Pentastomid parasites in fish in the Olifants and Incomati River systems , South Africa

LUUS-POWELL, WILMIEN J., JOOSTE, ANTOINETTE & JUNKER, KERSTIN. 2008. Pentastomid parasites in fish in the Olifants and Incomati River systems, South Africa. Onderstepoort Journal of Veterinary Research, 75:323–329 During parasitological field surveys of freshwater fish, sebekiid and subtriquetrid pentastome larvae were recovered from the body cavity or swim bladder of several fish species from various localities in Limpopo and Mpumalanga Provinces, South Africa. Sebekia wedli was recovered from the body cavity of Marcusenius macrolepidotus (Mormyridae) from Flag Boshielo Dam, Limpopo Province, and Alofia sp. and Subtriquetra rileyi were found in the swim bladder of Oreochromis mossambicus (Cichlidae) from the Phalaborwa Barrage, Limpopo Province. The latter species was also collected from the swim bladder of O. mossambicus in dams in the Phalaborwa region and the Ga-Selati River, Limpopo Province. A single specimen of Sebekia okavangoensis was present in the body cavity of Clarias gariepinus (Clariidae) in a dam on a sugarcane farm in the Komatipoort region, Mpumalanga Province. Pentastomid infections in the Mormyridae and Clariidae represent new host records.


INTRODUCTION
The Pentastomida represent an ancient taxon, comprising some 131 species in seven families (Almeida & Christoffersen 1999).Adults of most species inhabit the nasal passageways and lungs of snakes, lizards and crocodilians, while others are found in the air sacs of gulls and terns, and in the nasopharynx and sinuses of canids and felids (Bakke 1972;Banaja, James & Riley 1975;Riley 1986).Two spe-cies, Raillietiella bufonis and Raillietiella indica, use amphibians as final hosts (Ali, Riley & Self 1982).The pentastomid life cycle usually includes a vertebrate intermediate host in which larvae undergo several moults to reach the infective stage (Riley 1986;Winch & Riley 1986a;Riley & Huchzermeyer 2000), but insects have been reported as intermediate hosts of some raillietiellid parasites, while Reighardia sternae, a parasite of gulls, has a direct lifecycle (Banaja et al. 1975;Riley 1986).
In South Africa infective larvae of Sebekia wedli, Leiperia cincinnalis and Subtriquetra rileyi have been reported from the cichlids Tilapia rendalli (redbreast tilapia) and Oreochromis mossambicus (Mozambique tilapia) from the Phabeni Dam in the Kruger National Park (KNP) (Junker et al. 1998).Junker, Boomker & Bolton (1999) recorded the Nile crocodile, Croco dylus niloticus, as the definitive host of S. wedli and L. cincinnalis, as well as of Alofia nilotici, Alofia sim psoni, Sebekia cesarisi and Sebekia okavango ensis in South Africa.The final host of S. rileyi is as yet unknown.The pentastomid parasites included in this publication are products of a number of parasitological studies on freshwater fish from sev eral local ities in two provinces of South Africa.

MATERIALS AND METHODS
During fish health surveys, a total of 1 047 fish, belonging to 14 species were collected at nine localities in South Africa and examined for parasites (Tables 1 and 2).Fish were captured using gill nets of stretched mesh sizes, ranging from 30-120 mm, and transported live to the field laboratory where they were kept in con tainers with well aerated water.Immediately before dissection, fish were killed by decapitation.Encysted larvae were removed from their cysts and placed in water to unfold.In order to prevent them from contracting, they were fixed by adding small quantities of 70 % ethanol to the water over a period of approximately 1 h, after which they were transferred to 70 % ethanol.
Pentastomids were mounted and cleared in Hoyer's medium for identification.Measurements were taken from whole mounted specimens according to the schematic layout proposed by Riley et al. (1990).Hook and oral cadre morphology, combined with the number of annuli, were used as identification criteria.
The ecological terms prevalence, mean intensity and abundance are used in accordance with Bush, Lafferty, Lotz & Shostak (1997).Standard deviation was not calculated, since too few specimens were collected which renders standard deviation values meaningless (Rózsa, Reiczigel & Majoros 2000).

RESULTS
Pentastomid larvae were encountered at four localities in the Olifants River drainage system in Limpopo Province (Table 2).Eight encysted infective S. wedli larvae were collected from the body cavity of five of 29 Marcusenius macrolepidotus (Mormyridae) (bulldog) in Flag Boshielo Dam, constituting a new intermediate host record for this parasite.While a single cyst was located on the swim bladder, the remainder were found between fat deposits on the mesenteries.Cysts had a yellowish colour and closely resembled those of Clinostomum metacercariae, although slightly smaller in size.No free-living larvae were detected in the swim bladder.The prevalence of infection was 17.2 %, with a mean intensity of 1.6 (range 1-3) and an abundance of 0.3.
Subtriquetra rileyi was free-living in the swim bladder of O. mossambicus from dams in the PIC region.The prevalence of infection was 6.3 %, the mean intensity 1.9 (range 1-6) and the abundance 0.1.No pentastomid larvae were recovered from sharptooth catfish, Clarias gariepinus (Clariidae), the only other host examined at these sites at the time (Table 2).
At the Phalaborwa Barrage, S. rileyi, and a total of three infective larvae of an Alofia sp. were recovered from O. mossambicus.Concurrent infections of S. rileyi and Alofia sp.occurred in two hosts, with the former moving freely in the swim bladder while the Alofia sp. was encapsulated.Combined, the two species had a prevalence of 30 %, a mean intensity of 2.4 (range 1-5) and an abundance of 0.7.Clarias gariepinus and Labeo rosae (Cyprinidae) examined at this site harboured no pentastomid larvae (Table 2).This is the first published record of an Alofia species from fish intermediate hosts in South Africa.
In the Ga-Selati River, three S. rileyi larvae were recovered from one O. mossambicus of 36 examined, resulting in a prevalence of 2.7 %, and an abundance of 0.08.Larvae were moving freely in the swim bladder.No pentastomid larvae were found in C. gariepinus at this site (Table 2).
With respect to the Incomati River drainage system, Mpumalanga Province, a single infective larva of S. okavangoensis was encysted on the mesenteries of C. gariepinus, from a dam on a sugarcane farm in the Komatipoort region.This is the first report of C. gariepinus as intermediate host of a pentastomid.

DISCUSSION
The fact that S. wedli, S. okavangoensis and the Alofia sp. had reached the final larval stage, as evidenced by double hooks and rows of annular spines typical for infective sebekiid larvae (Winch & Riley 1986a), confirms that M. macrolepidotus, C. gariepinus and O. mossambicus are indeed the respective true intermediate hosts for these pentastomids.Similarly, Winch & Riley (1986a) found only the infective stage of Sebekia to be encysted.
Morphological characteristics of infective larvae of S. wedli from M. macrolepidotus correspond well with those recorded from O. mossambicus and T. rendalli (Junker et al. 1998).Ranging from 73-77, the number of annuli in infective larvae from M. macrolepidotus was similar to that observed in the latter hosts, namely 71-76 (Junker et al. 1998).
Other characteristics that support the identification of S. wedli were the single row of chloride cell pore caps on the anterior border of the annuli and the overall appearance of the oral cadre.The latter is readily distinguished from congeneric African species by appearing open anteriorly and having a less pronounced ovoid profile (Riley & Huchzermeyer 1995).
The arrangement of the chloride cell pore caps in an irregular field at the anterior border of the annuli, about 2-3 cells deep, and the mitre-shaped appearance of the oral cadre, observed in the specimen from C. gariepinus confirm its identification as S. oka vangoensis as described by Riley & Huch zermeyer (1995).
All infective Sebekia and Alofia larvae found during this study were encysted, the cysts being attached to either the swim bladder or the mesenteries.The literature suggests that sebekiids may occupy a number of sites in their intermediate hosts, and Overstreet et al. (1985) reported infective larvae of Sebekia mississippiensis under the connective tissues lining muscle, kidney, liver and swim bladder of a variety of fish intermediate hosts.
Sebekia wedli larvae appear to be host specific in Flag Boshielo Dam, as none of the other seven fish species examined at the time, including 12 O. mossambicus and three T. rendalli, harboured pentastomid larvae (Table 2).Sebekia wedli had a relatively high prevalence of 40.5 % in T. rendalli in the KNP, but only three of 119 O. mossambicus harboured this pentastome (Junker et al. 1998).Hence, it is possible that the sample size with respect to the other hosts in Flag Boshielo Dam was too small to detect infection.
Flag Boshielo Dam has a large population of Nile crocodiles.These crocodilians have long been established as a final host of S. wedli (Sambon 1922), andJunker et al. (1999) reported this pentastomid from Nile crocodiles in South Africa.Mormyrids are bottom feeders, favouring the muddy bottomed margins of rivers and floodplains (Skelton 2001), and are thus likely to ingest pentastome eggs, which are shed in crocodile faeces, and settle onto the bottom substrate.It remains a matter of speculation as to why S. wedli was not observed in intermediate hosts at any of the other dams, since both Loskop and Bly de Canyon Dams, as well as Tzaneen Dam also support crocodile populations.Possibly sample sizes at all these Dams were too small to detect pentastome infections.Prevalence data on pentastomids in Nile corocodiles in any of the above dams, including Flag Boshielo, are lacking.
Clarias gariepinus constitutes an important component of the diet of Nile crocodiles (Guggisberg 1972;Whitfield & Blaber 1979), and its own feeding habits would readily expose it to ingesting pentastome eggs.Riley & Huchzermeyer (2000) noted that Clari as spp.were common in swamp forest pools in the northern Congo Republic and postulated that it might serve as intermediate host for the pentastomids, Agema silvaepalustris, Alofia parva and S. oka vangoensis, that were recovered from swamp forest dwarf crocodiles, Osteolaemus tetraspis, in that region.However, no proof of the role of Clarias in the transmission of pentastomids has so far been presented.The prevalence of infection in C. gariepinus is low and only one of four hosts at the Komatipoort locality was infected, while a total of 112 collected at eight additional localities did not habour the parasite (Table 2).Ten C. gariepinus from the KNP also har-boured no pentastomids (Junker, unpublished data 1996).Winch & Riley (1986b) concluded that the absence of Subtriquetra subtriquetra in bottomfeed ing Tilapia spp. in Trinidad, despite its presence in Aequidens pulcher, another bottom-feeder in the same reservoir, was probably due to an immune response.This could also be an explanation for the low prevalence of pentastomids in C. gariepinus.(Junker 2002).Few data exist on the intermediate hosts of Subtriquetra, which has sofar been recovered from A. pulcher, O. mossambicus and T. rendalli (Winch & Riley 1986b;Junker et al. 1998).Fain (1961) reports S. subtriquetra from Brazilian fish in general.
While S. rileyi has now been reported from intermediate fish hosts from several localities in South Africa, its final host has not been identified.According to Riley et al. (1990) adult subtriquetrids are exclusive to crocodilians, but adult S. rileyi were absent in crocodiles harbouring other pentastome infections in the KNP, suggesting the involvement of a different final host, perhaps piscivorous terrapins or birds (Junker 2002).

TABLE 2
The prevalence of pentastomid infections in fish collected at various localities in South Africa Drainage