Introduction

Eye fluke diplostomids have complex life cycles, involving three hosts, of which two are intermediate and one is definitive (^{KARVONEN et al., 2006}; ^{VIOLANTE-GONZáLEZ et al., 2009}), and parasitise several fish species in the world (^{EIRAS, 1994}). Austrodiplostomum compactum (Lutz, 1928) is widely distributed in the Neotropical region, and A. compactum metacercaria parasitise several species of Brazilian freshwater fishes.

The A. compactum cercariae (first larval stage) infect aquatic snails and molluscs of the genus Biomphalaria Preston, 1910 in Mexico (^{VIOLANTE-GONZáLEZ et al., 2009}) and Brazil (^{PINTO; MELO, 2013}). Diplostomum and Austrodiplostomum metacercariae are frequently found in the eyes of freshwater fishes, which are the second intermediate hosts during the life cycle of the parasite (^{SEPPÄLÄ et al., 2004}). The adult stage parasitises the digestive tract of piscivorous birds, and Phalacrocorax brasilianus (Gmelin, 1789) [=Phalacrocorax olivaceus] is considered definitive host of this digenean in Brazil (TRAVASSOS et al., 1969; ^{NORONHA et al., 2009}; ^{MONTEIRO et al., 2011}).

According to ^{Eiras (1994)} and ^{Seppälä et al. (2005)}, the presence of Diplostomidae metacercariae in eyes of fishes can cause loss of vision, blindness and increases the coverage of cataracts of the lens area in fish with high rates of infection. This makes the fish more susceptible to predation and facilitates the transmission of the parasite to the definitive hosts (^{SEPPÄLÄ et al., 2004}).

However, a few Brazilian fish species were evaluated for the presence of A. compactum metacercariae and their possible effects. This study reports the occurrence of A. compactum in fish species from the Paranapanema River and its parasitological indexes, as well as several considerations regarding ecological aspects and a checklist for fishes parasitised in Brazil.

Materials and Methods

This study was conducted in the Chavantes reservoir (medium Paranapanema River), municipality of Ipaussu, São Paulo State, Brazil (23° 07′ 36″ S and 49° 59.23′ 10″ W) (Figure 1 and Table 1). Fish samples were collected monthly from March 2009 to February 2010 using gillnets (3 to 14 cm meshes between adjacent nodes) with 14 hours of exposure (from 5:00 PM to 7:00 AM). All captured fishes were frozen and transported to the laboratory where their eyes were removed and examined with a stereomicroscope. The metacercariae were collected from the vitreous humour, fixed in 70% alcohol under cover slip pressure and preserved in 70% alcohol (^{EIRAS et al., 2006}). The metacercariae were stained with carmine and cleared with eugenol for identification. Morphological analyses were performed with a computerised system for image analysis (Qwin Lite 3.2 – Leica Microsystems, Wetzlar, Germany).

Figure 1. Map of the hydrographical network in São Paulo State highlighting the Chavantes reservoir and sampling location on the Paranapanema River in the municipality of Ipaussu, São Paulo State, southeast Brazil. 

Twenty-three fish species were analysed: ‘peixe-cachorro’, Acestrorhynchus lacustris n=33 (Lütken, 1875), ‘canivete’, Apareiodon affinis n=41 (Steindachner, 1879), ‘lambari-do-rabo-amarelo’, Astyanax altiparanae n=37 Garutti & Britski, 2000, ‘lambari-do-rabo-vermelho’, Astyanax fasciatus n=30 (Cuvier, 1819), ‘saguiru-do-rabo-vermelho’, Cyphocharax modestus n=1 (Fernández-Yépez, 1948), ‘peixe-cadela’, Galeocharax knerii n=32 (Steindachner, 1879), ‘traíra’, Hoplias malabaricus n=9 (Bloch, 1794), ‘piava’, Leporinus amblyrhynchus n=11 Garavello & Britski, 1987, ‘ferreirinha’, Leporinus octofasciatus n=1 Steindachner, 1915, ‘pacu’, Piaractus mesopotamicus n=3 (Holmberg, 1887), ‘ximborê’, Schizodon nasutus n=17 Kner, 1858, ‘piranha’, Serrasalmus maculatus n=21 Kner, 1858, ‘sardela’, Triportheus nematurus n=6 (Kner, 1858), ‘caborja’, Hoplosternum littorale n=11 (Hancock, 1828), ‘cascudo’, Hypostomus regani n=4 (Ihering,1905), ‘mandiuva’, Iheringichthys labrosus n=17 (Lütken, 1874), ‘mandi-guaçu’, Pimelodus maculatus n=34 Lacépède, 1803, ‘bagre’, Rhamdia quelen n=1 (Quoy & Gaimard, 1824), ‘tucunaré’, Cichla kelberi n=42 Kullander & Ferreira, 2006, ‘joaninha’, Crenicichla haroldoi n=1 Luengo & Britski, 1974, ‘acará’, Geophagus brasiliensis n=2 (Quoy & Gaimard, 1824), ‘corvina’, Plagioscion squamosissimus n=30 (Heckel, 1840), and ‘tuvira’, Eigenmannia trilineata n=11 López & Castello, 1966.

The prevalence, mean intensity of infection and mean abundance were calculated according to ^{Bush et al. (1997)} for fish species with at least nine specimens collected. Parasite voucher specimens were deposited in the Coleção Helmintológica do Departamento de Parasitologia (CHIBB 6723, 6974-6985) at the Instituto de Biociências, Universidade Estadual Paulista (IB/UNESP), municipality of Botucatu, São Paulo State, Brazil. Fish voucher specimens were identified and deposited by Dr. Cláudio de Oliveira in the Coleção do Laboratório de Biologia e Genética de Peixes (LBP 4797, 9192, 9174, 9179, 9181, 9183, 9184, 9185, 9186, 9191, 9194) at IB/UNESP and the Coleção Ictiológica do Nupélia (NUP 6117), Universidade Estadual de Maringá, municipality of Maringá, Paraná State, Brazil.

A review of the fishes infected with A. compactum in Brazil was performed using books (^{EIRAS et al., 2010}) and by searching databases (SciELO, ISI, Scopus and Google Scholar). When available, data regarding the prevalence, mean intensity of infection and mean abundance along with their errors or standard deviations were added to the Table. In cases where these data do not exist but data does exist that enable their calculations, we performed the calculations according to the method described by ^{Bush et al. (1997)}.

Results

In the present study, the fish collected in the Paranapanema River showed a total of 357 A. compactum metacercariae, recovered from 43 fish specimens. Thirteen fish species were infected by at least one parasite (56.5% of the fish species analysed), C. kelberi, C. haroldoi, E. trilineata, G. brasiliensis, H. regani, H. littorale, H. malabaricus, I. labrosus, L. amblyrhynchus, P. mesopotamicus, P. maculatus, P. squamosissimus, and S. nasutus, of which, two were non-native species, C. kelberi and P. squamosissimus.

Of the thirteen species of fish parasitised (192 specimens), the prevalence was 22.3% (43 fishes). Six freshwater fish species were identified as new hosts for A. compactum metacercariae, C. haroldoi, E. trilineata, H. littorale, I. labrosus, L. amblyrhynchus, and P. mesopotamicus, totalling 46.2% of the infected fish species and 26.1% for all the fish species analysed. Three fish orders, Characiformes, Perciformes, and Siluriformes had four parasitised fish species, and Gymnotiformes had one infected fish species. The Cichlidae family had three infected species (23.1%), followed by Pimelodidae and Anostomidae, with two infected species each (15.4%) (Table 2).

The overall prevalence was 10.9%, and the highest prevalence was observed in P. squamosissimus (66.6%), followed by E. trilineata (54.5%) and I. labrosus (11.8%). The mean intensity of infection and mean abundance were low for all fish species, with P. squamosissimus having the highest values (13.1±6.1 and 8.7±4.2, respectively). All other fish species had values less than or equal to 2.0 for the mean intensity of infection and less than 1.0 for the mean abundance (Table 2).

Austrodiplostomum compactum metacercariae parasitised 36 Brazilian fish species and two taxa not identified (Cichla sp. and Hypostomus sp.) belonging to 13 families and four orders. The reports identified the fish species as follows: Characiformes (families: Anostomidae, three species; Curimatidae, one species; Erythrinidae, one species; and Serrasalmidae, four species); Siluriformes (families: Auchenipteridae, two species; Callichthyidae, one species; Loricariidae, eight species; and Pimelodidae, three species); Gymnotiformes (families: Gymnotidae, one species; and Sternopygidae, one species); and Perciformes (families: Centropomidae, one species; Cichlidae, 10 species; and Sciaenidae, one species). Eight infected fish species are non-native according to ^{Langeani et al. (2007)}: Cichla sp., C. kelberi, C. ocellaris, G. proximus, G. surinamensis, Metynnis maculatus, P. squamosissimus and Satanoperca pappaterra (Table 2).

Of the 36 fish species parasitised, only eight showed a mean abundance higher than five metacercariae. Of these fishes, six are non-native species of the order Perciformes. The non-native species, P. squamosissimus had the highest mean abundance of all the parasitised species (Table 2). The Paranapanema River showed the highest number of parasitised fish species (21), followed by the Paraná River (nine species), Tietê and Guandu Rivers (seven species each) and Grande River (two species). The other rivers had one fish infected species each (Table 2).

Discussion

Austrodiplostomum compactum metacercariae have been recorded in a wide variety of Brazilian fishes (Table 2). This study demonstrated that 56.5% of all fish species collected were infected. These fish species belong to four orders, Characiformes, Perciformes, Siluriformes and Gymnotiformes, and six new records of hosts to these metacercariae. Thus, the number of Brazilian fish species infected by A. compactum increased to 36, and the number of families parasitised for 13. Based on these observations, we hypothesise that this larval stage of the parasite has a low specificity for the second intermediate host (fish), similar to that reported by ^{Yamada et al. (2008)}. In addition, the six new hosts identified in the present study demonstrates that parasitological studies with fish communities should significantly expand the list of fish identified as second intermediate hosts for A. compactum.

The mean intensity of infection for the analysed fish species was lower than ^{Höglund and Thulin (1990)}, who reports that fish harbouring more than 40 Diplostomum metacercariae on their eyes are largely parasitized. According to several authors, the high infection rate with diplostomid metacercariae in fish can cause blindness or visual impairment (^{EIRAS, 1994}), cataracts (^{SHARIFF et al., 1980}; KARVONEN et al., 2004), problems with a food intake and growth (^{OWEN et al., 1993}) and increased susceptibility to predation (^{CROWDEN; BROOM, 1980}; ^{SEPPÄLÄ et al., 2004}). However, Karvonen et al. (2004) and ^{Owen et al. (1993)} report effects in fish with less than 10 Diplostomum spathaceum metacercariae (Rudolphi, 1819) and correlations between the number of metacercariae and the intensity of the effect. Another study reported that fish heavily infected with D. spathaceum occupied surface waters with more frequency, exposing them to higher levels of predation by birds. Therefore, fish with high infection rates would be subject to the effects of parasitism by A. compactum metacercariae, and fish with low infection show less intense effects of parasitism. Among these effects is the difficulty to detect prey, which results in fitness consequences for the individual (^{OWEN et al., 1993}).

The prevalence of A. compactum in the present study is low, with exception of the prevalence in E. trilineata and P. squamosissimus. Studies of the infection of A. compactum in P. squamosissimus showed a prevalence greater than 90% and a mean intensity of infection greater than 20 (^{KOHN et al., 1995}; ^{SANTOS et al., 2002}, ²⁰¹²; ^{MACHADO et al., 2005}; ^{PAES et al., 2010a}, ^b), whereas in the current study, the prevalence was 66.6% and the mean intensity of infection was 13.1±6.1. According to ^{Karvonen et al. (2006)}, the infection dynamics of D. spathaceum in fish is related to the snail population variance. ^{Voutilainen et al. (2009)} found a positive correlation between Lymnaea stagnalis (Linnaeus, 1758) density (an intermediate host for Diplostomum spp. in Finland) and Diplostomum sp. prevalence in lakes and ponds in Finland. Additionally, ^{Martins et al. (2002)} and ^{Santos et al. (2002}, ²⁰¹²⁾ suggested that high infection rates of A. compactum are dependent on high temperatures. According to ^{Berrie (1960)}, Diplostomum cercariae emerge in waters only at temperatures greater than 10 °C. Despite the high water temperature observed in the current study location (^{CARVALHO et al., 2012a}), which may lead to a high intensity of infection, the values of water transparency, conductivity and chlorophyll a (Table 1), suggest a small quantity of available nutrients in the local aquatic ecosystem. This can limit the size of the snail populations (first intermediate hosts), leading to the low values observed for the prevalence, mean intensity of infection and mean abundance for the majority fish species analysed.

The majority of fish species infected with A. compactum metacercariae in Brazil belong to the Perciformes (12 species), Siluriformes (13 species), and Characiformes (nine species), and the Cichlidae (10 species) and Loricariidae (eight species) families were the most infected. However, the other fish species analysed in this study also live in this habitat and were not infected. Therefore, the absence of A. compactum metacercariae infection in these fish species may be related to species-specific characteristics, which remain unknown.

For the fish species with higher mean abundances in Brazil, six are non-native species, and currently, P. squamosissimus has the highest mean abundance. This was also demonstrated by ^{Pojmanska and Chabros (1993)}, who showed that the prevalence of diplostomids was significantly lower in native fishes than non-native fishes. However, ^{Lacerda et al. (2012)} demonstrated an inverse situation in which the native fish H. malabaricus showed a higher abundance of Austrodiplostomum sp. than the introduced P. squamosissimus. ^{Lacerda et al. (2012)} also demonstrated that P. squamosissimus in native regions showed a prevalence and abundance of diplostomids that was significantly lower than P. squamosissimus in the region where they were introduced. According to ^{Lacerda et al. (2012)}, migratory birds, such as P. brasilianus are the final hosts for Austrodiplostomum; therefore, geographic barriers for the parasite may be reduced or eliminated. Rather than introducing a new parasite, P. squamosissimus may be acting as a new and suitable host for a local parasite. This may also occur with other species of fish, such as the six non-native Brazilian fish species, which had higher abundances.

The majority of infected fish species are in the Paranapanema River. Specifically, the Paraná basin is the most studied region for A. compactum in Brazil, and this is possibly related to the distribution of researchers in Brazil. Despite the large number of studies, further investigations are necessary to understand the relationship between environmental aspects and life strategies because diplostomiasis causes large aquaculture losses each year worldwide (^{PAPERNA; DZIKOWSKI, 1995}; ^{OVERSTREET; CURRAN, 2004}).