Community structure of metazoan parasites of silverside, Odontesthes bonariensis (Pisces, Atherinopsidae) from Argentina

The helminth communities of silverside, Odontesthes bonariensis (Valenciennes, 1835), from two Argentinean lagoons were studied and compared at component community and infracommunity levels. Nine helminth species were found: five digeneans (Austrodiplostomum cf. mordax, Ascocotyle (Phagicola) cf. diminuta, Ascocotyle sp., Thometrema bonariensis and Saccocoelioides sp.); two nematodes (Contracaecum sp. and Hysterothylacium sp.); one acanthocephalan (Wolffhugelia matercula) and one cestode (Cangatiella macdonaghi). Odontesthes bonariensis is a new host record for five parasite species. Richness, diversity and number of helminths in silversides from Salada Grande lagoon were higher than in those from Lacombe lagoon. This could be related with lagoon size, abundance of mollusks and fish-eating birds, and size and diet of silversides captured in each lagoon. In Salada Grande lagoon the helminth community of silversides was dominated by the allogenic and generalist species A. cf. mordax; while the autogenic and intermediate specialist species C. macdonaghi was dominant in Lacombe lagoon. Host sex did not affect richness, diversity or total abundance, whereas host size was positively correlated with these attributes, except diversity in Salada Grande lagoon.

The silverside, Odontesthes bonariensis (Valenciennes, 1835) is one of the most important freshwater commercial and sport fishing resources in the Pampean region of Argentina. It is the species most used for restocking due to its high adaptability and economic importance, and it has been introduced in numerous freshwater environments in Argentina and other countries (Japan, Italia, Peru, Bolivia and Chile) (grosman, 2001).
Silversides are zooplanktivores, although capable of exploiting other habitat resources such as insects, mollusks, other invertebrates and fishes, and even cannibalism, when plankton is scarce (escalante, 2001).
The goal of this work is to analyze the helminth community of O. bonariensis, at the component and infracommunity levels, in two lagoons from Buenos Aires Province, Argentina.
Ecological terminology follows Bush et al. (1997) and the specificity was classified according to DesDeVises et al. (2002). The description of infracommunity structure was based on helminth species richness, mean number of parasites, diversity, evenness, and dominance. Shannon's Diversity Index (H) was calculated using base 10 logarithms, and the Berger-Parker index of dominance was calculated following magurran (1988). All helminth species, irrespective of their rareness in helminth faunas, were considered for the calculation of these attributes. Possible effect of host sex on helminth species richness, diversity and total abundance was assessed using Mann-Whitney test (U), and possible influence of host standard length was tested by Spearman's rank correlation coefficient (r s ). A Kruskal-Wallis test (KW) was used to compare mean values of richness, diversity and number of helminth per infracommunity among seasons. The correlation between standard length of silversides and number of metacercariae of A. cf. mordax was tested by Spearman's rank correlation coefficient (r s ).

RESULTS
Composition of the helminth community. Nine species of helminths were found parasitizing O. bonariensis, five of them as larval stages. Five species were found in silversides from both lagoons: A. cf. mordax, C. macdonaghi, Wolffhugelia matercula Mañé Garzón & Dei-Cas, 1974 (Acanthocephala-Neoechinorhynchidae), Contracaecum sp. and Hysterothylacium sp. (Nematoda-Anisakidae). Four species were found exclusively in silversides from SGL: Thometrema bonariensis Lunaschi, 1988 (Stunkard & Haviland, 1924) and Ascocotyle sp. (Digenea-Heterophyidae). Three digenean species, one nematode species and one acantocephalan species are reported for the first time in O. bonariensis (Tab. I). The highest prevalence values were observed for Ascocotyle sp. and C. macdonaghi in silversides from SGL and LL, respectively; while the most abundant species were A. cf. mordax and C. macdonaghi in SGL and LL, respectively (Tab. II).
Colonization strategy. All helminth species found in this study have complex life cycles with at least Tab. I. Helminth parasites found in Odontesthes bonariensis from Salada Grande and Lacombe lagoons. Diet. The diet of O. bonariensis in both lagoons consisted primarily of microcrustaceans (Copepoda and Cladocera), while others items were rare (Tab. III). However, higher variety of food items was found in the silversides from SGL. Component community structure. Table  IV shows summarized data on helminth component community structure. Silversides from SGL showed higher helminth species richness and diversity compared to those from LL. Evenness was nearly equal for both lagoons, with very low values. The Berger Parker index was higher in LL, due to the high abundance of C. macdonaghi that represented 73% of the total number of helminths.
Infracommunity structure. Table V shows the results of the analysis of composition of the infracommunities. At site LL, 34% of the silversides were uninfected, whereas all silversides caught at SGL harbored at least one helminth species. In SGL, individual fish harbored between 1 and 6 (most frequently 4) helminth species, the maximum species richness (9) was not reached, and A. cf. mordax was the most abundant species in the infracommunities. In LL, individual fish harbored between 1 and 3 (most frequently 1) helminth species, the maximum richness (5) was not reached, and C. macdonaghi was the most abundant species in the infracommunities.
Host size (standard length) was positively and significantly correlated with helminth species richness, diversity and total abundance, according to Spearman's rank correlation coefficient, with the exception of diversity in SGL (Tab. VI). Helminth species richness, diversity and total abundance were not significantly correlated with sex of the silversides examined (Tab. VI).
The seasonal differences in mean values of richness, diversity and number of helminths per Tab. II. Prevalence (P), mean intensity (IM) and mean abundance (AM) of helminths in Odontesthes bonariensis from Salada Grande and Lacombe lagoons.

DISCUSSION
Some of the helminth species found are not habitual members of the helminth communities of O. bonariensis, but occur regularly in other freshwater fishes from other bodies of water in Buenos Aires Province. These include T. bonariensis, a common parasite of Cichlasoma facetum (Jenyns, 1842) (lunaschi, 1988); W. matercula, commonly found as parasite of Cnesterodon decemmaculatus (Jenyns, 1842) and Jenynsia lineata (Jenyns, 1842) (lunaschi & Drago, 1995), and Saccocoelioides sp., a genus with species reported from numerous freshwater fishes (sziDat, 1973;lunaschi, 1984, 2002martorelli, 1986). At the infracommunity level, the presence of these species does not contribute significantly to the evaluation of species  (Drago, 1997). Helminth assemblages in fish are generally depauperate and isolationist compared with those in endothermic vertebrates; factors such as ectothermy, low vagility and structural simplicity of the gut, among others, have been cited as possible reasons for these differences (KenneDy et al., 1986;KenneDy, 1990). In addition, the freshwater habitats are essentially isolated systems, in which the fish may move more or less freely within them, but natural movements of fish and their parasites between systems necessitates overcoming barriers are thus more limited and infrequent (KenneDy et al., 1991).
Richness and diversity of helminths of O. bonariensis were higher in SGL, despite the smaller sample size in this lagoon. This could be related with several factors: (1) the larger size of SGL and of the silversides collected in this lagoon (significant differences were found between the standard length of silversides from both lagoons; t=8.4; p <0.001). The influence of the size of aquatic environments has been discussed by several authors. According to the islandsize hypothesis (island size being defined as the area of an aquatic environment, number of hosts within a population, or the size of an individual host) large islands should contain more parasitic species (holmes & Price, 1986). Nevertheless, conflicting results have been obtained from this approach (marcogliese & cone, 1991). Some authors have argued that larger areas may sustain larger host populations and thus may favour the existence of higher parasite diversity (KenneDy, 1978). Moreover, localities with high diversity of host species may be more favourable for parasites with complex life cycles, because of the availability of more definitive host species in which the parasites can achieve full development (guégan et al., 2005); (2) greater abundance of mollusks and fish-eating birds in SGL, which increases the chances of allogenic species such as, A. (P.) cf. diminuta and Ascocotyle sp.; (3) the greater variety of food items found in the gut of silversides captured in SGL; which increases the chances of autogenic species such as T. bonariensis and Saccocoelioides sp., that were found only in hosts that had gastropods [Heleobia parchappei (d'Orbigny, 1835)] in the gut. The life cycles described for other derogenids and haploporids indicate that fishes become parasitized when they eat infected snails (martorelli, 1986, 1989). Drago The helminth community of silversides is dominated by the autogenic intermediate specialist species C. macdonaghi at the component and infracommunity levels in LL; while it is dominated by the allogenic generalist species A. cf. mordax in SGL. Is interesting to note that in LL the dominant species was too the most prevalent, but in SGL the dominant species was not the most prevalent. The specificity of dominant species has been studied in several communities of freshwater fishes; still, no clear pattern has been established because the dominant species may be specialist, generalist, or change across geographic regions for the same host species (KenneDy, 1990(KenneDy, , 1995(KenneDy, , 1997sures et al., 1999;KenneDy & hartVigsen, 2000). The allogenic-autogenic condition of dominant species has been analyzed in several opportunities, and in spite of the supposed limitations of autogenic life-style, these species can dominate helminth communities (esch et al., 1988, lynDon & KenneDy, 2001. The differences regarding dominant species in both lagoons suggest that specificity and colonization strategy of silverside parasites do not represent important factors for community dominance. The dominance of the allogenic species A. cf. mordax in SGL could be attributed to the greater body size of silversides caught in this lagoon. Moreover, the standard length of silversides from both lagoons has positive correlation with the number of metacercariae of this species (r s =0.8, p < 0.001 in SGL; r s =0.57, p < 0.001 in LL).   In some host-parasite systems, community structure may be influenced by host size through changes in diet or volume of ingested food, ontogenetic changes in immunocompetency and changes in the probability of contact with intermediate hosts (esch et al., 1990). The higher richness and number of helminths in the infracommunities from larger silversides could be related with the higher variety of food items consumed by these fishes, particularly gastropods and their distribution in the lagoon (larger fish inhabit deep waters while smaller individuals inhabit the coastal region). Moreover, the increased abundance of larval endoparasites in larger fishes can be attributed to the occurrence of cumulative infection processes (isaac et al., 2000;guiDelli et al., 2003). These processes can explain the increasing number of helminths throughout the life cycle of silversides, especially for larval stages (A. cf. mordax, A. (P.) cf. diminuta, Ascocotyle sp., Hysterothylacium sp. and Contracaecum sp.).