basal seta (see Fig. 9A); pleopods IV and V (Fig. 1D, F) distinctly articulated at base, unisegmented, bearing 1 inner and 1 outer seta midway and 1 terminal and 1 subterminal seta (see Fig. 9B).
Uropod (Fig. 1A, D, G) biramous; protopod approximately 2.8 times longer than wide, with inner distal spinu-lose spine, 3 outer setae and 2 dorsal surface setae; exopod 2-segmented, proximal segment with fine seta and 2 spines along outer margin and single fine inner spine, distal segment bearing 2 short outer, 2 long inner and 3 unequal distal setae; endopod about 1.3 times as long as exopod, 5-segmented, first segment with 6 outer setae and 1 distal and 1 middle spine, second to fourth segments each with 3 outer setae of unequal length and inner distal spine; fifth segment bearing single outer seta, 4 setae and 1 spinulose spine subterminally along inner margin, and 3 long setae terminally.
Stomach (Fig. 1C) with complex structure, and difficult to observe in detail with light microscope; some elements observed, i.e., lateral ampullae (?) with fine hairs, lateral teeth (?) with acutely pointed processes, ventral pyloric ridge(?) with fine hairs posteriorly, and 4 longitudinal rows of minute spinules. Gut posterior to stomach almost uniformly straight with constriction near anus (Fig. 1A, H).
Female reproductive system tentatively identified as follows: ovary(?) in cephalothorax paired; oviducts parallel to gut (Fig. 1A, H); seminal receptacle(?) present in pereions V and VI; oviduct(?) directed anteroventrally to pereion V, opening at base of pereiopod V. In a paratype 3 pairs of fully mature eggs present at posteriormost part of oviduct (Fig. 1H). Paired gonopores opening near bases of pereiopods V (sixth thoracic somite) on posteroventral surface (Fig. 7A, B).
SEM observation.-Four partly damaged adult females were observed with scanning electron microscopy (Figs. 6Fig. 7Fig. 8-9). Since most SEM micrographs complement the above description, only selected points are mentioned here. The mandibular gnathobase is expanded laterally with a honeycomb structure and numerous hairs (Fig. 6C, E). Gonopores were difficult to see under the light microscope, and could be observed with SEM only in the specimens whose pereio-pods were cut off. The pores are located near the base of pereiopods V (on the sixth thoracic somite) on the sternite as in other malacostracans (Wilson, 1991) (Fig. 7A, B). The gonopore is slit-like, approximately 0.02 mm long, although the specimens examined were highly modified during dehydrolation and critical-point drying. The coxae of pereiopods are clearly separate from the body unlike in Hirsutia bathyalis (Fig. 7D). Setal elements on the mouthpart appendages are greatly diversified (Fig. 8A–D). Unisegmented pleopods IV and V are distinctly separate from the body (Fig. 9B).
Remarks.-The present new species is similar to Thetispelecaris remex from caves of Bahamas, but can be distinguishable from it by the following features: (1) the shape of the carapace (more narrowed anteriorly in T. remex); (2) the antennal scale bears 6 setae (only 4 in T. remex); (3) the maxillipedal palp carries 3 spiniform and 2 fine setae terminally (only 3 slender setae in T. remex); (4) the number of exopodal segments of pereiopods IV and V is 5 (4 in T. remex); (5) the armature elements on pereiopods I to VII; (6) pereiopods III to VI with basis and ischium almost fused (separate in T. remex); (7) pleopods IV and V with 2 setae terminally (3 in T. remex).
All 16 type specimens examined of the new species were identified as female on the basis of the presence of eggs within the body or of an embryo in the marsupium. Although reproductive strategies of the Bochusacea are totally unknown, no record of males in the order may suggest parthenogenesis, as known in another peracaridan group, the Tanaidacea (Schram, 1986). This idea is supported also by a fact that 16 individuals of Thetispelecaris remex all were females. No male is known also in the genus Hirsutia (Sanders et al., 1985; Just and Poore, 1988). In contrast the male is known in the Mictocarididae (which formerly belonged to the Mictacea but is now assigned to the suborder within the Cosinzeneacea). In the holotypic female of the new species only a single embryo remains within the marsupium formed by plumose setae of the epipods (oostegites). Others were possibly lost during collection and/or processing, because three pairs of fully matured eggs were observed in the oviduct of a paratypic female and would probably be released simultaneously into the marsupium.
Etymology.-The specific name “yurikago.” (Japanese, meaning baby cradle) alludes pereiopods II-VI with a highly developed epipod which bears more numerous marginal setae than in Thetispelecaris remex and seems to function in brooding.
Close phylogenetic relationships of marine crustaceans living in shallow-water caves and in the deep-seas have attracted much attention from biologists (Stock and Vermeulen, 1982; Iliffe et al., 1984; Hart et al., 1985; Stock, 1986, 1993, Wilkens et al., 1986; Boxshall, 1989; Boxshall and Jaume, 2000). The origins of cavernicolous forms have been intensively debated by these and other authors. Some groups such as thaumatocyridid ostracods (Wilkens et al., 1986), misophrioid copepods (Boxshall and Jaume, 2000), pardaliscid and ingolfiellid amphipods (Stock and Vermeulen, 1982; Stock, 1986, 1993; Wilkens et al., 1986) and galatheid decapods (Calman, 1912; Wilkens et al., 1986) clearly show phylogenetic affinities between taxa from both habitats. Previous studies and present results reveal that the Bochusacea exhibits a close phylogenetic relationship between the cavernicolous and deep-sea inhabitants.
Some cave-living invertebrates were previously supposed to have originated from the deep-sea (Iliffe et al., 1984; Boxshall, 1986; Wilkens et al., 1986). However, Stock (1986) has strongly rejected their deep-sea origin theory on the basis of paleobiogeography, and proposed that the common ancestor of both lineages could have been distributed in shallow waters, and then invaded both habitats and evolved independently. This idea is supported also by a detailed cladistic analysis of misophrioid copepods (Boxshall and Jaume, 2000). The misophrioid genus Misophriopsis is distributed not only in hyperbenthic regions from deep oceanic (3000 m) to shallow coastal waters (less than 10 m), but also in caves (Boxshall and Jaume, 2000). This range extension is also explained by horizontal/vertical explorations of the shallow-water hyperbenthic ancestor (Boxshall and Jaume, 2000). According to Stock (1986, 1993), the anoxia in the deep waters (200–2000 m) in the Atlantic, which occurred in the Oligocene/Miocene boundary, would have entirely wiped out the Tethyan fauna. He argued that the deep-sea origin for cave-dwelling animals is therefore unlikely, and that the present cavernicolous and deep-sea inhabitants probably evolved from the shallow-water ancestor during or after the Miocene when the deep waters were reventilated (Stock, 1986, 1993).
It is noteworthy that both Thetispelecaris and Hirsutia exhibit very similar mouthpart structures. This suggests that both taxa essentially utilize similar types of food items irrespective of their different habitats. Sanders et al. (1985) supposed, on the basis of mouthpart and pereiopod morphology, that Hirsutia is a facultative carnivore. Fryer (1965) carefully analyzed the feeding mode of the Thermosbaenacea from a functional morphological viewpoint, concluding that Thermosbaenacea is not a filter feeder but employs a feeding mode consisting of scraping, brushing and pushing for detritus. Since the mouthparts of Bochusacea are basically similar to those of Thermosbaenacea (Bowman and Iliffe, 1985), the feeding of the latter is presumably similar to that of the former, i. e. bochusaceans are deposit feeders, as already pointed out by Just and Poore (1988). This is also supported by the new observation that the guts of 16 specimens of Thetispelecaris yurikago were full of fine particles. Just and Poore (1988) reported the gut of H. sandersetalia collected from 1500 m depth to be packed with fine-grained sedimentary material. In both genera, the well-developed, long pereiopod II with heavily sclerotized serrate spines terminally, may be used in grasping and manipulating large detrital particles in association with the anteriorly directed pereiopod I rather than holding prey (cf. Sanders et al., 1985; Just and Poore, 1988).
We express our sincere thanks to Drs. Geoffrey A. Boxshall, Robert R. Hessler, and Gary Poore for giving us valuable comments on the manuscript. We are also indebted to Dr. Gina EbanksPetrie for permitting our survey on Grand Cayman Island, and to Mr. John Slaspinsky for cooperation at sea. This is a contribution to the DIVERSITAS-IBOY project entitled “Exploration and Conservation of Anchialine Faunas”.