Introduction
Farmed fish are often exposed to inadequate environmental factors including excessive organic matter and high population density which result in reduced dissolved oxygen ( GHIRALDELLI et al., 2006 ; ZANOLO & YAMAMURA, 2006 ). Unfavorable conditions induce a stress response in fish ( TEROVA et al., 2005 ) which predisposes them to a number of diseases including parasitic infestations ( ENGERING et al., 2013 ). High parasite loads can cause irreversible damage to both integument and gills of infected fish ( PAPERNA, 1991 ).
Crustaceans are one of the most significant pathogens in aquaculture especially those of the subclasses Branchiura and Copepoda ( JITHENDRAN et al., 2008 ). and occur in fish farms and in natural habitats of fish as well ( LUQUE et al., 2013 ; TAVARES-DIAS et al., 2015 ), Argulus and Dolops are the genera of branchiuran crustaceans most frequently found in fish ( MARTINS et al., 2002 ; TAVARES-DIAS et al., 2007 ). These crustaceans are blood-sucking parasites and are capable of attaching to the integument and gills of fish ( THATCHER & BRITES-NETO, 1994 ).These arthropods move rapidly, and break through the epithelium of the fish skin using their oral appendages ( CARVALHO et al., 2004 ). Lernaea and Lamproglena are the genera of copepods most often described in fish ( LUQUE & TAVARES, 2007 ). These crustaceans attach firmly to the integument and gills of their fish hosts causing marked epithelial damage and severe vascular injury ( INNAL et al., 2017 ).
Mortality outbreaks were reported in production fish parasitized by crustaceans ( CARNEVIA & SPERANZA, 2003 ), such as reports of epizootics in Atlantic salmon infested by sea lice (Copepoda, Caligidae), in which they cause intense spoliation, loss of epithelium, hemorrhages, increased secretion of mucus, and necrosis of the affected tissue ( COSTELLO, 2006 ). Disease outbreaks and mortality caused by crustaceans have been reported in farmed fish ( CARNEVIA & SPERANZA, 2003 ). In epizootics of sea lice (Copepoda, Caligidae) infestation in Atlantic salmon, affected fish had severe spoliation, loss of the skin epithelium, hemorrhage, increased mucus secretion, and tissue necrosis ( COSTELLO, 2006 ). In addition to the damage caused by branchiuran crustaceans and copepods on their fish hosts, these arthropods can also act as disease vectors. Caligus rogercresseyi (sea louse) was recently described as a mechanical vector of Infectious salmon anemia virus ( NYLUND et al., 1993 ; OELCKERS et al., 2014 ).
Parasitic crustaceans may also serve as substrate for epibionts such as Epistylis spp. ( SILVA et al., 2011 ; AZEVEDO et al., 2014 ; CÔRREA et al., 2016 ). These protozoa are often considered commensal organisms in fish farms. However, in adverse environmental conditions they can cause damage to the integument of Oreochromis niloticus (Nile tilapia) ( VALLADÃO et al., 2015 ) and Pseudoplatystoma reticulatum (barred sorubim)/Pseudoplatystoma corruscans (spotted sorubim) hybrids ( PÁDUA et al., 2016 ).
The objectives of the present study are to report the association between Epistylis spp. and parasitic crustaceans in the headstander Megaleporinus obtusidens (known as piava or piapara in Brazil) and to describe the lesions produced by ciliate and crustacean co-infections in this fish host.
Materials and Methods
Fish and study area
Megaleporinus obtusidens (Characiformes: Anostomidae) (known as piava or piapara in Brazil) is an omnivore native fish with a wide geographical distribution, and occurs in the watersheds of the southern and southeastern regions of the country ( HARTZ et al., 2000 ).
This species of fish is ideal for intensive cultivation, and has good consumer acceptance due to the taste of its meat and its size as well; it can weigh up to 7.5 kg ( GLUSCZAK et al. 2006 ; ADORIAN et al. 2017 ).
In this study, fish examined for the presence of parasites originated from an extensive fishing establishment located in the county of Araraquara, State of São Paulo, southeast Brazil. They were cultivated in earth ponds and marketed to several fish-and-pay fishing camps in the area. The producer claimed that several fish from this farm died without any apparent cause, and that affected fish were coughing at the water entrance of the tanks.
Parasite analysis
Five moribund adult piavas were captured. In order to avoid detachment of parasites from the skin, fish were euthanized immediately after being captured; the spinal cord was severed with a knife or scalpel. Mucus was collected from the skin and gills of these fish by gently scrapping off their body and branchial surfaces with a glass slide. Mucus samples were examined under the light microscope. Parasitic crustaceans that were found in these specimens were gently removed from the surface of the glass slide with anatomical tweezers. Some parasites were stored in 70% alcohol while others were fixed in 2% glutaraldehyde for subsequent scanning electron microscopy analysis. Identification of parasitic crustaceans was based on the morphological characters of these arthropods and according to the information published by Robinson & Avenant-Oldewage (1996) and Thatcher (1991) . Images were captured using a Nikon microscope E200® equipped with a Motic 5.0 image capture system. The ecological descriptors are in accord those published by Bush et al. (1997) .
Histopathological analysis
Gill specimens were fixed in 10% formaldehyde and then preserved in 70% alcohol.
These tissue samples were dehydrated in alcoholic solutions of increasing concentrations and then embedded in paraffin. Sections of 5 μm were obtained using a microtome, and were stained with hematoxylin and eosin. All lesions were photographed under a light microscope.
Scanning electron microscopy analysis
Crustacean and gill samples were submitted for scanning electron microscopy analysis. Specimens were fixed in glutaraldehyde with 2% sodium cacodylate buffer and washed overnight in osmium tetroxide. Samples were dehydrated in alcoholic solutions of increasing concentrations. Critical point drying was used to remove water from these tissues. After that, samples were carefully placed in aluminum specimen stubs and bathed with gold. Images were captured using a JEOL JSM- 5410 electron microscope.
Results
Parasite identification
All fish were parasitized with the crustaceans Dolops carvalhoi ( YAMAGUTI, 1963 ) and Lernaea cyprinacea ( EIRAS et al., 2010 ) and by the epibiont Epistylis spp. with prevalence rates of 100% for all three parasites. The mean parasite intensity of Dolops carvalhoi was 155.6 ± 61.31 (106-229). We were unable to count Lernaea cyprinacea since these copecods are fairly small and difficult to visualize with the naked eye, and fish were considered too large [average weight 2.22 kg ± 0.37 (1.8-2.7) and total length 60.6 cm ± 8.11 (54-74)] to be examined under the stereomicroscope.
Host-parasite relationship
Large numbers of the ciliate Epistylis spp. were found attached to the integumentary surface of both D. carvalhoi and L. cyprinacea ( Figure 1 ). Epistylis spp. was also attached to the secondary lamellae of each gill filament from these fish. Tissue damage of varying severity was present in all fish infested with parasites. Microscopic lesions observed in the parasitized gills included hyperplasia and hypertrophy of the lamellar epithelium, an inflammatory infiltrate composed of eosinophils, telangiectasia, foci of hemorrhage and necrosis, extensive fusion of the secondary lamellae, and detachment of the lamellar epithelium (epitheilai lifting) ( Figure 2 ).

Figure 1 a) Dolops carvalhoi in ventral position with the intense presence of the adhesion of the protozoan Epistylis spp., scale bar: 1 mm. b) In detail a hook of the crustacean Dolops carvalhoi demonstrating the presence of Epistylis spp., scale bar: 200 µm. c) Panoramic view of a Lernaea cyprinacea, scale bar: 1 mm. d) Adhesion of Epistylis spp. in the cephalothorax of the crustacean Lernaea cyprinacea , scale bar: 200 µm.

Figure 2 a) Lernaea cyprinacea adhered to a branchial filament, causing intense hypertrophy and hyperplasia epithelial, scale bar: 200 μm. b) Crustacean parasite and Epistylis spp. causing lamellar damage, with areas of hemorrhage and necrosis, and severe inflammatory infiltrate, scale bar: 200 μm. c) Epistylis spp. near areas of epithelial hyperplasia and scaling, scale bar: 200 μm. d) In detail, Epistylis spp. parasitizing branchial filament, scale bar: 50 μm.
Discussion
Fish ectoparasites are important mechanical vectors in aquatic environments, and are capable of transferring pathogens from infected fish to healthy fish or between wild fish and farmed fish ( XU et al., 2007 ). Crustaceans are mechanical vectors of a number of pathogens in aquatic ecosystems including viruses, bacteria, and parasites. Parasitic crustaceans can have a major economic impact on commercial fish culture, and can pose a significant threat to species diversity within an aquatic habitat ( OVERSTREET et al., 2009 ). Gnathiid isopods, for example, are vectors of haemoggregarines which are haemoprotozoans that can cause mild chronic infection in fish from coral reefs ( DAVIES & SMIT, 2001 ). The pathology of hemogreggarine infections in fish, albeit largely unknown, may have marked deleterious effects on fish farms ( SMIT et al., 2006 ). The interaction between the branchiurian Argulus foliaceus (fish louse) and Rhabdovirus carpio, which causes spring viraemia of carp, is another example of a crustacean acting as a vector for diseases in fish. Crustaceans carrying Rhabdovirus carpio attach to the body surface of Cyprinus carpio (common carp) and transmit the virus to these fish ( AHNE, 1985 ). High mortality rates have been recorded for spring viremia of carp ( DIKKEBOOM et al., 2004 ).
To date, there are no publications reporting that the crustacean D. carvalhoi is capable of transmitting pathogens to fish even though other parasitic crustaceans may act as basibionts for Epistylis spp. including Ergasilus chelangulatus ( AZEVEDO et al., 2014 ). In the present study, L. cyprinacea was found in fish skin mucus samples examined by light microscopy. Free-living crustaceans may also act as basibionts for ciliates. An example of such interaction is the association between the shrimp Macrobrachium rosenbergii ( MANDAL et al., 2015 ) and the planktonic copepods Pseudodiaptomus stuhlmanni ( JONES et al., 2016 ), Thermocyclops minutus, and Notodiaptomus amazonicus ( CABRAL et al., 2016 ).
The ciliate epibionts Epistylis spp. are considered commensal protozoan organisms ( VALLADÃO et al., 2015 ) but have been associated with a decrease in food conversion efficiency and increased susceptibility to predation in crustaceans ( VISSE, 2007 ). Furthermore, Epistylis spp. contributed to mortality of the copepod Boeckella triarticulata in an environment with limited food availability ( XU & BURNS, 1991 ). Fish parasitized by crustaceans for prolonged periods of time may develop chronic stress. Fish are more susceptible to infections, including epistiliasis, due to stress-induced immunosuppression ( PICKERING & POTTINGER, 1989 ; TORT, 2011 ; WALKER et al., 2004 ). Parasitic crustaceans act as carriers for Epistylis spp., and disseminate this protozoan organism in fish farms.
This ciliate can cause a variety of lesions in fish such as damage to the scale epithelium including hemorrhage and necrosis ( WELICKY et al., 2017 ). Epistylis spp. causes injury to the integument of the Nile tilapia Oreochromis niloticus which is characterized by the presence of a subepithelial infiltrate of lymphocytes and mast cells and epithelial hyperplasia and hypertrophy ( VALLADÃO et al., 2015 ). Microscopic lesions that are seen in the integument of cultured cichlid fish and siluriformes (catfish) in association with the cilitate Epistylis include hydropic degeneration, multifocal necrosis, proliferation of skin mucous (globet) cells, and the presence of a mixed inflammatory infiltrates composed of granulocytes and mononuclear cells ( PÁDUA et al., 2016 ). Studies have shown that Epistylis spp. is involved in severe outbreaks of Aeromonas hydrophila infection in farmed fish with high mortality. The ciliate is capable of propagating bacteria which adhere to their peduncles ( HAZEN et al., 1978 ; MILLER & CHAPMAN, 1976 ).
The gill lesions found in our study are consistent with those previously described by other researchers in cases of macroparasite infection of fish ( BYRNE et al., 2002 ; MARQUES & CABRAL, 2007 ). Branchial lesions in which the parasite is present are severe whereas those in adjacent areas without the parasites are mild. Since branchial lesions are usually focal, most of the unaffected gill surface area is still functional which allows fish to survive. Death ultimately occurs only in those cases of heavy parasite infestation in which extensive and severe gill lesions are present.
It is imperative to actively monitor parasitic diseases in farmed fish as they can cause significant problems to global aquaculture production resulting in major economic losses. In farmed fish, parasitic crustaceans cause erosions on the integument and gills and can also spread many pathogens including Epistylis spp. Further studies are needed to assess the impact of Epistylis spp. on fish health. The mechanisms involved in the attachment of parasitic crustaceans on the integument of fish should be investigated in order to further our understanding of the interactions between these arthropods and their fish hosts. Crustaceans parasites are important mechanical vectors of Epistylis spp. infection and Epistylis-Dolops-Lernaea complex affects the health of fish.