Observations on the Infection by Kudoa sp . ( Myxozoa , Multivalvulida ) in fishes caught off Rio Grande , Rio Grande do Sul State , Brazil

It is reported the parasitization of Kudoa sp. (Myxozoa, Multivalvulida) within the somatic muscles of the fish Odontesthes bonariensis (Valenciennes, 1835), Micropogonias furnieri (Desmarest, 1823) and Mugil liza Valenciennes, 1836, captured off Rio Grande, Rio Grande do Sul State, Brazil. Other species of fish caught at the same place were not infected. The parasites formed elongated plasmodia inside the muscle fibres. Myoliquefaction or any host tissue reaction were not observed. The morphological and morphometric features of the spores are described, the parasites are compared with other Kudoa species, and the importance of the infection is discussed.


Introduction
Kudoa spp.(Myxozoa, Multivalvulida) are frequent parasites of marine and estuarine fish.A recent synopsis of the species reported the existence of 95 nominal species (Eiras, Saraiva, & Cruz, 2014b) infecting a wide range of hosts all over the world.Since then, at least 6 new species were described, one of them infecting a freshwater host (Azevedo et al., 2015).
A little more than half of the parasites locate within the skeletal muscle fibers, and a number of species are especially important once they cause liquefaction of the host muscle within a short period after the death of the host which becomes useless for consumption (Cruz, Silva, & Saraiva, 2011;Gilman & Eiras 1998;Henning, Hoffman, & Manley, 2013;Moran, Whitaker, & Kent, 1999;Romero & Burgos 1996;Yokoyama & Itoh 2005;Yokoyama, Whipps, Kent, Mizuno, & Kawakami, 2004).This can occur in both wild and cultured fish, and the infection may be decisive for the economic success of fisheries and farming.

Material and methods
Fresh fish specimens were acquired from fishermen at Rio Grande fish market.In the laboratory they were identified, measured (total length) and inspected for Kudoa infection.Small samples of muscle, collected from both sides of the specimens, were crashed between two glass slides and inspected at the stereomicroscope.When detected, the plasmodia were preserved in formalin at 4%, and latter 15 to 25 spores were examined at the microscope and measured according to Burger and Adlard (2010).The spores dimensions were compared using Kruskall-Wallis test, followed by multiple comparisons when significant differences (p < 0.05) were detected.The eventual myoliquefactive action of the parasites was assessed by visual inspection of the muscle to evaluate the integrity of the muscle fibers, and by palpation of the fish to detect softening of the muscle.

Results
The fish species examined, number and total length of the specimens are indicated in Table 1.
The parasites were located inside the skeletal muscles fibers in the anterior part of both sides of the body forming small intracellular elongate plasmodia, with both extremities round or slightly tapering.The dimensions of the plasmodia were similar in all the hosts, measuring 0.72 -1.0 mm in length by about 0.07 mm wide.All the plasmodia were in the same developmental stage containing only mature spores.The spores were also morphologically similar in all the hosts.They were bell-like shaped in lateral view.In apical view they were stellate, quadrate, having four equal radiating shell valves with margins smoothly curved, without projections or extensions.The four polar capsules were slightly pyriform, elongated, equally sized with pointed anterior extremities.The number of coils of the polar filament was not visible in optic microscopy.
Despite the similar morphology of the spores of the three different hosts there were consistent differences concerning the morphometry of the spores.Considering the three forms observed the specimens from O. bonariensis presented the highest dimensions and those of M. furnieri were smaller than M. liza for spore width and spore thickness.M. furnieri spores were clearly smaller than those from O. bonariensis and M. liza, and those from M. liza presented intermediate values -(Table 2).1.2 ± 0.2 (0.7 -1.3) 15 ab Significant differences p = 0.000 p = 0.000 p = 0.000 p = 0.000 p = 0.002 p = 0.000 p = 0.017

Discussion
The morphometry of the spores and statistical data lead us to the conclusion that the parasites from the three hosts correspond at least to two different species, one infecting M. furnieri, and another infecting O. bonariensis and M. liza.However, the existence of three different species can not be ruled out, and a molecular study of the specimens would be necessary to clarify this question.
Our specimens were compared with all the Kudoa spp.described for Brazilian hosts, muscle parasites from hosts from other sampling places, and parasites with other location than the fish muscles.
From the comparison with species from Brazilian hosts it is obvious that our specimens belong to different species due to differences in the size of the spores and polar capsules.
From the about 100 nominal species described so far a little more than half of the species parasitize the muscles of the hosts.The comparison of the present material with all the species strongly suggests they correspond also to different species.
The specimens from O. bonariensis are similar to K. petala from Sillago sihama in China, though a little smaller in length and a little bigger in width.The polar capsules of K. petala are drop-like, while in our material they are elongated, and K. petala is coelozoic in the gall bladder (Zhou & Zhao, 2008).Therefore the identity of both species is unlikely.Our specimens are also similar to K. scomberi in the somatic muscles of Scomber japonicus in Japan, forming plasmodia seen as minute white cysts, having spores longer (6.4) and thicker (8.1) than our specimens and larger drop-like polar capsules (Li et al. 2013).For these reasons the two species are considered different.Some similarities exist with K. shiomitsui detected in the pericardial cavity and heart of Takifugu rubripes (Egusa & Shiomitsu 1983).However, the spores from Brazilian hosts are not so wide, and the different location within the hosts separates both forms.Interestingly, they are very similar in dimensions to K. permulticapsula (6.0 x 9.1 x 7.7) infecting the somatic muscles of S. commerson in Australia (Whipps, Adlard, Bryant, & Kent, 2003).However, K. permulticapsula has 13 polar capsules (occasionally 14, rarely 15), which distinguishes immediately the two species without further considerations.
Concerning the specimens infecting M. liza, they do not match any species, namely those infecting Mugil spp.: the dimensions of the spores and polar capsules are higher than those of K. bora infecting the muscles of M. japonicus (Nigrelli, 1946).They are smaller than K. haridase in the gall bladder of M. persina (Sarkar & Gosh 1991), and larger than K. intestinalis from the muscular layer of intestine of M. cephalus (Maeno, Nagasawa, & Sorimachi, 1993), and smaller than K. tetraspora located around the optic lobes of M. cephalus (Narasimhamurti & Kalavati 1979).Other differences relate to the size and shape of the polar capsules and the shape and dimensions of the plasmodia.As a consequence our forms can not be identified with any of the species infecting Mugil spp., conclusion which is re-inforced by the different location of the parasites within the hosts, excepting K. bora which infects also the muscles.Comparison with other species failed to show forms identical to the present material.
The parasites of M. furnieri are distinct from all the other species.The most identical one is K. azoni infecting Pleurogrammus azonus in the Sea of Japan (Amur Bay and Ussuri Bay).However it locates between, and not inside the muscle fibres, and causes post-mortem myoliquefaction (Aseeva, 2004), and can not be identified to the parasites from M. furnieri.
It is concluded that the Kudoa spp.observed infecting O. bonariensis, M. liza and M. furnieri are most probably species not described hitherto, and correspond to two or three different species.However, and due to lack of molecular studies, we refrain for the moment further considerations about their identity.
Considering the high diversity of marine fish from Brazil, estimated at around 1,300 species, it is surprising that so few species of Kudoa were reported so far.The same situation occurs generally in South America.As far as we are aware only 6 species of Kudoa were reported for South American countries besides Brazil (Argentina, Chile, Peru and Uruguai), and Kudoa sp. were found in Brazil, Chile and Peru, infecting in all only a total of 14 different host species.Clearly more research is needed on these parasites.It is important to emphasize that, besides the scientific interest of the parasites, the study of its biology is important due to the host myoliquefaction caused by a number of species which may impair the commercialization and the farming of some host species.

Conclusion
The Kudoa spp.observed in fish caught off Rio Grande increase the knowledge about these parasites infecting marine fish in Brazil.Taking into account the low number of species described for the country it is emphasized the importance of the study of these parasites.Furthermore, the fish infection by Kudoa may have important economic consequences concerning the exploitation of wild and farmed fish due to the dramatic changes which some species induce in the host muscle.This is one more and important reason for the study of these parasites in Brazil.

Table 1 .
Fish species examined for Kudoa infection, number and total length (cm) of the specimens.