Exoskeleton, Distribution, and Movement of the Flexible Setules on the Myodocopine (Ostracoda: Myodocopa) First Antenna

Home , Conchoecia, Cypridinidae, Halocypridina, Myodocopa, Myodocopida

EXOSKELETON, DISTRIBUTION, AND MOVEMENT OF THE FLEXIBLE SETULES ON THE MYODOCOPINE (OSTRACODA: MYODOCOPA) FIRST ANTENNA

Andrew R. Parker

ABSTRACT

The halophore, halothalium, and s-seta are termed herein and occur on the myodocopine first an-

tenna, probably the most systematically significant myodocopine appendage. The morphology of Downloaded from https://academic.oup.com/jcb/article/18/1/95/2418948 by guest on 29 September 2021 the exoskeleton and distribution of these structures are described, following scanning electron and light microscopic analyses and literature comparisons. The movements of these and other myodocopine first antennal structures are studied, using video recordings of an exemplary species. Halo- phores are setules with a characteristic exoskeletal ultrastructure comprising a layer of very fine rings, with walls approximately circular in cross section, covered by an outer, probably elastic, layer or sheath. There is a single pore at the halophore tip. This organization permits great flexibility. If dendrites innervate halophores, the flexibility may aid in sampling for water-borne chem- icals or mechanoreception. The s-seta (formerly the sensory seta) is a seta arising from the fifth article of the myodocopine first antenna that frequently possesses numerous long halophores. The s- seta is always and only present in the Myodocopina. The collective halophores distributed over the whole of one first antenna are termed the halothalium. Analogies of the structures described in this study are made with other crustacean structures.

The ostracod suborder Myodocopina, of at least some of the first antennal setae was within the order Myodocopida, comprises five believed to be tactile (Vannier and Abe, 1993). extant families; Cypridinidae, Cylindrole- In addition, the function of the seta arising berididae, Philomedidae, Rutidermatidae, and from the fifth article has long been designated Sarsiellidae (Kornicker, 1975). Myodoco- as sensory (e.g., Skogsberg, 1920; Sars, 1922). pines occur in marine or brackish environ- However, no such evidence has been pre- ments world-wide at all depths (Cohen, sented to substantiate these sensory claims. 1982). Most (including juveniles) can swim, Presumably, based on morphological simi- but are benthic or epibenthic (only 3-6 of the larities, original assumptions were made from 26 described cypridinid genera are wholly or comparisons with homologous or convergent mostly planktonic) for much of their lives structures in other crustacean taxa. Poulsen (Cohen, 1982). They all exhibit sexual di- (1962) followed the terminology for myo- morphism (Cohen and Morin, 1990). They docopine limbs as established by previous are "filter-feeders" (Kornicker, 1975), or more myodocopine workers, but remained uncon- accurately comb-feeders, because they typi- vinced of the implication about their function. cally operate at low Reynolds numbers (Co- hen, 1989), scavengers, active predators, "col- Morphology of the Myodocopine First An- lectors" (collect detritus from the sediment; tenna.-The myodocopine first antenna con- Walker, 1972), detritus feeders (Cannon, sists of eight articles (some occasionally 1933; Kornicker, 1975), or parasites (Bennett fused). The more proximal articles bear var- et al., 1997). The first antenna is probably the ious (often short) setae, and three of the four most significant appendage in myodocopine distal articles (namely 5, 7, and 8) bear about systematics. seven relatively long setae. It is these longer setae that project through the anterior end of Previous Terminology.-The first antenna the open carapace and probably reach out be- (antennule) of Myodocopina includes about yond the boundary layer of flow (ostracods seven long terminal and subterminal setae of typically operate only at low Reynolds num- previously unconfirmed function (Fig. 1 ). The bers; see Koehl and Strickler, 1981). These first antenna of vargula hilgendorfzi Poulsen setae are historically termed: the sensory seta (Cypridinidae) extended when the animal was (arising from the fifth article); the a- (usu- resting on the substrate. Thus, the function ally short), b- (usually mid-long) and c-setae Downloaded from https://academic.oup.com/jcb/article/18/1/95/2418948 by guest on 29 September 2021

Fig. I. Azygocypridina lowryi, adult female, left first antenna, medial view (eighth article not in view). New terminology (for explanation see discussion section) is used in the labeling; traditional terminology of s-seta is within parentheses.

(from the seventh article); and the d-, e-, f-, and glion to the muscles, but other antennulary g-setae (from the eighth article) (Fig. 1). Some nerve fibers terminate in cells in the postero- setae are missing in a few myodocopine taxa. medial corner of the ganglion. From here, Most are annulate, and may bear spines, se- other fibers extend to the distal end of the tules, or "suckers" (the last in males only). The ganglion, passing out as a bundle running to sensory seta of the fifth article usually appears the tip of the limb. Between these two bun- morphologically distinct. This seta of medium dles of nerves lies a group of four or five length is sometimes widened (or bulbous) at bipolar cells, giving rise to medium-sized gi- the base or has a separated basal section, and ant fibers (Cannon, 1931). The antennulary usually has at least six long thin flexible se- nerves are similar in the cypridinid Giganto- tules. These setules, previously unconvention- cypris mulleri Skogsberg (see Cannon, 1940). ally termed filaments (see Watling, 1989) have However, nothing more detailed regarding the been described as unringed, although in the lit- precise innervation of the first antennal setae erature they are often illustrated with stippling, has been published. giving them a granular appearance (e.g., Poulsen, 1962). This is how they appear at Aims.-The aims of this study are to: (1) de- magnifications less than about 250x. scribe the morphology of the halophore ex- This study demonstrates that the setules of oskeleton using the scanning electron micro- the b- to g- and the sensory first antennal se- scope (SEM), (2) describe the distributions of tae of Myodocopina share a characteristic ul- halophores within myodocopine taxa using trastructure and all are termed "halophores" the literature and light microscopy, (3) de- herein. The sensory seta is termed the "s-seta." scribe the movement of the myodocopine first The collection of halophores on one first an- antennal setae and setules using video record- tenna is termed the "halothalium." The terms ings, (4) compare the flexible setules of the halophore, s-seta, and halothalium, designated myodocopine first antenna with other crus- in this paper, are used hereafter to avoid un- tacean structures, and (5) review the litera- necessary confusion (an explanation of these ture on, and standardize the terminology of, terms can be found in the discussion section). myodocopine first antennal setae/setules. Techniques suitable for the study of nervous Myodocopine First Antennal Nervous Sys- structures are not employed. tem.-The nervous supply of the first antenna of the cypridinid Doloria levis Skogsberg MATERIALS AND METHODS arises from two swellings on the ventral side Specimens Examined.-Living specimens of Azygo- of the forepart of the nerve ring, marking the cypridina lowryi Kornicker were collected in single- deuterocerebral part of the brain (Cannon, chamber baited traps designed by Keable (1995). These 1931). The antennulary basal ganglion is the were set overnight at depths of 200 m (34°31.48'S, 151°13.22'E) and 300 m (34°31.80'S, 151°15.60'E) off largest in the body of D. levis and occupies the Wollongong (New South Wales, Australia) coast. The most of the basal article. What appear to be ostracods were removed from the traps and immediately motor nerves pass directly through the gan- transported to the laboratory in fresh aerated sea water. Table 1. Material examined (museum specimens fixed in 7% Formalin solution and preserved in 70% ethanol). AM represents Australian Museum (Sydney); NMNH represents National Museum of Natural History (Washington, D.C.). Downloaded from https://academic.oup.com/jcb/article/18/1/95/2418948 by guest on 29 September 2021 Behavioral observations were carried out within 4 h, during which time the ostracods appeared to be in a healthy condition. Preserved museum specimens (fixed in 7% Formalin solution and preserved in 70% ethanol) of ostracods and other crustaceans were examined (Table 1).

Anatomical Study.-For SEM examination, preserved museum specimens (Table 1 ) were cleaned, using 5 half-second exposures to ultrasound, critical-point dried, using a BioRad CPD 750, then coated with gold. These specimens were examined using a Cambridge Instruments S 120 SEM. Fig. 2. Diagrammatic sections of the halophore exo- Accidentally broken halophores of A. lowryi were stud- skeleton. A, longitudinal section (shaded areas represent Downloaded from https://academic.oup.com/jcb/article/18/1/95/2418948 by guest on 29 September 2021 ied to view the internal surface of their exoskeletons. In cross sections of the walls of the rings); B, cross section some cases, the rings which comprise the exoskeleton of through t-t. the halophores were themselves broken, and their internal architecture was opportunely examined. The setae and setules of myodocopine limbs other than the first antennae were also examined under the SEM to The bases of the halophores are not finely an- search for external morphologies similar to those of halo- nulate, except where they arise terminally on phores. Azygocypridina lowryi was mainly used in this a seta. Each halophore bears a thin outer layer investigation because of its relatively large size (carapace (sheath) (Figs. 2, 3F), which follows the length about 11 mm). Preserved museum specimens of grooved external contours of the halophore species from other ostracod taxa and crustacean classes (Table 1) were examined similarly. set by the rings. Some myodocopines bear first antennal setae (other than the b- to g- or Behavioral Study.-Video recordings were made of sev- s-setae) with unringed setules, although these eral living specimens ofA. lowryi in a large Petri dish (19- setules are evenly tapered and much thinner cm diameter) containing fresh sea water (3-cm depth) to determine the movement and action of the first antennal than halophores (appearing almost like long setae/setules. Lighting was provided by a fiber-optic source spines, though sometimes flexible to some ex- about 15 cm from the ostracods. Behaviorial observations tent). Conchoecia belgicae Miiller (Myo- were recorded, using an Olympus SZH stereomicroscope docopa: Halocypridina) bears terminal first an- connected to a Panasonic FIO video camera. tennal setae with "wrinkled" surfaces which

Literature Comparisons.-Illustrations of first antennae may indicate the presence of fine rings be- and s-setae from different myodocopine species were ob- neath. However, these setae appear to have col- tained from the literature. These were compared to dem- lapsed walls, which were not observed in halo- onstrate the gross morphological variation in halophores, phores under the same conditions (see ultra- and the variation in distribution of halophores on the first antennae. structure section of Materials and Methods). The b-, c-, f-, g-, and s-setae (and some- RESULTS times the "bases" or shafts of the d- and e- setae) exhibit an exoskeleton comprising Exoskeletal Morphology large connected sclerotized rings (with walls Of the material examined in this study approximately rectangular in cross section). (Table 1), only the myodocopines exhibited A few setules on the b- and c- first antennal setules, confined to the b- to g- and s-setae setae of male cypridinids bear "suckers" (see of their first antennae, with the same ex- Kornicker, 1983; the function of these cup- oskeletal morphology as illustrated in Fig. 2. or disc-shaped structures, with no associated These setules, or halophores, arise terminally musculature, has not been verified). These se- or nonterminally on the first antennal setae. tules are annulate with ring widths greater The halophores are composed, throughout than 1 pm. The coarsely annulate first an- most or all of their length, of a series of fine tennal setae and setules are rigid in compar- rings (circular in cross section) lying side-by- ison with the flexible halophores. side (Fig. 4A). There is a minute pore at the Internal Structures tip of each halophore (Fig. 3E). The rings of the halophores range from about 110-670 nm In addition to confirming the ring-type con- in thickness within Myodocopina (e.g., Fig. struction of their exoskeleton, broken halo- 5; Table 2) and are themselves either hollow phores revealed thin interior structures, of un- or contain a material which is less optically known ultrastructure, running approximately dense than that of the outer section (Fig. 4B). centrally through the halophores (Fig. 6). Downloaded from https://academic.oup.com/jcb/article/18/1/95/2418948 by guest on 29 September 2021

Fig. 3. Azygncypridina lowryi, adult female. A, whole animal, left carapace removed: al = first antenna; 7th = seventh limb; B, tip (articles 5-8) of right first antenna, medial view: s = s-seta; C, bases of halophores of the s-seta (note most proximal sections without fine annulations); D, midsection of halophores of the s-seta; E, tip of a halophore of the s-seta showing pore (arrowed); F, split (or dried) halophores of the s-seta showing outer layer (proba- ble elastic-type material, sheath, covering rings; arrowed). Scales: A = 2 mm; B = 500 11m; C-F = 10 11m.

Movements of Cypridinid First - Antennal Setae muscles attached to the sixth, seventh, and eighth articles, these setae move toward each Video recordings of a first antenna of a liv- other, so that all become parallel, then quickly ing specimen of A. lowryi revealed that the return to their original relaxed positions (Fig. c- and s-setae move in an equal and opposite 7). This cycle takes less than half a second manner to the f-seta, and the g-seta also ex- to complete. As the long rigid setae move uni- hibits slight movement. At rest, the s- and c- formly, their halophores flex readily and sub- setae lie parallel, diverging from the f-seta at tend a much greater arc, consequently pass- an angle of about 30°. Upon contraction of the ing through a larger volume of water. Downloaded from https://academic.oup.com/jcb/article/18/1/95/2418948 by guest on 29 September 2021

Fig. 4. Halophores of adult myodocopines in the families Cypridinidae (A-D) and Cylindroleberididae (E). A, Azy- gocypridina lowryi, broken halophores of the s-seta showing ring construction; B, A. lowrvi, broken halophore of the s-seta showing a broken ring (note ring appears hollow); C, Paradoloria sp., halophore of the s-seta (note rings thinner than in A. lowryi); D, male Skogsbergia sp., halophores arising from the f-seta (note the halophores of the male f- and g-setae are themselves branched in this species; the female f- and g-setae more closely resemble that of A. lowryi); E, Archasterope sp., halophores of the s-seta. Scales: A, B, C, and E = 2 pm; D = 20 Nm.

Fig. 5. Halophore of the s-seta of an adult female Azygocypridina lowryi, showing the width of the rings at various positions along the halophore (of length L). Table 2. Width of three individual adjacent rings (w) of a halophore, at different regions along its length (L), from the s-seta in three cypridinid (Myodocopina) taxa. Downloaded from https://academic.oup.com/jcb/article/18/1/95/2418948 by guest on 29 September 2021 family determination and phylogenetic studies (Kornicker, 1978). However, when com- Fig. 6. Paravargula sp., adult male. Broken base of third paring the s-setae of different myodocopine most proximal halophore of the left s-seta, showing in- families, the possibility of convergence must terior structures (base at the left of the picture). The fact be considered (Kornicker, 1986). For exam- that dendrites presumably innervate the halophore would ple, one form of s-seta occurs only in the fam- seem to indicate that the function of the halophore might ilies Philomedidae and Sarsiellidae, but since be chemoreception. However, unless the interior structures are bundles of axons, their number would indicate these two clades have no immediate common that they are not dendrites. Scale = 5 11m. ancestry, the similarity of their s-setae must result from convergence (Kornicker, 1978, 1986). Variation of the S-Seta (General) The s-seta is confined to the Myodocopina. Halophores occur in all myodocopines. The However, comparable structures exist within s-seta is present on the fifth article of the first the Halocypridina (Myodocopa). Halocypri- antenna of all myodocopine taxa, although its dines frequently possess groups of long thin- form exhibits much variation. The s-seta is ev- walled setae ("filaments") on their first an- ident in the first instar (Fig. 8C, D; Kornicker tennae, which may be finely annulate (Angel, and Iliffe, 1989). Although present in both sexes, 1970; Martens, 1979). For example, species the s-seta is sexually dimorphic to different ex- of Euconchoecia (Halocypridina) have a pad tents and may also show slight variation be- of setae on the fifth article of the first antenna. tween that of the right and left limbs (Fig. 9D). This is probably a parallelism, because halo- The s-seta varies markedly among the phores are setules and therefore cannot be ho- Philomedidae and is a useful character in sub- mologous. In C. belgicae, these setae exhibit

Fig. 7. Diagrammatic terminal part of the first antenna of Azygocypridina lowryi, showing the movement of setae as determined from video recordings (no halophores illustrated). Dotted lines represent positions to which the terminal and subterminal setae move during flexion, solid lines represent resting positions (only setae making significant movements are illustrated). 5 = fifth article; c, f, g, and s represent corresponding setae. Downloaded from https://academic.oup.com/jcb/article/18/1/95/2418948 by guest on 29 September 2021

Fig. 8. Cypridinid first antennae and s-setae. A, Skng.sbergia minuta Poulsen, female left first antenna, lateral view; B, Skngsbergia minuta, male left first antenna, medial view (note many halophores present on f- and g-setae); C, Gi- gantocypris agassizi Miiller, first instar, first antenna (note halophores present only on d- and e-setae (not labeled) terminally); D, G. agassizi, third instar, first antenna (note many halophores present); E, Cypridina acuminata Mfiller, male left first antenna, medial view; F, Pterocypridina .sex Kornicker, male left s-seta, medial view; G, Bathyvargula parvi.spinosa Poulsen, male right s-seta, medial view; H, B. parvispinosa, female right s-seta, medial view. Subfig- ures all from Poulsen (1962), except F from Kornicker (1983). 4 = fourth article; a-g, and s represent corresponding setae. Scales: A, B, and E = 0.1 mm; C = 0.5 mm; D = I mm; G and H = 0.2 mm. Downloaded from https://academic.oup.com/jcb/article/18/1/95/2418948 by guest on 29 September 2021

Fig. 9. Cylindroleberidid (A-D), philomedid (E-I), and rutidermatid (J, K) first antennae and s-setae. A, Dia.sterope grisea Brady, tip of male left first antenna, lateral view (not all setae shown); B, Diasterope grisea, tip of female left first antenna, medial view (not all setae shown); C, Homasterope maccaini Kornicker, male and female s-setae; D, Skogsbergiella spinifera Skogsberg, s-setae of left and right female first antennae; E, Harbansus paucichelatus Kornicker, male left first antenna, medial view; F, Harbansus paucichelatus, female left first antenna, medial view; G, Harbansus dayi Kornicker, male s-seta; H, Paraphilnmede.s unicnrnuta Poulsen, male s-seta; I, Scleroconcha ar- cuata Poulsen, male s-seta; J, Scleraner trifax Kornicker, male s-seta; K, Rutiderma species B Kornicker, male s- seta. Subfigures A-D from Kornicker (1975); E-1 from Kornicker (1978); J, K from Kornicker (1994). 4 = fourth article; a-g and s represent corresponding setae; I = left; r = right. Scales = 100 ltm. a different external morphology to halophores opteroninae), the female s-setal shaft divides when viewed at high magnification. Podoco- into two distal shafts, each bearing two ter- pan ostracods bear no structures comparable minal halophores (Kornicker, 1981 ). In some to the s-seta. species, such as Asteropterygion (Aster- opteroninae), a group of long halophores Detailed Morphological Variation of the arises subterminally, reaching to the tips of S-Seta and the Distribution of Halophores the terminal halophores (Kornicker, 1981). on the First Antenna The male s-setal shaft is often stouter with Cypridinidae.-The shaft of the s-seta is of- numerous halophores arising throughout its ten widened proximally and usually bears two length, giving a plumose appearance. Excep-

groups of halophores: several long halophores tions occur in the genus Homasterope, where Downloaded from https://academic.oup.com/jcb/article/18/1/95/2418948 by guest on 29 September 2021 joined to the setal shaft proximally (some- the male s-seta is similar to that of the fe- times as a double row) and fewer halophores male (Fig. 9C). Furthermore, the female s- (often short) joined more distally and sparsely. seta is bare (no halophores) in the genus Mi- Halophores usually appear slender and not ta- croasteropteron, and the d- or e-seta of the pered. In most cypridinids, the male s-seta is eighth article of the first antenna is also miss- similar to that of the female or comprises a ing in this genus. few more halophores, although the shaft and halophores may be slightly wider in the male. Philomedidae.-The s-seta lacks long halo- A typical s-seta for this family exists in phores in females but has numerous long Cypridina acuminata Muller (Fig. 8E). How- halophores in males. The fifth joint of the first ever, notable exceptions occur in some gen- antenna (from which the s-seta arises) is well era such as Bathyvargula (Fig. 8G, H) and developed, rectangular or trapezoidal in the Metavargula, in which the male s-setal halo- female, small and triangular in the male (Fig. phores differ from those of the females in be- 9E, F). ing broader and flatter, and the shaft of the Kornicker (1978) divided the male philome- male s-seta may have an almost bulbous base. did s-seta into two types: type 1, with an elon- The cypridinids are the myodocopines with gated widened proximal (one-third to two- the oldest fossil record (McKenzie, 1972), thirds of the total length) shaft region bear- and Kornicker (1978) suggested that their s- ing long halophores (Fig. 91); type 2, with a seta is the primitive type. bulbous proximal (about one-fifth of the total In some genera, such as Cypridina (Fig. 8E), length) shaft segment bearing the halophores, the c- and f-setae are unusually long and, there- with the proximal edge of the bulb projecting fore, their halophores are capable of reaching backward (Fig. 9E, G). The male s-seta of the farther out when the antenna is extended genus Paraphilomedes (Fig. 9H) appears tran- [Cypridina is planktonic (e.g., Poulsen, 1962)]. sitional between the two types, but is assigned An exception to the general arrangement of to type 1 since the bulbous shaft section does halophores making up the halothalium within not strictly project backward (Kornicker, the Cypridinidae is shown by the male first 1978). Kornicker (1978) used this character in antennae of Skogsbergia, Pterocypridira, part to define two subfamilies: Philomedinae Heterodesmus, Siphonostra, and Paravargula. (type 1 s-seta) and Pseudophilomedinae (type The males only of these genera typically have 2 s-seta). numerous (long in most genera) halophores Kornicker (1978) considered the type 1 s- arising from the widened bases of the f- and seta to be plesiomorphic within the Philome- g-setae (Fig. 8B). Therefore, in males of these didae, since it appears to be intermediate be- genera the greatest proportion of the halotha- tween the male cypridinid s-seta and the type lium occurs on the f- and g-setae (as opposed 2 s-seta, and considered the type 2 s-seta to to the s-seta in other cypridinids). be a synapomorphy.

Cylindroleberididae.-A high degree of sexual Rutidermatidae.-The female s-seta is either dimorphism of the s-seta is usually exhibited bare (having presumably secondarily lost its (Fig. 9A, B). The female s-setal shaft bears halophores) or possesses two short proximal 4-14 terminal halophores and up to 22 situated and two terminal halophores. The male s-seta more proximally (often shorter) (Kornicker, is stout with numerous long halophores, ap- 1981). In Asteropella and Actinoseta (Aster- pearing plumose. The male s-seta is most similar to the type allel margins which constrict at the shaft tip 1 s-seta of the philomedid male (Kornicker, where the base of the long halophore arises 1978). Males of Rutiderma and Scleraner can (e.g., in A. lowryi the halophore of the d-seta be distinguished by the extent of coverage of is about six times the length of the setal shaft) the broad proximal part of the s-setal shaft by (Fig. 1). Myodocopine first antennal setae halophores. In Scleraner (Fig. 9J), halophores may bear more than one halophore terminally occur throughout the length of this part. In (e.g., Fig. 9C). Nonterminal halophores ex- Rutiderma (Fig. 9K), they are confined to the hibit bases without finely ringed walls (e.g., distal third of this part or less (Kornicker, Figs. 5, 3C). This probably provides increased 1994). structural strength at the area most vulnera-

ble to breakage (not appropriate to terminal Downloaded from https://academic.oup.com/jcb/article/18/1/95/2418948 by guest on 29 September 2021 Sarsiellidae.-The female s-seta is either bare halophores). (having presumably secondarily lost the halo- Halophores occur on most terminal and phores) or with two or three short halophores. subterminal myodocopine first antennal setae. The male s-seta exhibits a proximal bulbous Thus, it appears inappropriate to term only shaft segment bearing numerous halophores. one of these setae as "the sensory seta" (in- The male s-seta resembles the type 2 s-seta ferring a function), as in previous nomencla- of the male Philomedidae, although the two ture. This has also led to confusion in dis- are considered to be convergent (Kornicker, cussion of this seta. Therefore, I term this 1978). seta, arising from the fifth article of the my- DISCUSSION odocopine first antenna, the s-seta (the "s" of the previous terminology "sensory" is re- Myodocopine First Antennal Setae and tained for historical value, while the single Setule Nomenclature letter designation of the other long terminal This study shows that in all myodocopines and subterminal first antennal setae is fol- examined (Table 1) all setules (except those lowed). The s-seta usually consists of a shaft bearing suckers in male cypridinids) of the b- plus halophores, but is so termed even if it to g- and s- first antennal setae exhibit the lacks halophores (probably a secondary loss). same exoskeletal morphology, which includes The s-seta is always and only present in the a layer of very fine rings (Fig. 2). Therefore, Myodocopina. they probably have a similar function. I term I term the collective halophores of one first an individual setule of the myodocopine b- to antenna the halothalium (from the Greek, g- and s- first antennal setae (except those meaning abundance of halophores) (plural bearing suckers in male cypridinids) a halo- halothalia). phore (from the Greek, meaning ring-bearing). The Greek word halo, meaning ring, is Functional Morphology of the Halophores used rather than the more appropriate annu- Long thin structures penetrate each halo- lus (Latin for ring) in this nomenclature to phore. These are possibly dendrites or en- avoid confusion with most terminal and sub- veloping cells. However, transmission elec- terminal first antennal setae, which are de- tron microscopy is required to identify these scribed as annulate. The rings of these setae structures. Halophores exhibit a terminal pore are much wider than those of the halophores through which water-borne molecules can en- and have thicker walls of approximately rec- ter and be potentially detected if the halo- tangular cross section. Halophores may phores possess chemosensory cells. Setae branch off a seta terminally (i.e., at the ter- with a terminal (or subterminal) pore are minal end of many b- to g- and s-setae). Here likely to be chemosensory (Dahl, 1973; Guse, the seta-setule interface occurs at the position 1978; Hindley and Alexander, 1978; Altner where the fine annulations of the halophores et al., 1983), although the presence of a ter- begin. Alternatively, and more commonly minal pore alone does not designate a (e.g., Fig. 9G-K), halophores may branch off chemosensory function. Mechanoreceptive a seta nonterminally. The d- and e-setae, and chemoreceptive sensory cells are often which were each previously believed to be present in the same sensillum (Schmidt and one long bare seta, are actually each com- Gnatzy, 1984). Therefore, halophores may be posed of a short setal shaft with a terminal bimodal chemosensory/mechanosensory sen- halophore. The shaft has approximately par- silla. Table 3. General morphology of s-setae and halothalia in myodocopine families. Downloaded from https://academic.oup.com/jcb/article/18/1/95/2418948 by guest on 29 September 2021

The unusual annulations of the halophores The thin outer layer (sheath) that covers the allows for great flexibility. The tip of a halo- rings of a halophore presumably has some phore passes through a relatively large arc elastic properties and, therefore, allows the during flexion, which may optimize their abil- rings to separate to a relatively high degree, ity to detect chemical and/or physical stim- thus enabling the halophore to maintain in- uli. The halophores are initially moved by tegrity (Fig. l0A). When halophore rings sep- their host when the setae from which they arate during flexion, the outer sheath becomes arise move (due to contraction of the muscles smooth, no longer showing the grooved con- of the sixth, seventh, and eighth articles of touring of the underlying rings (Fig. l0A). the first antenna). The s-seta is probably If each ring is separated from its neighbor by moved by the sixth article pressing against it an angle where the sheath becomes totally (see Fig. 7). Such setal movement may also smooth (but not stretched) along one side of be affected by the surrounding boundary layer a halophore, the halophore could then exhibit of flow. This boundary layer may prematurely great curvature. Therefore, the halophore halt the movement of certain setae as they ap- sheath may not have to exhibit great elastic proach each other. Alternatively, it may cause properties; only enough to resume its relaxed otherwise sedentary setae to be dragged along position (following the contours assigned by with setae moved by article movement. The the rings) after flexion of the halophore. How- first antennal setae move freely at their joints ever, if the separation of adjacent rings shown and are not individually controlled due to an in Fig. 10A is achieved, where x = 12°, the absence of attached muscles (all crustacean sheath must stretch (i.e., move beyond its setae are without muscles). These conclusions smooth state). The sheath must also have sig- are drawn from video recordings of an ostra- nificant tensile strength and be responsible, cod in a very small volume of water, i.e., un- either wholly or in part, for the binding to- natural conditions. However, although be- gether of individual rings. haviors such as swimming would be greatly affected by such conditions, the movement of Analogies with Other Myodocopine individual setae/setules would be less af- Flexible Structures fected. Therefore, the range of first antennal The halophore ultrastructure is most simi- setal/setule movements reported herein is lar to the ultrastructure of the seventh limb probably very similar to that in situ. setae (Fig. 11) within the Myodocopina. Downloaded from https://academic.oup.com/jcb/article/18/1/95/2418948 by guest on 29 September 2021

Fig. 10. Diagrammatic longitudinal sections of two adjacent rings of annulate setae or setules at "rest" or relaxed (linear) and during flexion. A, seta (or setule) containing rings with walls of circular cross section, e.g., a halophore; B, seta (or setule) containing rings with walls of rectangular cross section, e.g., a myodocopine seventh limb seta. f = pivoting point (fulcrum) of adjacent setal or setular rings when "rested" and/or during flexion; f = adjusted fulcrum of adjacent halophore rings during flexion; x = angle subtended by adjacent rings during setal or setular flexion; y = one-half angle subtended by adjacent rings to provide maximum separation of rings without stretching the sheath; z = distance of sheath "extension" required during separation of adjacent rings by x', to be accounted for by stretching. At "rest", x = 0; during flexion, x = 12°. cross section have an adjustable pivoting point or fulcrum, f (Fig. 10A), where the whole setal or setular structure is stable when f is in any of a range of positions (i.e., the walls of the adjacent rings "roll over" each other). For example, in Fig. 10A, the fulcrum moves from point f when the halophore is relaxed to point /'during flexion, but the structure is stable in both positions. However, adjacent rings with walls of rectangular cross

section have a fulcrum, f (Fig. lOB), which Downloaded from https://academic.oup.com/jcb/article/18/1/95/2418948 by guest on 29 September 2021 is fixed. Consequently, these adjacent rings may "slip" from their finely balanced positions when their angle of separation, x (Fig. 10B), is anything but very small. Therefore, this type of structure becomes more unstable as x (i.e., setal flexion) increases. In addition, when the walls of the rings have a rectangular cross section, the surrounding sheath must stretch during ring separation, making setal flexion additionally dependent on the elastic properties of the sheath. The general ring-type setal construction appears successful in permitting both flexibility (to different degrees) and strength. Its analogy may be the result of selection pres- sures, because the same physical properties may be required of different structures, even though their functions vary.

Comparable Crustacean Structures At low magnification halophores appear similar to aesthetascs (Leydig, 1862), com- Fig. 11. Azygocypridina lowryi, adult female seventh mon throughout the Crustacea, resembling limb. A, section near tip showing setae; B, midsection those described by Laverack and Ardill of a seta showing ring-type construction with thin con- (1965) and Laverack (1968) in their filamen- tinuous outer layer. Scales: A = 50 11m; B = 5 11m. tous, thin-walled appearance. However, aesthetascs exhibit a very different ultrastructure; they lack finely annulate walls (Ghiradella et These setae are used for cleaning and stir- al., 1968a, b). The typical marine aesthetasc ring free embryos within the female carapace is a long slender filament (Ghiradella et al., (Vannier and Abe, 1993), functions which re- 1969). The aesthetascs of Cancer productus quire structural flexibility. The seventh limb Randall (Decapoda) show transverse folds in setae tend to be approximately linear when their outer walls which may contribute flex- relaxed (where the straight adjoining edges of ibility to the filaments (Ghiradella et al., adjacent rings lie parallel; Fig. 10B), and 1969). Similarly, aesthetascs of some cirri- spring back to this resting position after flex- pede cyprids exhibit a wrinkled external mor- ion. However, the rings in the walls of these phology (Glenner et al., 1989). seventh limb setae are approximately rectan- The aesthetasc "Y" present in certain gular in cross section and, therefore, proba- podocopan ostracods bears two distinct re- bly have limited flexibility compared to the gions of different external cuticle morphol- thinner rings with walls of circular cross sec- ogy, one smooth, the other with transverse tion of the halophore (Fig. 10). This is be- folds. Perhaps these different morphologies cause adjacent rings with walls of circular provide different selective properties of the cuticle (Andersson, 1975). Therefore, the cu- the many, often difficult to obtain, crustaceans used in this ticle of the aesthetasc "Y" probably has a dif- study. I am also grateful to the Australian Museum Trust ferent function from the more uniform halo- and the Smithsonian Institution for funding this study. phore cuticle, which is reflected in the greater LITERATURE CITED flexibility of the halophore. Altner, I., H. Hatt, and H. Altner. 1983. Structural prop- Antennal setae with fine helical sections of erties of bimodal chemo- and mechanosensitive setae on the pereiopod chelae of the crayfish, Austropota- exoskeleton have been described in Mysi- mobius torrentum.â��Cell and Tissue Research 228: dacea (Crouau, 1981). In addition to their 357-374. number and type of innervating dendrites, the Andersson, A. 1975. The ultrastructure of the presumed different cuticular structures of mysid setae chemoreceptor aesthetasc "Y" of a cypridid ostra- correspond to functional differences. The re- code.-Zoologica Scripta 4: 151-158. Downloaded from https://academic.oup.com/jcb/article/18/1/95/2418948 by guest on 29 September 2021 Angel, M. V. 1970. Bathyconchoecia subrufa n. sp. and gions of first antennal setae with fine helical B. septemspinosa n. sp., two new halocyprids (Ostra- cuticle in Antromy.sis juberthiei Bacescu (My- coda, Myodocopida) from the tropical north Atlantic sidacea) are flexible, and are innervated by che- and the description of the larval development of B. sub- moreceptor and mechanoreceptor cilia (Crouau, rufa.â��Crustaceana 19: 181-199. 1981). Therefore, these setae would detect Bennett, M. B., M. R. Heupel, S. M. Bennett, and A. R. variations in turbulence in the environment. Parker. 1997. Sheina orri (Myodocopa: Cypridinidae), an ostracod parasitic on the gills of the epaulette shark, However, halophores differ from these mysid Hemiscyllium ocellatum (Elasmobranchii: Hemiscylli- setae in that they are: annulate rather than he- dae).-International Journal for Parasitology 27: lical, with annulations continuing to the very 275-281. Cannon, H. G. 1931. On the anatomy of a marine os- tip of the structure; setules not setae; and ca- tracod, Cypridina (Doloria) levis Skogsberg.â��Dis- pable of bending in any direction. covery Reports 16: 435-482. The s-seta is comparable to the callyno- ��â ��â ��â ��â . 1933. On the feeding mechanism of certain ma- phore found among the eucaridan and per- rine ostracods.â��Transactions of the Royal Society of acaridan crustaceans (Lowry, 1986). Thin- Edinburgh 57: 739-764. ��â ��â ��â ��â . 1940. On the anatomy of Gigantocypris mÃ¼l- walled sensory setae ("hairs") (Dahl, 1979) leri.â��Discovery Reports 19: 185-244. constitute this brushlike organ. These setae Cohen, A. C. 1982. Ostracoda.â��In: S. Parker, ed., Syn- are innervated by dense bundles of nerve opsis and classification of living organisms. Pp. 181-202. fibers (Dahl, 1979) and terminate at a pore. McGraw-Hill Book Co., New York, New York. ��â ��â ��â ��â . 1989. Comparison of myodocopid ostracodes in They are also termed aesthetascs (Lowry, two zones of the Belize Barrier Reef near Carrie Bow 1986). However, the s-seta differs quite Cay with changes in distribution 1978â��1981.â��Bulletin markedly from the callynophore in that its of Marine Science 45: 316-337. halophores arise from a setal shaft, whereas ��â ��â ��â ��â , and J. G. Morin. 1990. Patterns of reproduc- tion in ostracodes: a review.â��Journal of Crustacean the aesthetascs of the callynophore arise from Biology 10: 184-211. the articles (possibly fused) themselves of the Crouau, Y. 1981. Cytology of various antennal setae in first antenna. a troglobitic Mysidacea (Crustacea).-Zoomorphology Further comparisons of halophores with 98: 121-134. other crustacean structures may be made fol- Dahl, E. 1973. Presumed chemosensory hairs on talitrid amphipods (Crustacea).â��Entomologica Scandinavica lowing determination of the innervating struc- 4: 171-180. tures. This requires transmission electron mi- ��â ��â ��â ��â . 1979. Deep-sea carrion feeding amphipods: croscopy and represents the next stage in evolutionary patterns in niche adaptation.-Oikos 33: halophore study, where the function(s) of 167-175. Ghiradella, H., J. Case, and J. Cronshaw. 1968a. Fine halophores may be revealed. structure of the aesthetasc hairs of Coenobita com- pressa Edwards.-Journal of Morphology 124: ACKNOWLEDGEMENTS 361-386. I thank Dr. James Lowry (Australian Museum, Syd- ��â ��â ��â ��â , J. Cronshaw, and J. Case. 1968b. Fine struc- ney) and Dr. Noel Tait (Macquarie University, Sydney) ture of the aesthetasc hairs of Pagurus hirsutiusculus for critical reading of the manuscript and advice, Dr. Dana.-Protoplasma 66: 1-20. Louis Kornicker (National Museum of Natural History, â��â��â�� , â��â��â�� , and ��â ��â ��â ��â . 1969. Surface of the cu- Smithsonian Institution, Washington, D.C.) for advice on ticle on the aesthetascs of Cancer.-Protoplasma 69: ostracod morphology and Dr. Duane Hope (NMNH, 145-150. Smithsonian Institution) for advice on sensory structures. Glenner, H., J. T. HÃ¸eg, A. Klysner, and B. Brodin Larsen. Thanks are also due to Mr. Geoffrey Avern (Australian 1989. Cypris ultrastructure, metamorphosis and sex in Museum) for SEM micrograph production and Mr. Ray seven families of parasitic barnacles (Crustacea: Cir- Cameron (Macquarie University) for help with filming. ripedia: Rhizocephala).â��Acta Zoologica 70: 229-242. Cordial thanks go to Dr. James Lowry and Dr. Penny Guse, G.-W. 1978. Antennal sensilla of Neomysis inte- Berents (Australian Museum) for the opportunity to study ger (Leach).â��Protoplasma 95: 145-161. Hindley, J. P. R., and C. G. Alexander. 1978. Structure Lowry, J. K. 1986. The callynophore, a eucaridan/per- and function of the chelate pereiopods of the banana acaridan sensory organ prevalent among the Am- prawn, Penaeus merguiensis.â��Marine Biology 48: phipoda (Crustacea).â��Zoologica Scripta 15: 333-349. 153-160. McKenzie, K. G. 1972. Contribution to the ontogeny and Keable, S. J. 1995. Structure of the marine invertebrate phylogeny of Ostracoda.-Proceedings of the I. P. U., scavenging guild of a tropical reef ecosystem: field XXIII International Geological Congress. Pp. 165-188. studies at Lizard Island, Queensland, Australia.â��Jour- Martens, J. M. 1979. Die pelagischen Ostracoden der nal of Natural History 29: 27-45. Expedition Marchile I (Suedost-Pazifik), 2: Systematik Koehl, M. A. R., and J. R. Strickler. 1981. Copepod und Vorkommen (Crustacea: Ostracoda: Myodocop- feeding currents: food capture at low Reynolds num- ida).-Mitteilungen aus Hamburgischen zoologischen ber.-Limnology and Oceanography 26: 1062-1073. Museum und Institut 76: 303-366. Kornicker, L. S. 1975. Antarctic Ostracoda (Myo- Poulsen, E. M. 1962. Ostracoda-Myodocopa, 1: Cypri- docopina), parts 1 and 2.â��Smithsonian Contributions

diniformes-Cypridinidae.-Dana Reports 57: 1-414. Downloaded from https://academic.oup.com/jcb/article/18/1/95/2418948 by guest on 29 September 2021 to Zoology 163: 1-720. Carlsberg Foundation, Copenhagen, Denmark. ��â ��â ��â â . 1978. Harbansus, a new genus of marine Ostra- Sars, G. O. 1922. Ostracoda.-An account of the Crus- coda, and a revision of the Philomedidae (Myodoco- tacea of Norway with short descriptions and figures of pina).-Smithsonian Contributions to Zoology 260: 1-75. all the species 9: 1-277. Bergen Museum, Bergen, Nor- ��â ��â �� â â . 1981. Revision, distribution, ecology, and on- way. togeny of the ostracode subfamily Cyclasteropinae Schmidt, M., and W. Gnatzy. 1984. Are the funnel-canal (Myodocopina: Cylindroleberididae).â��Smithsonian organs the 'campaniform sensilla' of the shore crab, Contributions to Zoology 319: 1-548. Carcinus maenas (Decapoda, Crustacea)?-Cell and ��â ��â �� â â . 1983. The ostracode family Cypridinidae and Tissue Research 237: 81-93. the genus Pterocypridina.â��Smithsonian Contributions Skogsberg, T. 1920. Studies on marine ostracods, 1: to Zoology 379: 1-29. cypridinids, halocyprids and polycopids.â��Zoologiska ��â ��â �� â â . 1986. Sarsiellidae of the western Atlantic and Bidrag fran Uppsala, supplement 1: 1-784. northern Gulf of Mexico, and revision of the Sarsiel- Vannier, J., and K. Abe. 1993. Functional morphology linae (Ostracoda: Myodocopina).â��Smithsonian Con- and behavior of Vargula hilgendorfii (Ostracoda: My- tributions to Zoology 415: 1-217. odocopida) from Japan, and discussion of its crustacean ��â ��â ��â â . 1994. Ostracoda (Myodocopina) of the SE Aus- ectoparasites: preliminary results from video record- tralian continental slope, part 1.â��Smithsonian Con- ings.â��Journal of Crustacean Biology 13: 51-76. tributions to Zoology 553: 1-200. Walker, K. R. 1972. Trophic analysis: a method for ��â ��â �� â â , and T. M. Iliffe. 1989. Troglobitic Ostracoda studying the function of ancient communities.-Jour- (Myodocopa: Cypridinidae, Thaumatocyprididae) from nal of Palaeontology 46: 82-93. anchialine pools on Santa Cruz Island, Galapagos Is- Watling, L. 1989. A classification system for crustacean lands.-Smithsonian Contributions to Zoology 483: setae based on the homology concept.â��In: B. E. Fel- 1-40. genhauer, L. Watling, and A. B. Thistle, eds., Func- Laverack, M. S. 1968. On the receptors of marine in- tional morphology of feeding and grooming in Crus- vertebrates.-Oceanography and Marine Biology 18: tacea. Pp. 15-26. A. A. Balkema, Rotterdam, The 11-56. Netherlands. , and D. J. Ardill. 1965. The innervation of the RECEIVED: 29 January 1997. aesthetasc hairs of Panulirus argus.â��Quarterly Jour- ACCEPTED: 22 April 1997. nal of Microscopical Science 106: 45-60. Leydig, F. 1862. Naturgeschichte der Daphniden (Crus- Address: Division of Invertebrate Zoology, Australian tacea, Cladocera).â��Laupp � Siebeck, TÃ¼bingen, Ger- Museum, 6 College Street, Sydney, New South Wales, many. Australia 2000.