Introduction

Hoplias malabaricus Bloch, 1794 (Thraira), and Hoplerythrinus unitaeniatus Spix & Agassiz, 1829 (Aimara), are benthopelagic species of Erythrinidae with wide distribution in South and Central America. Both of these fish are sedentary and occur in several types of fluvial and lacustrine environments, especially in shallow water environments and near submerged or marginal vegetation, where prey abundance tends to be high, thus increasing their success in catching their prey. These fish are piscivorous when adults, but juveniles also feed on plankton, crustaceans, insects and seeds. These species are habitually active at twilight and night and do not undertake migration (^{SANTOS et al., 2006}; ^{SOARES et al., 2011}; ^{BENIGNO et al., 2014}). These fish are also commonly found in the Amazon river system, which is a complex ecosystem that includes tributary rivers and streams and floodplains and periodically flooded forest alongside them.

Hoplias malabaricus and H. unitaeniatus are important fishery resources in the feeding of human riverine populations from Amazon, and they are fish used in aquaculture and as ornamental fish (^{BENIGNO et al., 2014}). In Brasil, the farmed production of H. malabaricus in 2001 was 926.5 tons and the fished production was approximately 10,000 tons, whereas the fished production of H. unitaeniatus was 316.3 tons (^{BRASIL, 2013}). Thus, development of fishery programs requires information on the health of natural stocks. Knowledge of the parasites of natural populations of fish makes it possible to evaluate the impact of these parasites on their hosts, given that many parasites may influence the structure, size, feeding, growth rate and reproduction of natural populations (^{TAKEMOTO et al., 2009}; ^{MOREIRA et al., 2009}; ^{TAVARES-DIAS et al., 2014a}), along with the quality and acceptance of infected fish in the consumer market (^{BENIGNO et al., 2014}). Furthermore, studies on parasitic infections in fish populations provide an important increase in the information available on the parasite-host system.

In H. malabaricus, there have been reports of infections due to species of crustaceans, monogeneans, nematodes, digeneans, acanthocephalans, cestodes and hirudineans (^{CHAMBRIER et al., 1996}; ^{TAKEMOTO et al., 2009}; ^{ROCHA, 2011}, ^{ROSIM et al., 2011}; ^{BENIGNO et al., 2012}; ^{GRAÇA et al., 2013}; ^{BENIGNO et al., 2014}). On the other hand, H. unitaeniatus has been parasitized by species of crustaceans (^{MALTA, 1984}; ^{LEAL et al., 2010}), nematodes (^{MARTINS et al., 2005}; ^{BENIGNO et al., 2012}) and acanthocephalans (^{TAKEMOTO et al., 2009}). However, the structure of the parasite communities and infracommunities of these two hosts has not been studied. The present study compared the parasite communities and infracommunities of H. malabaricus and H. unitaeniatus from Amazon river system in Brazil.

Materials and Methods

Results

All the specimens of H. malabaricus were parasitized by one or more parasites, such as Ichthyophthirius multifiliis Fouquet, 1876 (Ciliophora); Piscinoodinium pillulare Schäperclaus, 1954, Lom, 1981 (Dinoflagellida); Tetrahymena sp. (Ciliophora); Braga patagonica Schödte & Meinert, 1884 (Cymothoidae); Urocleidoides eremitus Kritsky, Thatcher & Boeger, 1986 (Monogenoidea); metacercariae of Clinostomum marginatum Rudolphi, 1819 (Clinostomidae); larvae and adults of Procamallanus(Spirocamallanus) inopinatus Travassos, Artigas & Pereira, 1928 (Camallanidae); larvae of Contracaecumsp. (Anisakidae) and larvae of Nomimoscolex matogrossensis Rego & Pavanelli, 1990 (Monticelliidae). All the specimens of H. unitaeniatus were also parasitized by one or more parasite such as I. multifiliis; P. pillulare; Argulus pestifer Ringuelet, 1948 (Argulidae); Urocleidoidessp.; Anacanthorus sp.; Whittingtonocotyle caetei Santos-Neto, Rodrigues & Domingues, 2015; Whittingtonocotyle jeju Santos-Neto, Rodrigues & Domingues, 2015; (Monogenoidea), metacercariae of C. marginatum; larvae of Contracaecum sp.; larvae and adults of P.(S.) inopinatus; larvae and adults of Gorytocephalus spectabilis Machado Filho, 1959 (Neoechinorhynchidae) and larvae of N. matogrossensis (Table 1). However, for both hosts, I. multifiliis and P. pillulare predominated and the component community was dominated by ectoparasite species. There was an overdispersion of the parasites in H. malabaricus and H. unitaeniatus, except for Tetrahymena sp. in the gills of H. malabaricus, which showed random dispersion. In addition, Urocleidoides sp., Whittingtonocotyle spp. and N. matogrossensis presented random dispersion in H. unitaeniatus (Table 2).

The diversity of parasites was similar for H. malabaricus and H. unitaeniatus (Table 3). For H. unitaeniatus, hosts parasitized by three to five parasite species predominated and for H. malabaricus, hosts parasitized by four and five parasite species predominated (Figure 1). In H. malabaricus, the length of the hosts did not show any correlation with the species richness (rs = 0.263, p = 0.139), Brillouin index (rs = 0.163, p = 0.363) and evenness (rs = 0.283, p = 0.110). In H. unitaeniatus, the length of the hosts also did not show any correlation with the species richness (rs = –0.044, p = 0.817), Brillouin index (rs = –0.261, p = 0.164) or evenness (rs = –0.261, p = 0.164).

Figure 1 Species richness of parasites in two Erythrinidae fish from Amazonas river system (Brazil). 

In H. malabaricus, the mean Kn (1.01 ± 0.03; t = –4.28; p = 0.006) was greater than the standard, but in H. unitaeniatus, the Kn (1.00 ± 0.03; t = –5.04; p = 0.499) did not differ from the standard (Kn = 1.00), according to the t-test. The length and weight of H. malabaricus showed a weak positive correlation with the abundance of U. eremitus and larvae of Contracaecum sp. For H. unitaeniatus, only the weight of the hosts showed a weak correlation with the abundance of I. multifiliis (Table 4). However, for both hosts, none of the parasite species showed any correlation (p > 0.05) with Kn.

Discussion

The parasitic fauna of H. malabaricus was constituted by 10 taxons: three species of Protozoa, one of Isopoda, one of Monogenoidea, one of Cestoda, two of Nematoda and one of Digenea. The parasitic fauna of H. unitaeniatus comprised 12 taxons: two species of Protozoa, one of Argulidae, four of Monogenoidea, one of Cestoda, two of Nematoda, one of Acanthocephala and one of Digenea. However, the diversity was similar for the two Amazonian hosts. The parasite communities in H. malabaricus and H. unitaeniatus were dominated by the ectoparasites I. multifiliis and P. pillulare, in the same way as has been reported for other fish in the same area, i.e. the eastern Amazon region (^{BITTENCOURT et al., 2014}; ^{TAVARES-DIAS et al., 2014a}). However, among the endohelminths of H. malabaricus and H. unitaeniatus, larvae predominated, especially those of Contracaecum sp. and Clinostomum marginatum, which are parasites with low host specificity. Thus, these two fish species are also hosts for different species of endohelminths that are transmitted via the trophic chain. These fish can be used as hosts by different endohelminth species, but the diversity of the community of these endoparasites results from the interactions between the life histories and ecology of the hosts and parasites, among other factors (^{TAKEMOTO et al., 2009}; ^{MOREIRA et al., 2009}; ^{BITTENCOURT et al., 2014}; ^{TAVARES-DIAS et al., 2014a}).

The parasitic infections presented an overdispersion in H. unitaeniatus and H. malabaricus. Overdispersion, which are common in freshwater fish, suggest that the parasite-host relationship is stable and is influenced by environmental factors relating mainly to spatial-temporal changes in physicochemical parameters and to differentiated host susceptibility to parasites due to differences in immunological, behavioral and genetic characteristics among the hosts (^{MOREIRA et al., 2009}; ^{LOPES et al., 2009}; ^{POULIN, 2013}), along with differences in the dimensional proportions of sites in these hosts for parasitic infections. However, infection by Tetrahymena sp. in H. malabaricus, and by Urocleidoides sp., Whittingtonocotyle spp. and N. matogrossensis in H. unitaeniatus presented a random distribution, possibly due to the lower opportunity that these parasites had for colonizing such hosts. Therefore, the overdispersion pattern has some extremely important implications for the parasite population and its evolutive dynamics, and also for its hosts.

The gills of H. malabaricus and H. unitaeniatuswere the main sites of parasitic infection, due to the presence of protozoans, monogenoideans, crustaceans and metacercarial ectoparasites. In both hosts, I. multifiliis and P. pillulare predominated, but H. unitaeniatus presented higher intensity and abundance levels for I. multifiliis, and these levels increased with increasing host weight. Infections due to these two protozoan species are highly influenced by the characteristics of the local environment and by the sedentary behavior of these two hosts, which generally live close to vegetation in order to ambush their prey (^{SANTOS et al., 2006}; ^{SOARES et al., 2011}). These protozoa are known to proliferate in eutrophic environments like the Igarapé Fortaleza basin (^{BITTENCOURT et al., 2014}), which was the habitat de H. malabaricus and H. unitaeniatus in the present study. This was the first report of I. multifiliis and P. pillulare in H. unitaeniatus and the first report of P. pillulare for H. malabaricus.

Protozoa of the genus Tetrahymena are commonly found in organic material at the bottom of water bodies and they parasitize fish, amphibians, crustaceans and turbellarians (^{HERBERT & GRAHAM, 2008}; ^{COLORNI, 2008}). Tetrahymena sp. was only found in the gills of H. malabaricus, with greater prevalence (12.1%) than what has been described for Carnegiella strigata (0.89%) in the Negro river, central Amazon region (^{TAVARES-DIAS et al., 2010}). The present study is the second report on these protozoa for fish in Brazil. ^{Dickerson (2012)} stated that fish parasitized by Tetrahymena sp. may present protection against infection by I. multifiliis. This may have been the reason for the lower intensity and abundance of I. multifiliis that was observed in H. malabaricus, in comparison with H. unitaeniatus. On the other hand, factors relating to the innate immunity of H. malabaricus also cannot be ruled out.

In the gills of H. malabaricus, there was low parasitism of B. patagonica, which is a cymothoid without specific parasitic activity that also parasitizes other fish in the Amazon region (^{TAVARES-DIAS et al., 2014b}). The first report of A. pestifer for H. unitaeniatus comprised a low level of gill parasitism, similar to what has been described for Pseudoplatystoma trigrinum in Manaus, state of Amazonas (^{LOPES et al., 2009}). However, this parasitism of A. pestifer was lower than what was described for Pseudoplatystoma fasciatum and P. trigrinumfrom Janauacá Lake, (^{MALTA, 1984}) and for Pseudoplatystoma punctifer from Manaus region, both in state of Amazonas (^{LOPES et al., 2009}). Thus, A. pestifer is an argulid with low parasitic specificity and it depends on its hosts for geographic dispersion and distribution (^{MALTA, 1984}).

In the gills of H. malabaricus, the levels of infection due to U. eremitus were high in comparison with the levels of Urocleidoides sp., Anacanthorus sp., W. caetei and W. jeju in H. unitaeniatus. In the gills of H. unitaeniatus, the coexistence of four species of monogenoideans was due to the low levels of infection and random dispersion of Urocleidoides sp. and W. caetei and W. jeju. Urocleidoides eremitus has been found to be present in H. malabaricus in several hydrographic basins in Brazil, along with Urocleidoides naris, Urocleidoides cuiabai, Urocleioides brasiliensis and Dactylogyridae gen. sp. (^{ROSIM et al., 2011}). However, U. naris, U. cuiabai and U. brasiliensis seem not to have any distribution in the Amazon basin. Recently, in addition to these species of monogeneans, ^{Graça et al. (2013)} found Cosmetocleithrum bulbocirrus, Vancleaveus janauacaensis and Anacanthorus sp., along with another Dactylogyridae gen. sp. in H. malabaricus in the Paraná river (states of Paraná and Mato Grosso). In the present study, Urocleidoides sp. and Anacanthorus sp. in H. unitaeniatus are new species, and W. caetei and W. jeju are monogenoideans described recently by ^{Santos-Neto et al. (2015)}. The present study not only expands the distribution of U. eremitus to the eastern Amazon region, but also provides the second report of species of monogeneans for H. unitaeniatus.

In fish, heterogeneity of the composition of the endohelminth fauna and the presence of different parasitic stages can mainly be correlated with the hosts’ geographic distribution; habitat and way of life; age and longevity; position in the trophic chain; volume of food ingested; ontogenetic changes to immunocompetence and diet; and likelihood of contact with infective intermediate hosts in the environment (^{TAKEMOTO et al., 2009}; ^{MOREIRA et al., 2009}; ^{BITTENCOURT et al., 2014}; ^{TAVARES-DIAS et al., 2014a}).

Hoplias malabaricus and H. unitaeniatus are piscivorous hosts when adults, but they feed on plankton, crustaceans, insects and seeds when they are juveniles (^{SANTOS et al., 2006}; ^{SOARES et al., 2011}). Among the endohelminths of these hosts, larval stages predominated, particularly those of Contracaecum sp., along with metacercariae of C. marginatum. These parasites have zoonotic potential (^{BENIGNO et al., 2014}). Other fish species that are intermediate or paratenic hosts in the same region as that of the present study have commonly shown predominance of the larvae of digeneans and nematodes (^{TAVARES-DIAS et al., 2014a}; ^{BITTENCOURT et al., 2014}), as has H. malabaricus in other regions (^{TAKEMOTO et al., 2009}; ^{ROCHA, 2011}). Contracaecum sp. uses species of microcrustaceans as intermediate hosts and fish as secondary or paratenic intermediate hosts, while piscivorous birds are the definitive hosts (^{MOREIRA et al., 2009}; ^{MORAVEC, 2009}).

In Brazil, C. marginatum uses the gastropod Biomphalaria spp. as the primary intermediate host and fish as the secondary intermediate host, and the cycle is completed in piscivorous birds, which form the definitive hosts (^{PINTO & MELO, 2013}). Thus, in wild fish populations, the parasite transmission is through ingestion of prey (intermediate hosts) and variability of the feeding behavior of predatory fish like H. malabaricus and H. unitaeniatus may have a strong influence on the distribution of parasite species.

In wild populations of H. malabaricus (^{TAKEMOTO et al., 2009}; ^{ROCHA, 2011}) and H. unitaeniatus (^{BENIGNO et al., 2012}) in Brazil, the nematode fauna has been characterized by the presence of larvae of Contracaecum sp. and Eustrongylides sp., which are both important parasites from a public health point of view because of their zoonotic potential (^{TAKEMOTO et al., 2009}; ^{BENIGNO et al., 2012}). ^{Takemoto et al. (2009)} also recorded this nematode in H. malabaricus in the Paraná river basin. In the present study, H. malabaricus and H. unitaeniatus were infected with larvae of Contracaecum sp. and larvae and adults of P. (S.) inopinatus. Both of these are nematodes with low parasitic specificity. There were similar levels of infection with larvae of Contracaecum sp. in H. malabaricus and H. unitaeniatus, but this parasitism was much less than what was described by ^{Benigno et al. (2012)}, for both of these hosts on Marajó island, state of Pará (Brazil). Procamallanus (S.) inopinatus, a nematode that parasitizes fish in both their larval and adult stages (^{TAKEMOTO et al., 2009}), was found in H. unitaeniatus and H. malabaricus in the present study, but the infection levels in H. unitaeniatus were higher than in H. malabaricus. The present study has provided the first report of P. (S.) inopinatus for H. unitaeniatus.

For H. unitaeniatus and H. malabaricus, there were similar levels of infection by metacercariae of C. marginatum in the gills and intestine. The parasitism levels were also similar to those of metacercariae of Ithyoclinostomum dimorphum and metacercariae of Clinostomatopsis sorbens in the mesentery and/or musculature of H. unitaeniatus and H. malabaricus on Marajó Island, state of Pará, which is also in the eastern Amazon region (^{BENIGNO et al., 2014}). However, in other regions of Brazil, H. malabaricus has been infected with metacercariae of I. dimorphum, Clinostomum complanatum, Austrodiplostomum compactum and Sphincterodiplostomum musculosum, along with adults of Pseudosellacotyla lutzi (^{ROCHA, 2011}). These results indicate the role that these two hosts play in the life cycles of these digeneans.

The life cycle of the acanthocephalans require a species of aquatic vertebrate as the definitive host and a microcrustacean (amphipods, copepods, isopods or ostracods) as an intermediate host (^{ROCHA, 2011}; ^{TAVARES-DIAS et al., 2014a}). Hoplias malabaricus has been found to be parasitized mainly by Quadrigyrus machadoi, Q. brasiliensis and Q. torquatus (^{TAKEMOTO et al., 2009}; ^{ROCHA, 2011}) and H. unitaeniatus only by Q. brasiliensis (^{TAKEMOTO et al., 2009}). These are acanthocephalans with low parasitic specificity. The present study has provided the first report of G. spectabilis for H. unitaeniatus, with the presence of cystacanths and adults. The levels of infection were similar to those of cichlids in the same region as the present study that were parasitized by G. spectabilis (^{BITTENCOURT et al., 2014}; ^{TAVARES-DIAS et al., 2014a}). These results indicate that H. unitaeniatus is preying on microcrustaceans that contain infective forms of G. spectabilis, a parasite that has an unknown life cycle. On the other hand, H. malabaricus was not being infected with acanthocephalans, since the transmission and transportation of these parasites depends on the efficiency of the intermediate hosts, which acquire parasites through transferring them to other hosts. In addition, the size, development stage and species of the microcrustaceans ingested also significantly influence the likelihood that a fish might acquire infection due to acanthocephalans, which are endohelminths with a complex life cycle.

Planktonic microcrustaceans and cyclopoid copepods serve as intermediate hosts for larvae of species of Proteocephalidae. Fish become infected through ingesting these crustaceans, and the fish thus infected are especially the small ones that serve as food for larger predatory fish, which are hosts that carry higher levels of infection due to cestodes (^{SCHOLZ, 1999}). The levels of infection due to N. matogrossensis were similar between H. malabaricus and H. unitaeniatus, but were higher than those reported by ^{Chambrier et al. (1996)} for H. malabaricus in Paraguay. In H. malabaricus and H. unitaeniatus, the presence of larvae of N. matogrossensis indicates that these fish are intermediate hosts for this proteocephalid. This is the first report of N. matogrossensis for H. unitaeniatus.

The Kn of H. malabaricus and H. unitaeniatus was not negatively influenced by the presence of ecto and endoparasites, thus reflecting the low pathogenicity of the parasite community encountered. Hosts that acquire resistance to parasites through adaptation may also not be affected with regard to their body conditions (^{GUIDELLI et al., 2011}; ^{TAVARES-DIAS et al., 2014a}), depending on the organ infected and the parasite species, abundance and pathogenicity. However, the high Kn of H. malabaricus suggests that these fish were consuming greater quantities of food when they ingested the infective forms, which were transmitted via the trophic route, and thus had greater weight than expected.

In summary, the differences in parasitic fauna composition in this study were caused by occurrence of Tetrahymena sp. and B. patagonicain H. malabaricus, and absence of these parasites in H. unitaeniatus, along with absence of G. spectabilis and A. pestifer in H. malabaricus. All of these species lack parasitic specificity. Moreover, the parasite community was dominated by ectoparasites, which were favored by the way of life of H. unitaeniatus and H. malabaricus, which are hosts occupying a high position in the trophic chain, thereby contributing towards the presence of endohelminths. Thus, the results from this first investigation on the parasitic communities of these two piscivorous fish indicate that both of these hosts play an important role in the life cycles of these nematodes, digeneans, acanthocephalan and cestode, which were found at moderate infection levels. Parasite abundance and species richness were not affected by host body size, because the correlations were weak, as shown by the extremely low correlation coefficient values (body size explained less than 40% of the variance relating to parasite abundance). Thus, this indicates that factors other than host body size are more important determinants of variation of parasite abundance and species richness for the populations of both of these hosts. This was the first report on parasite diversity in wild H. unitaeniatus and H. malabaricus and it showed that these species had similar parasite communities, characterized by low species diversity, low evenness and species richness, and by high prevalence and diversity of ectoparasite species.