Redescription and life cycle of Gangesia parasiluri (Cestoda: Proteocephalidae), a parasite of the Far Eastern catfish Silurus asotus

Transcription

1 FOLIA PARASITOLOGICA 46: 37-45, 1999 Redescription and life cycle of Gangesia parasiluri (Cestoda: Proteocephalidae), a parasite of the Far Eastern catfish Silurus asotus Takeshi Shimazu Nagano Prefectural College, Miwa, Nagano , Japan Key words: Gangesia parasiluri, Cestoda, morphology, life cycle, Japan Abstract. Gangesia parasiluri Yamaguti, 1934 (Cestoda: Proteocephalidae) is redescribed on the basis of adults obtained from the intestine of Silurus asotus Linnaeus (Teleostei: Siluridae) from Lake Suwa, Nagano Prefecture, central Japan. Its life cycle was studied in the field and laboratory. Rostellar hooks of the adults showed a wide variation in number, ranging from 35 to up to 57. Plerocercoids were found in the rectum of Chaenogobius urotaenia (Hilgendorf) and Rhinogobius brunneus (Temminck et Schlegel) (Teleostei: Gobiidae) from the same lake. Procercoids were formed in the haemocoel of Mesocyclops leuckarti (Claus) (Copepoda: Cyclopidae) 7 days post infection at C. They developed into plerocercoids in the intestine of Pseudorasbora pumila pumila Miyadi (Teleostei: Cyprinidae), R. brunneus and S. asotus. Plerocercoids from naturally and experimentally infected fishes were fed to S. asotus, from which immature worms were recovered. It is considered that the life cycle involves three hosts: a copepod as the intermediate host in which procercoids are formed, small fish as paratenic hosts which retain plerocercoids and transport them into S. asotus, and S. asotus as the definitive host in which adults develop. Rostellar hooks of the adults were much fewer, much larger and arranged in fewer circles than those of the plerocercoids. It is suggested that the former are newly formed and replace the latter in an early stage of development of plerocercoids into adults in S. asotus. Yamaguti (1934) described a proteocephalid cestode, Gangesia parasiluri Yamaguti, 1934, as a new species from the intestine of the Far Eastern catfish Parasilurus asotus (Linnaeus) (= Silurus asotus Linnaeus) in Japan. Since then, this tapeworm has been recorded from northeastern parts of Asia (Dubinina 1971, 1987, Chen 1984) but not from Japan. Demshin (1985) experimentally studied the development of its procercoids in a copepod intermediate host, Mesocyclops leuckarti (Claus). In a lake in Japan, adults of the tapeworm were found in S. asotus, and plerocercoids of a Gangesia species were obtained from gobiid fishes. The rostellar hooks of the plerocercoids were much smaller and much more numerous than those of the adults. In order to confirm whether the plerocercoids were conspecific with the adults of G. parasiluri, the life cycle of the tapeworm was studied in the field and laboratory. The purposes of this paper are to describe adults, plerocercoids and procercoids obtained; to show that the plerocercoids belong to G. parasiluri; and to discuss the life cycle of the tapeworm. MATERIALS AND METHODS Adults from naturally infected Silurus asotus. Worms were obtained from the intestine of Silurus asotus (Teleostei: Siluridae) from Lake Suwa, Nagano Prefecture, central Japan, at irregular intervals from May 1975 to July Plerocercoids from naturally infected fishes. Plerocercoids were obtained from the rectum of gobiids, Chaenogobius urotaenia (Hilgendorf) and Rhinogobius brunneus (Temminck et Schlegel) (Teleostei: Gobiidae), from the same lake at irregular intervals from May 1989 to October Experimental hosts used. Male and female adults of Mesocyclops leuckarti (Copepoda: Cyclopidae) were collected in small ponds in Nagano, Nagano Prefecture, where no S. asotus lived. Many R. brunneus (27-31 mm in standard body length) and Pseudorasbora pumila pumila Miyadi (Teleostei: Cyprinidae) (26-32 mm) were collected in a pond in Nakano, Nagano Prefecture, where no catfish lived, either. Before use, five S. asotus ( mm) had been reared in an aquarium for about 2-4 months after capture in Lake Suwa, feeding on live R. brunneus and P. pumila pumila mentioned above. Experimental infection of oncospheres to copepods. Live gravid worms spontaneously laid eggs, each of which contained a fully-formed oncosphere, when they were immersed in distilled water just after having been dissected out of S. asotus. Newly laid eggs and copepods were kept together in distilled water in a glass container for 1 day. The copepods alone were then reared in pond water from the same ponds in Nagano in another glass container in the dark at C to supply food to the copepods and to reduce activity of them. They were examined for procercoids at irregular intervals. Similar infection experiments were made four times from late June to mid-july in 1994 to Experimental infection of procercoids to fishes. Five R. brunneus and five P. pumila pumila were put into two separate glass containers including many copepods, all of which proved to harbour fully-formed procercoids days post infection at C. Similarly infected copepods were fed by stomach tube to one S. asotus. All these fish were Address for correspondence: T. Shimazu, Nagano Prefectural College, Miwa, Nagano , Japan. Phone: ; Fax: ; tshimazu@cb3.so-net.ne.jp 37

2 examined the next day to see whether procercoids were infective to them and to obtain plerocercoids with the scolex almost the same as that of procercoids. Experimental infection of plerocercoids to S. asotus. Many 1-day-old plerocercoids from the rectum of one of five experimental R. brunneus were given by stomach tube to one S. asotus, which was examined for worms 5 days later. Each of three S. asotus was given by stomach tube 15 plerocercoids from the rectums of C. urotaenia and R. brunneus from Lake Suwa. One and two of them were examined for worms 5 and 7 days later, respectively. Specimens. Most worms were fixed in hot 5% formalin, stained with Delafield s haematoxylin and mounted in Canada balsam. Some were flattened, fixed in either AFA or Schaudinn s fluid, stained with Heidenhain s iron haematoxylin, and mounted in Canada balsam. Some others were fixed in hot 5% formalin and mounted unstained in Canada balsam. Some scoleces and proglottides were serially sectioned (7 or 10 µm thick) and stained with haematoxylin and eosin. The type specimens of G. parasiluri were borrowed from the Meguro Parasitological Museum (MPM), Tokyo (holotype, MPM Coll. No ; paratype, 22794). Specimens were observed using light microscopy often with a differential interference contrast. Representatives of the specimens studied have been deposited in the National Science Museum, Tokyo (NSMT-Pl 4060, 4061, 4065, , ). RESULTS Adults from naturally infected Silurus asotus Though the prevalence and intensity of infection of worms in S. asotus were not accurately recorded at each time of examination, 15 of 56 S. asotus (about mm in standard body length) examined from 27 May to 8 July in 1995 were infected with 1-4 worms visible to the naked eye. Worms found were in various degrees of development. Most worms found in April (mid-spring) and May were immature, and a few were mature. Many worms obtained in June (early summer), July and October (mid-autumn) were gravid; and some were immature or mature. Redescription of Gangesia parasiluri Figs. 1-8 Based on gravid specimens, measurements of ten in millimetres unless otherwise stated (Figs. 1-6). Length of body up to 87, consisting of up to 190 proglottides. Scolex long by wide. Rostellum like a biconvex lens from side view (Figs. 4, 5), rarely anteriorly truncated, long by wide. Rostellar hooks rose-thorn-shaped, arranged in concentric circles at about mid-level of margin of rostellum; hooks of almost same size arranged in 1 complete or 2 irregular circles in apparently normal state (Fig. 4); small and large hooks numbering up to 57 in total somewhat irregularly alternating in some specimens (Fig. 5); very large hooks (Fig. 6, arrowheads) with broad bases and small blades mixed with very small hooks (Fig. 6, arrows) like those of procercoids and plerocercoids as described later; gaps of various sizes without any hooks often seen in circles; no hooks present in extreme instances. Base of hook broad, slightly concave, square or triangular, 7-16 µm long by µm wide, possibly lying under tegument; blade of hook curved, µm long, possibly protruding from tegument (Fig. 1). Suckers in diameter; hooklets numerous, rose-thorn-shaped, with blade about 2 µm long and base about 1 µm long, arranged in about 7 regular rows on anterior two-thirds of rim of each sucker (Figs. 4, 5). Neck slender, sometimes widens posteriorly to scolex, long by wide. Strobila acraspedote. Immature proglottides wider than long; mature ones (Fig. 2) usually longer than wide, long by wide, with ratio of length to width being 1: ; gravid ones (Fig. 3) longer than wide, long by wide, with ratio being 1: Lateral nerve cord located between vitelline follicles and longitudinal muscle bundles on each side of medulla. Longitudinal muscle bundles more sparsely arranged in both lateral fields than in other fields. Main longitudinal excretory canals internal to vitelline follicles. Testes per proglottis, varying in number from proglottis to proglottis even in a single strobila, in diameter, preovarian, distributed continuously in dorsal plane of medulla. External seminal vesicle ventral to testes, much convoluted. Cirrus pouch long by wide, occupying 29-38% of proglottis width. Internal seminal vesicle thin-walled, folded. Ejaculatory duct thick-walled, folded, lined with gland cells. Cirrus unarmed, lined with gland cells, long when everted. Genital atrium small. Genital pore irregularly alternating, located at 36-47% of lateral margin of proglottis from its anterior end. Ovary dumbbellshaped, each lobe long by wide, further multilobulated. Oocapt and ootype-complex posterior to ovarian isthmus. Uterine duct undulating in dorsal plane of medulla, emptying dorsally into median stem of uterus midway between cirrus pouch and ovary. Median stem of uterus I-shaped, formed in mature proglottides, extending in ventral plane of medulla from ventral side of ovary to anterior border of proglottis, giving off lateral branches on each side in gravid proglottides; uterine pores 5-8 per proglottis. Vagina winding anteriorly along median line from ootype to level of cirrus pouch between uterus and testes, slightly dilated, sometimes including sperm, running ventral to cirrus pouch, opening into genital atrium anteriorly or posteriorly to cirrus pore, with well-developed sphincter. Vitelline follicles small, in a dense band extending entire length of proglottis on each lateral side, mostly in medulla but some extending into cortex in part; vitelline ducts transverse, ventral to ovary; common vitelline duct ventral to ovarian isthmus and ootype. 38

3 Shimazu: Life cycle of Gangesia parasiluri Figs Gangesia parasiluri, gravid adults from intestine of naturally infected Silurus asotus from Lake Suwa, Nagano, central Japan. Fig. 1. Rostellar hooks, lateral view (A), en face view (B), length of blade (a), length and width of base (b and c). Scale bar = 0.01 mm. Fig. 2. Mature proglottis, ventral view. Scale bar = 0.5 mm. Fig. 3. Gravid proglottis, ventral view. Scale bar = 0.5 mm. Eggs (Figs. 7, 8). Oncosphere almost globular,18-28 µm in diameter; embryonic hooks consisting of slender handle 6-8 µm long, knob-like guard about 1 µm long and curved blade about 3 µm long, outer two pairs slightly smaller than inner one. First or outermost egg envelope (Figs. 7, 8a) thin, hyaline, µm in diameter, sometimes incompletely swollen with funnellike depressions. Second (Fig. 8b) thick, µm in diameter, with granules of various sizes between it and third. Third (Fig. 8c) thin, about 30 µm in diameter. Fourth or innermost (Fig. 8d) very thin, tightly enclosing oncosphere. Type specimens The type specimens borrowed were the holotype, a whole-mounted gravid worm from the small [sic] intestine of Parasilurus asotus (= Silurus asotus) from Ainoki-mura, Nakaniikawa-gun, Toyama Prefecture, on 29 October 1929; and the paratype, a whole-mounted immature worm from the small [sic] intestine of P. asotus from Lake Ogura, Kyoto Prefecture, on 14 November 1931 (Yamaguti 1934, Shimazu 1990, 1993). It seems from the plural paratypes (p. 48) that Yamaguti (1934) designated one or more other paratypes; but they have not been located. My reexamination of the specimens revealed that (1) Yamaguti s original description and figures are almost correct; (2) mature and gravid proglottides of the holotype were usually longer than wide, mm long by mm wide, with the ratio being 1: ; (3) the rostellar hooks appeared to be 35 in both specimens (instead of 37 in Yamaguti s description for the holotype), with base µm long by µm wide and blade µm long in the holotype; (4) the suckers bore numerous hooklets with blade about 2 µm long on the anterior two-thirds (instead of the anterior part in Yamaguti s description) of their margins; (5) the testes numbered (instead of in Yamaguti s description) per proglottis in the holotype; (6) the cirrus pouch was mm long by mm wide, occupying 30% of proglottis width, in the holotype; (7) the lateral uterine branches numbered at least 14 (instead of not more than 12 in Yamaguti s description) on each side; and (8) some of the vitelline glands extended into the cortex. 39

4 Figs Gangesia parasiluri. Figs Gravid adults (continued). Fig. 4. Scolex, rostellar hooks arranged in a single circle. Scale bar = 0.1 mm. Fig. 5. Scolex, small and large rostellar hooks irregularly alternating in 2 irregular circles. Scale bar = 0.1 mm. Fig. 6. Rostellum, very small (arrows) and very large (arrowheads) rostellar hooks mixed. Scale bar = 0.05 mm. Fig. 7. Mature eggs, living. Scale bar = 0.05 mm. Fig. 8. Mature egg, highly pressed, showing first or outermost (a), second (b), third (c) and fourth or innermost (d) egg envelopes and oncosphere, living. Scale bar = 0.03 mm. Fig. 9. Five procercoids (arrows) in haemocoel of experimentally infected Mesocyclops leuckarti, living, 7 days post infection at C. Scale bar = 0.03 mm. Procercoids from experimentally infected copepods Figs. 9, 10 Eggs were readily ingested by copepods, in the haemocoel of which oncospheres developed into procercoids (Fig. 9). Fully-formed procercoids were obtained 7 days post infection at C and 9 days at 20 C. In one experiment made in July 1994, 26 of 40 copepods were infected with 1-17 procercoids per host 7 days post infection at C. The morphology and 40 measurements (in micrometres) of ten whole-mounted procercoids were as follows (Fig. 10). A thin hyaline membrane surrounding body proper and cercomer. Body proper obovoid, long by wide. Rostellum long by wide; rostellar hooks numerous, numbering about , minute, rose-thorn-shaped, with base 2-3 long and blade about 4 long, arranged in about 5 irregular circles. Suckers in diameter, at about mid-level of body

5 Shimazu: Life cycle of Gangesia parasiluri proper, each bearing numerous minute hooks arranged in about 7 regular rows on its rim except in posterior seventh part. Calcareous corpuscles numerous. Embryonic hooks scattered in body proper. Cercomer globose, hollow in center, in diameter; large cells arranged in a single layer in cercomer wall. Plerocercoids from experimentally and naturally infected fishes Figs Many plerocercoids were recovered from the intestine of three of five Pseudorasbora pumila pumila and one S. asotus and from the rectum of one of five Rhinogobius brunneus 1 day post infection. The plerocercoids from P. pumila pumila and S. asotus were almost the same in morphology and measurements as, though slightly larger than, those of the body proper of the above-described procercoids from experimental infections (Figs. 11, 12). Plerocercoids were found in the rectum of Chaenogobius urotaenia and R. brunneus from Lake Suwa in all seasons. Though the prevalence and intensity of infection of worms in these gobiids were not accurately recorded at each time of examination, they were much higher in C. urotaenia than in R. brunneus. The morphology and measurements (in micrometres) of ten whole-mounted plerocercoids were as follows (Figs. 13, 14): body long by wide; rostellum long by wide; rostellar hooks minute, with base 2-4 long and blade 3-4 long, numerous, numbering , arranged in 3-4 irregular circles; suckers in diameter, each bearing numerous minute hooks arranged in 6-7 regular rows on anterior two-thirds of its rim; calcareous corpuscles numerous; some embryonic hooks remaining in body. Immature worms from experimentally infected Silurus asotus Figs. 15, 16 At 5 days post infection, six and seven worms, respectively, were found in the intestine of two S. asotus fed plerocercoids from experimentally infected R. brunneus and from naturally infected C. urotaenia and R. brunneus. At 7 days post infection, one, one and five worms, respectively, were obtained from the intestine of three S. asotus fed plerocercoids from naturally infected C. urotaenia and R. brunneus. The thirteen 5-day-old worms did not show strobilization and measured (in micrometres): body long; rostellum long by wide; suckers in diameter. Of them, one (shrunk, 253 long) had rostellar hooks of two types: the hooks of type 1 (Fig. 15, arrow) were similar to those of the above-described plerocercoids in shape and size, though a little fewer, and arranged in 2-3 irregular circles; and the hooks of type 2 (Fig. 15, arrowhead) were very small, sparse, with only blades, arranged in 2 irregular circles posteriorly to the hooks of type 1. In other two ( long), the hooks of type 1 (Fig. 16, arrow) were much fewer than those of the preceding specimen. However, the hooks of type 2 (Fig. 16, arrowhead) were much larger than those of the preceding specimen, but still with only blades: those of the anterior circle numbered (missing in part) with blades 4-9 long; and those of the posterior circle numbered with blades about 2 long. The remaining ten ( long) had only the hooks of type 1. The seven 7-day-old worms were also still in a state before the onset of strobilization and measured: body long (some being shrunk); rostellum long by wide; suckers in diameter. Their rostellar hooks were similar to those of the abovedescribed gravid specimens in shape, size, number and arrangement. DISCUSSION Identification The present adult specimens are similar in morphology and host species to Yamaguti s (1934) original description and the two type specimens reexamined of Gangesia parasiluri, though they are different from the latter in numbers of the rostellar hooks (35-57 instead of 32-37), testes ( instead of per proglottis), and lateral uterine branches (14-22 instead of at least 14). These differences are considered to be within the limits of intraspecific variability in a single species. Consequently, the present adult specimens are identified as G. parasiluri. The plerocercoids from natural infections also belong to this species because they were almost identical with those from experimental infections in morphology and measurements, especially in the shape, size, number and arrangement of the rostellar hooks. Yamaguti (1934) obtained a gravid strobila without the scolex from the small [sic] intestine of Parasilurus asotus (= Silusus asotus) from Lake Ogura. He described and illustrated a whole-mounted and seriallysectioned fragments of various lengths of the strobila as a paratype of his new species, Proteocephalus parasiluri, or now Paraproteocephalus parasiluri (Yamaguti, 1934) Shimazu, 1993 (see Shimazu 1993); but Shimazu (1993) regarded the strobila as G. parasiluri. In addition, unpublished specimens of this tapeworm have been known from S. asotus at the following localities in Japan: (1) a river in Kujukurihama, Chiba Prefecture (MPM Coll. No. 9848); (2) the Tama River (locality not specified) (MPM Coll. No ); (3) Lake Kizaki, Nagano Prefecture (NSMT-Pl ); (4) a stream in Koshoku, Nagano Prefecture (NSMT-Pl 4059); (5) Lake Biwa at Onoe, Shiga Prefecture (NSMT-Pl 4062); and (6) the Nukui River at Babadai, Higashihiroshima, Hiroshima Prefecture (NSMT-Pl 4063) (all my identification, unpubl. data). Gangesia parasiluri has been recorded from northeastern parts of Asia. The material that was 41

6 Figs Gangesia parasiluri. Fig. 10. Procercoid from experimentally infected Mesocyclops leuckarti, living, 9 days post infection at 20 C. Scale bar = 0.1 mm. Figs. 11, 12. Plerocercoid from intestine of experimentally infected Pseudorasbora pumila pumila, entire body (Fig. 11), scolex (Fig. 12), 1 day post infection, fixed in hot formalin. Scale bars = 0.1 mm (Fig. 11) and 0.05 mm (Fig. 12). Figs. 13, 14. Plerocercoids from naturally infected Chaenogobius urotaenia from Lake Suwa, entire body (Fig. 13), scolex (Fig. 14), fixed in hot formalin. Scale bars = 0.1 mm (Fig. 13) and 0.05 mm (Fig. 14). Figs. 15, 16. Rostella, immature worms from experimentally infected S. asotus, 5 days post infection, showing rostellar hooks of type 1 (arrow) and type 2 (arrowhead), fixed in hot formalin. Scale bars = 0.05 mm. 42

7 Shimazu: Life cycle of Gangesia parasiluri reported by Dubinina (1971, 1987) from P. asotus from the Amur River basin, Russia, is morphologically similar to the type specimens re-examined and present material. There were rostellar hooks; and the smaller their number, the larger their size. This tendency was not clear in the present material. The testes numbered and more per proglottis. I agree with Dubinina (1971) that the number of testes should not be considered an important feature because the testes are difficult to count due to their large number and because they vary in number in different proglottides of the same strobila. In Dubinina s material, eggs were µm in diameter (of the second envelope counting from the outermost one?), and oncospheres were µm in diameter. On the other hand, mature eggs in the material that Demshin (1985) obtained from P. asotus from water bodies of Prikhankai plain, Russia, were µm in the outermost envelope, 36 µm in the second, and µm in the innermost, though nothing was mentioned about the third (counting from the outermost). Chen (1984) described G. parasiluri from P. asotus, S. soldatovi and P. asotus S. soldatovi from the Liao River in Liaoning Province, northeastern China. In Chen s material, the rostellar hooks numbered and measured µm long; each sucker had marginal hooklets on its anterior half; the testes were per proglottis; the lateral uterine branches were on each side; and eggs measured 26.6 µm in outer diameter (in outermost envelope?) and 13 µm in inner diameter (in innermost envelope or oncosphere?). Chen s material does not seem to differ significantly from the type specimens re-examined and present material in spite of its smaller eggs. The larva figured (fig. 1A), about 14 mm long (calculated from the figure), possesses the rostellar hooks arranged in a single circle. Gupta and Arora (1979) obtained a material from Callichrous bimaculatus and Wallago attu from Ambala (Haryana), India, and described it as G. parasiluri. Compared with the type and present specimens, it has a large scolex; large suckers without any marginal hooklets; a large rostellum; few (16-32), large (11-32 µm long) rostellar hooks; and large eggs (93 µm in diameter). Nothing is said about the lateral uterine branches. Their material seems to need further morphological study before its species is definitely determined. Life cycle Demshin (1985) was successful in tracing the development from oncosphere to procercoid of G. parasiluri from Prikhankai plain in Mesocyclops leuckarti in the laboratory. Demshin s procercoids that were recovered days post infection at C are similar in general morphology and measurements to the present procercoids from experimentally infected M. leuckarti. However, there is a distinct difference in the armature of the rostellum between the two: arrow-shaped processes (apparently primordia of the rostellar hooks) 14 µm long in a single circle on the rostellum with bristles (Demshin 1985) instead of about minute hooks with bases and blades (about 4 µm long) in about 5 irregular circles on the rostellum without bristles (this paper). Demshin s description is questionable. Evidently, the copepod M. leuckarti serves as a good experimental intermediate host for G. parasiluri (Demshin 1985, this paper). A natural copepod host is still unknown. The present study shows that procercoids can infect not only Rhinogobius brunneus and Pseudorasbora pumila pumila but also Silurus asotus to develop into plerocercoids in these fishes. In Lake Suwa, plerocercoids were really found in the rectum of Chaenogobius urotaenia and R. brunneus in all seasons (this paper) and once in the intestine of Pseudorasbora parva (Temminck et Schlegel) (NSMT-Pl 4064, unpubl. data). Moreover, plerocercoids survived in the rectum of R. brunneus for six months without any further development in the laboratory (NSMT-Pl 4546, unpubl. data). Probably, these gobiids and cyprinids become infected with plerocercoids by ingesting copepods harbouring procercoids. The gobiids, particularly C. urotaenia, might become infected with plerocercoids by eating smaller fish harbouring plerocercoids, though this has not yet been confirmed. Plerocercoids are retained for a long period without any further development at least in the gobiids. S. asotus can become infected with the tapeworm by swallowing copepods harbouring procercoids, suggesting a two-host life cycle in G. parasiluri (this paper). However, this may be unlikely to take place because S. asotus, except larvae and very small juveniles, feeds mainly on small fish, shrimp, crayfish, frogs, etc. All the 20 immature worms recovered from four S. asotus 5 and 7 days post infection are considered those that developed from plerocercoids experimentally given to S. asotus because they were too small to start strobilization and because S. asotus had been reared in an aquarium for about 2-4 months before use. It seems from an ecological point of view that small fish such as the gobiids and cyprinids play an essential role to transport plerocercoids to S. asotus. Consequently, three hosts might be ecologically required for the tapeworm to complete its life cycle: a copepod as the intermediate host in which procercoids are formed; small fish as paratenic hosts which retain plerocercoids and transport them into S. asotus; and S. asotus as the definitive host in which adults develop. It appears from the present data that the tapeworm mature in the definitive host at least from early June (early summer) to late October (mid-autumn) in Lake Suwa. This might be implicated in the presence of a seasonal development of the tapeworm. 43

8 Rostellar hooks In G. parasiluri, the processes like hooks with bases and blades were arranged in a band on the rostellum of the procercoid, plerocercoid and adult; and similar processes were also present on the rim of each sucker (this paper). They are called rostellar hooks and hooklets, respectively, consistently in this paper. In some other proteocephalideans, the spines and hooklets are in fact giant, hook-like microtriches of the tegument (de Chambrier and Vaucher 1997, T. Scholz pers. comm.). It is unknown whether the rostellar hooks and hooklets of G. parasiluri are specialized microtriches or true hooks accompanied with muscles because light microscopy was used in the present study. The rostellar hooks in the present procercoids and plerocercoids were almost the same in shape and size, though they were slightly fewer in total number and in number of circles in the plerocercoids than in the procercoids. However, the rostellar hooks of the plerocercoids were distinctly different from those of the adults in size (3-4 versus µm long in blade), total number ( versus 35-57) and number of circles (3-4 irregular versus 1 complete or 2 irregular). Nothing is known about the formation of rostellar hooks in gangesiids. In Taenia crassiceps (Zeder, 1800) Rudolphi, 1810, each rostellar hook originates from a single tegumental microthrix; the blade is first formed; and then the guard and handle are added (Mount 1970). If this is the case also in G. parasiluri, this and the present observations on the immature specimens from experimental S. asotus suggest that (1) the minute hooks of type 1 with both blades and bases in 3-4 irregular circles were the fully-formed procercoid-plerocercoid hooks; and (2) the large hooks of type 2 with only blades in 2 circles posterior to the hooks of type 1 were in an early stage in the formation of the adult hooks; (3) the procercoid-plerocercoid hooks no longer develop further and subsequently fall off from the rostellum or degenerate; and (4) the adult hooks are newly formed, grow rapidly to full size, and replace the procercoidplerocercoid hooks by the time worms attain a body length of about 2-3 mm before starting strobilization about 7 days post infection in the definitive host. Of immature worms from naturally infected S. asotus, one (NSMT-Pl 4537), 624 µm in body length, before the onset of strobilization, possessed the rostellar hooks of two types as described above (unpubl. data). In some gravid worms, very small hooks and very large hooks were mixed (Fig. 6). Most presumably, the former were procercoid-plerocercoid hooks which had remained without falling off or degenerating, and the latter were adult hooks. The procercoid and plerocercoid of the tapeworm had several alternating circles of numerous rostellar hooks. In this armature of the rostellum, they resemble the adults of two other gangesiine genera, Silurotaenia Nybelin, 1942 and Electrotaenia Nybelin, 1942, from siluroid fishes in Europe and Africa, respectively (Nybelin 1942). This resemblance might be of phylogenetic importance. Further studies are needed of the nature and formation of the rostellar hooks of the procercoid, plerocercoid and adult of the tapeworm using transmission and scanning electron microscopy. Acknowledgements. I am grateful to Dr. Toshihiko Mizuno, Toyonaka, Osaka, for identifying the copepod; Dr. Tomáš Scholz, Institute of Parasitology, České Budějovice, Czech Republic, for translating the Russian texts into English; Prof. J. C. Pearson, Brisbane, Australia, for valuable comments on the manuscript; Mr. Jun Araki, MPM, for the loans of the specimens; and Dr. Marc Zehnder, University of Neuchâtel, Neuchâtel, Switzerland, for a copy of Nybelin s paper. The abstract of this paper was read at the Annual Meeting of the Japanese Society of Parasitology in Chiba in April 1997 (Shimazu 1997). REFERENCES CHEN Y.S. 1984: The parasitic cestodes of fishes from Liao He in China. In: Institute of Hydrobiology, Academica Sinica (Ed.), Parasitic Organisms of Freshwater Fish of China. Agr. Publ. House, Beijing, pp (In Chinese, with English title.) DE CHAMBRIER A., VAUCHER C. 1997: Révision des cestodes (Monticellidae) décrits par Woodland (1934) chez Brachyplatystoma filamentosum avec redéfinition des genres Endorchis Woodland, 1934 et Nomimoscolex Woodland, Syst. Parasitol. 37: DEMSHIN N.I. 1985: Postembryonic development of the cestode Gangesia parasiluri (Proteocephalata, Proteocephalidae), a parasite of the Amur catfish (Parasilurus asotus). Zool. Zh. 64: (In Russian, with English summary.) DUBININA M.N. 1971: Cestodes from fishes of the Amur basin. Parazitol. Sb. 25: (In Russian.) DUBININA M.N. 1987: Phylum Plathelminthes. Class Cestoda Rudolphi, In: O.N. Bauer (Ed.), Key to the Parasites of Freshwater Fishes of the USSR. Vol. 3. Parasitic Metazoa, Part 2. Nauka, Leningrad, pp (In Russian.) GUPTA N.K., ARORA S. 1979: On a new species of the genus Gangesia Woodland, 1924 (Cestoda: Proteocephaloidea) from a freshwater fish Wallago attu, along with a survey of its related forms. Res. Bull. (Sci.) Panjab Univ. 30: MOUNT P.M. 1970: Histogenesis of the rostellar hooks of Taenia crassiceps (Zeder, 1800) (Cestoda). J. Parasitol. 56: NYBELIN O. 1942: Zur Helminthenfauna der Süsswasserfische Schwedens. II. Die Cestoden des Welses Göteborgs Kungl. Vet. O. Vitterh. Samh. Handl., 6. f., s. B, 1:

9 Shimazu: Life cycle of Gangesia parasiluri SHIMAZU T. 1990: Some species of the genus Proteocephalus (Cestoidea: Proteocephalidae) from Japanese freshwater fishes, including a description of a new species. Jpn. J. Parasitol. 39: SHIMAZU T. 1993: Redescription of Paraproteocephalus parasiluri (Yamaguti, 1934) n. comb. (Cestoidea: Proteocephalidae), with notes on four species of the genus Proteocephalus, from Japanese freshwater fishes. J. Nagano Pref. Coll. No. 48: 1-9. SHIMAZU T. 1997: Life cycle of Gangesia parasiluri (Cestoidea: Proteocephalidae). Parasitol. Int. 46 (Suppl.): 126. YAMAGUTI S. 1934: Studies on the helminth fauna of Japan. Part 4. Cestodes of fishes. Jpn. J. Zool. 6: Received 11 March 1998 Accepted 16 April