AN ATTEMPT TO DETERMINE THE INTERMEDIATE HOST FOR POMPHORHYNCHUS LAEVIS (ACANTHOCEPHALA) IN THE BALTIC SEA

Background. A number of fish species in the Baltic Sea are known as definitive hosts for Pomphorhynchus laevis but it is unclear which of the Gammarus species is the intermediate host of this parasite. The aim of the present paper was to identify this host in brackish waters of the Baltic Sea. Material and methods. A total of 531 scuds (G. salinus, G. zaddachi, and G. duebeni) were collected from the Gulf of Gdaƒsk and the Pomeranian Bay, to determine the infection rate of those amphipods in the natural environment. Under experimental condition the scuds were exposed to infection with P. laevis in two different arrangements. In treatment one, the amphipods were kept, from May to July, in the same tank with infected flounder. In treatment two, 197 scuds were exposed to eggs of P. laevis, taken from dissected female acanthocephalans. Results. Scuds sampled from two areas of the Baltic Sea were not infected with P. laevis. Out of three Gammarus species cohabiting with infected flounder only G. zaddachi became infected. None of the scuds exposed directly to the eggs of the parasites became intermediate host of the acanthocephalan studied. Conclusion. Gammarus zaddachi is probably an intermediate host for Pomphorhynchus laevis in the Baltic Sea.


INTRODUCTION
A number of marine and freshwater fishes have been reported as definitive hosts of acanthocephalan Pomphorhynchus laevis (Zoega in Müller, 1776). The occurrence of this parasite in the waters of the Polish Exclusive Economical Zone of the Baltic Sea has been associated mainly with the flounder, Platichthys flesus (Linnaeus, 1758) (cf. Mulicki 1947, Sulgostowska et al. 1987, Sulgostowska and Styczyƒska-Jurewicz 1996, Zió∏kowska et al. 2000, Chibani et al. 2001.
In flounder, this parasite attaches in the terminal portion of the intestine, perforating its wall, with the proboscis. As a consequence of the perforation, the proboscis protrudes into the peritoneal cavity of the fish, while the rest of the body remains in the lumen of the intestine. Under conditions of very intensive infections, the worms may completely obstruct the patency of the intestine. In rare instances, single P. laevis can be found on the surface of the liver.
In freshwater bodies of the Great Britain the intermediate host of this acanthocephalan was identified as Gammarus pulex (cf. Hine and Kennedy 1974) and G. duebeni (cf. Kennedy et al. 1978). In France the freshwater intermediate hosts were G. pulex and G. roeseli (cf. Bauer et. al. 2000), whereas in Italy-Echinogammarus stammeri (cf. Dezfuli et. al. 2000). According to Ja˝d˝ewski (1975) G. pulex has been considered a collective species.

MATERIAL AND METHODS
The present paper was divided into three parts. Part one: Natural infection of Gammarus spp. with larvae of P. laevis in the southern Baltic Sea.
The scuds were sampled from a site in the Gulf of Gdaƒsk, within the period of September 2001-August 2002 with the aid of a dredge towed by a fishing cutter. A total of 531 scuds were collected and they were later kept in aerated tanks, filled with sea water, taken from the sampling site. The amphipods were fed every 4 days with a feed for ornamental fishes. A total of 346 scuds were randomly selected for two infection experiments involving eggs of P. laevis. The remaining 185 Gammarus specimens were necropsied to determine their infection level with larvae of the acanthocephalan studied, which would reflect the infection level of scuds in the natural environment.
In August 2002, a total of 36 scuds were collected in the same way as before, from the Pomeranian Bay (fisheries sub-divisions J-4 and J-5). They were necropsied, following their identification up to species level.
Part two: infection of scuds with the larvae of P. laevis under laboratory conditions Experiment one Forty live flounder were collected in March 2002 from the fisheries base at UnieÊcie. The fish were caught in fisheries sub-divisions J-4 and J-5. Ten randomly selected flounder were examined to prove their 100% infection with P. laevis. The remaining fish were transported in aerated tanks to a laboratory in Gdaƒsk. A total of nine fish arrived safely in good condition and they were placed in a 400-l, aerated tank with a constant water temperature of 10°C. Throughout the entire time-period of this experiment the fish were fed live shrimp, every 5 days, which also coincided with the water change in the tank. In May 2002, a total of 149 specimens of Gammarus spp. caught in the Gulf of Gdaƒsk were placed in the tank with the flounder. The scuds were protected by a cubic steel-wire cage (15 ◊ 15 ◊ 15 cm). The cage was covered by a plastic net. The stretched mesh size was 1 mm. Such arrangement enabled undisturbed flow of water and acanthocephalan eggs and it also protected the scuds against flounder predation. This culture was continued until July 2002. Past this date, the flounder were necropsied to determine the level of their infection. The fish were measured and aged, based on their otoliths. After their identification up to species level, the scuds were dissected in search for P. laevis larvae.
Experiment two A total of 218 mature females of P. laevis were recovered from 50 flounder caught in fisheries sub-divisions J-4 and J-5 in June 2002. The parasites were placed in a physiological solution in a petri dish, ruptured, and their eggs, containing acanthor were released. The dish was subsequently transferred to a brackish-water tank containing 197 scuds. The scuds were necropsied after 30 days, following their specific identification.
Part three: determining the specific composition of the flounder food. Thirty-two flounder were collected from the fisheries sub-divisions J-4 and J-5 and necropsied to determine their food in the natural environment.

RESULTS
Part one. Among the 50 scuds collected from the Gulf of Gdaƒsk in the autumn-winter season and 135 in the summer, no one was infected with the larvae of P. laevis. In this number were 41 G. salinus, 52 G. zaddachi, and 92 G. duebeni. A similar result was obtained after examination of G. salinus from the fisheries sub-divisions J-4 and J-5.
Part two Experiment one After two-month-long cohabitation of scuds (G. salinus, G. zaddachi, and G. duebeni) and flounder harbouring the adults of P. laevis, only one species of the amphipods (G. zaddachi) became infected with the acanthocephalan (Table 1). The infected scuds hosted from 1 to 5 encysted larvae of P. laevis in their hemocoel. The infection experiment yielded a total of 60 specimens of P. laevis, representing various phases of development. The length of the smallest cyst was 488 µm, while its width was 126 µm. The dimensions of the larva inside the cyst were 177 and 66 µm, respectively (Fig. 1). Lager cysts contained larger larvae and the increment in size was associated with progressive complication of their structure. A cyst, which was 771 µm long and 246 µm wide, contained a larva measuring 721 µm in length and 136 µm in width. Observations of its structure revealed the presence of partly developed proboscis and primordial testes-the anterior-and posterior one (Fig. 2). The subsequent phases of development are marked by a gradual filling of the space inside the cyst by the body of P. laevis larva (Fig. 3). The length of such larvae ranged from 1085 to 1348 µm. At this stage the body of the larva was distinctly divided into the anterior part (presoma), consisting of the proboscis and the neck and the posterior one (metasoma) containing primordia of the reproductive system. The length of presoma was 278 µm, while its width reached 139 µm.
Further complication of the external morphology of the larva consisted in appearance of the characteristic widening of the neck forming a spherical bladder called bulbus. At the same time the proboscis structure became also more complex, showing the presence of hooks. The reproductive system developed further and the larva was much larger than those described above. Depending on the specimen it was 2061-3619 µm long, with the proboscis attaining 254-291 µm. The bulbus diameter was about 120 µm (Fig. 4).
The length of larva, the most advanced in its development, was 4015 µm. Its proboscis was 259 µm long and 182 µm wide. The diameter of the bulbus was 240 µm. The neck was 605 µm long and 240 µm wide and the metasoma was 360 µm wide. The proboscis was armed with 18 longitudinal rows of hooks, 12 hooks in each row (Fig. 5).
By the end of this experiment, all flounder were necropsied and all of them harboured adult forms of P. laevis. A total of 178 P. laevis were found (71 males and 107 females). The mean infection intensity of this parasite was 19.8. The maximum intensity reached 32. In addition to the acanthocephalans, the fish also hosted the following two nematode species in their intestines: Cucullanus heterochrous and Hysterothylacium aduncum. The livers contained third stage larvae of H. aduncum. Four fish showed symptoms of lymphocystis. The fish were 3-to 5-year-old and their length ranged from 22 to 34 cm.

Experiment two
After 30 days of the experiment, no larvae of P. laevis were found inside the scuds exposed. All 197 scuds were identified as G. duebeni.
Part three The diet of flounder caught in the fisheries sub-divisions J-4 and J-5 consisted mainly of blue mussel, Mytilus edulis, which was in most cases accompanied by Gammarus salinus and Gammarus sp. Less frequently found were Macoma balthica, Mya arenaria, Crangon crangon, and in one case-Cardium glaucum. Each of the flounder studied hosted specimens of P. laevis. The mean infection intensity was 31.4. The maximum value of the infection intensity was 101. Other endoparasites found were: acanthocephalan Echinorhynchus gadi and nematodes Cucullanellus minutus, C. heterochrous, and H. aduncum. The length of the flounder ranged from 20 to 27 cm and they were 3 to 5 years old.

DISCUSSION
In the course of the present study, no scuds infected with larvae of P laevis were found in the natural environment. Under experimental conditions we were able to infect only one species-G. zaddachi.
Among the amphipods reported to be intermediate hosts of P. laevis in freshwater bodies were: G. pulex (cf. Hine and Kennedy 1974, Kennedy 1984, Kennedy 1996, Siddal and Sures 1998, G. pulex and G. roeseli (cf. Bauer et al. 2000), G. pulex and G. duebeni (cf. Kennedy et al. 1978), and Echinogammarus stammeri (cf. Maynard et al. 1998, Dezfuli et al. 2000. On the other hand, Kennedy (1996) was not able to find P. laevis larvae in freshwater G. zaddachi. This problem can be solved only after detailed studies on the biology of the intermediate-and final hosts and the parasite itself, because the successful transmission and development of P. laevis depends on their reciprocal relations.
A number of authors who studied this acanthocephalan paid attention to differences between Pomphorhynchus specimens acquired from marine environment with those from freshwater. Lundström (1942) and Kennedy (1984) believed that saltwater-and freshwater forms are different, though difficult to differentiate. Engelbrecht (1957) identified the specimens of Pomphorhynchus recovered by him from the Baltic Sea as P. laevis forma tereticollis, whereas Gibson (1972) reported acanthocephalans from the North Sea simply as Pomphorhynchus sp. If we assume that the two forms of this acanthocephalan represent separate sub-populations it is possible that two different environments affect the process of intermediate hosts selection by P. laevis. Another piece of evidence to support the above hypothesis is the fact that presently described attempts of infecting G. duebeni in brackish-water environment failed, while this amphipod, collected from a river, harboured larvae of P. laevis (cf. Kennedy et al. 1978).
It cannot be concluded, based on the above-mentioned data, that P. laevis is a nonspecific parasite capable of infecting different species of the genus Gammarus (cf. Bauer et. al. 2000). On the contrary, it is likely that different populations of P. laevis select a defined scud species, which guaranties them a successful completion of their life cycle. P. laevis is able to affect the behaviour of its intermediate hosts, which is a proof of its narrow specificity. The invertebrates hosting larvae of this acanthocephalan demonstrated an increased activity, positive reaction to light (swimming towards the light source), and drifting. Non-infected scuds exhibited the opposite reactions (Maynard et al. 1998, Dezfuli et al. 2000. Such behaviour of those invertebrates certainly increases the chances of transmission of P. laevis. It is easier for a fish to notice (and catch) a scud, intensively moving in the water column, rather than that penetrating the bottom. The behavioural changes of the infected amphipods can explain the high prevalence of P. laevis in flounder, despite the failure in finding the infected scuds in the bottom samples. It is presently suggested that the infection rate of Gammarus spp. with P. laevis in the Baltic Sea is low. Sulgostowska and Vojtková (1992) in their parasitological survey of 863 specimens if Gammarus spp. from the Gulf of Gdaƒsk and the Bay of Puck, were not able to find even a single scud infected with P. laevis. A similar result obtained Voigt (1991) who studied parasite fauna of scuds from the eastern Baltic Sea.
The chances of infection of flounder with this parasite can also be affected by migrations of this fish. In late autumn they migrate from the shore towards the closest depth, where they spawn in spring and after that they came back to the coastal feeding grounds (Ci´glewicz 1947). Their diet changes, on their route, which is associated with the distribution of invertebrates on the sea floor. According to Mulicki (1947) the food composition of flounder depends on the living area. During their regular migrations in the sea they may encounter a population of G. zaddachi, which is infected with larvae of P. laevis. They may also infect the scuds with eggs of P. laevis, which established themselves in the flounder intestines in the previous year.

CONCLUSSION
The results of the present study suggest that Gammarus zaddachi is probably an intermediate host for Pomphorhynchus laevis in the Baltic Sea.