DISCUSSION

The intestinal helminth parasite community of the stocked eel population in Neusiedler See showed characteristics similar to those in natural eel populations in rivers of the UK and mainland Europe (Kennedy, 1993; Kennedy et al. 1997, 1998; Sures et al. 1999; Sures & Streit, 2001; Kennedy & Moriarty, 2002). The component community consisted of 5 helminth species, all of these species have previously been reported from eels in Europe. This number of species is comparable with data on the intestinal parasite fauna reported for eels from rivers in Belgium (Schabuss et al. 1997), Germany (Sures et al. 1999; Sures & Streit, 2001), Ireland (Kennedy & Moriarty, 2002), Italy (Kennedy et al. 1997), Portugal (Saraiva & Eiras, 1996) and the Netherlands (Borgsteede et al. 1999). Higher component community richness (9, 11, 8 and 8 species) was reported from England (Kennedy, 1993, 1997; Norton et al. 2003) and Italy (Kennedy et al. 1998), respectively.

The intestinal parasite community in both sampling sites in Neusiedler See was heavily dominated by acanthocephalans. Such domination of a community by an acanthocephalan is a common characteristic of intestinal helminth communities in eels and has been reported from the UK and mainland Europe (Kennedy, 1990; Schabuss et al. 1997; Kennedy et al. 1998; Sures et al. 1999; Kennedy & Hartvigsen, 2000; Sures & Streit, 2001; Kennedy & Moriarty, 2002).

Within the 8 years of study significant changes in the structure and diversity of the parasite community in Neusiedler See were observed. The parasite community in Illmitz was heavily dominated by A. lucii from 1994 until 1998 but in 2001 A. anguillae dominated the parasite community, after increasing continuously from 1995 onwards. The diversity indices increased and more eels were infected with more parasite species with each year of the study. In the southern sampling site, the community structure remained rather constant with a continuous domination of A. anguillae and a very low proportion of A. lucii-infected eels. The diversity characteristics increased with each year but to a lower extent than in Illmitz. Comparing the component community of eels in Neusiedler See with the investigation in Lough Derg, Ireland (Kennedy & Moriarty, 2002), it is obvious that the Shannon-Wiener index and Simpson's index in Neusiedler See exhibited higher values (maximum 2·07 and 4·02 respectively) and varied over a higher range than in Lough Derg (maximum values 0·30 and 1·15). The Berger Parker dominance index in Illmitz decreased from a maximum value of 0·95 to a minimum of 0·3 and the dominating species changed within the study period, whereas in Lough Derg the high domination of A. lucii ranging from 0·93 to 1·0 remained constant throughout the 18 years of study. Thus the diversity in Neusiedler See was higher and varied to a higher extent between years than in Lough Dergh, while the dominance of a single species was lower and decreased throughout the study. The diversity in Neusiedler See was more similar to diversity characteristics reported from component communities of the 13-year study on the river Clyst (Kennedy, 1993) and short-term studies from coastal lagoons in Italy (Kennedy et al. 1997), the river Rhine (Sures et al. 1999; Sures & Streit, 2001) and the rivers Thames and Test (Norton et al. 2003).

On the intestinal helminth infra-community level, a higher maximum number of species per eel (4 species), similar mean number of species and mean Brillouin's index values, and lower mean number of parasites per eel were recorded in Neusiedler See compared to results from Lough Derg. The number of eels infected with 0 or 1 parasite species decreased and the percentage of eels infected with 2 or more species increased in both sampling sites in Neusiedler See throughout the 8 years of study, whereas the number of eels uninfected or with a single-species infection remained at a rather constant high level in Lough Derg. Results of the investigations on the eel parasite community of the River Tiber (Kennedy et al. 1998) showed lower diversity characteristics of the infracommunity than in Neusiedler See.

This increase in richness and diversity is very surprising as the eel population was predicted to decline and become extinct within 10 years after the prohibition of eel stocking in 1993 (Herzig et al. 1994). The infection levels and the diversity of the parasite community did not decline as a reaction to the predicted changes in the host population. So the question arose whether the eel population was in decline at all. Recent results from fish ecological investigations of the reed belt showed that despite a slight decrease in the eel population between 1994 and 1997, the eel stock in 2002 was at a similar level as in the early 1990s (Wolfram et al. 2001; Wolfram & Mikschi, 2003). In contrast to the findings of this study, the results of the fish ecological investigations showed that the length frequency distribution of eel did not change from 1988 to 2002 and the average size of eels did not increase (Wolfram et al. 2001; Wolfram & Mikschi, 2003). There was no decrease of specimens in the smaller size classes and small eels (<15 cm) were caught regularly until 2002, indicating constant illegal stocking.

Corresponding to such ongoing illegal stocking measures probably by local fishermen, the host population did not decrease after 1993. Thus, the characteristics of the intestinal parasite community reflected the unchanged eel population and moreover we observed an increase in helminth parasite diversity over the last decade.

The most important factors associated with the infection parameters of the dominating acanthocephalans in the 2 sampling sites in Neusiedler See were the year of study, season of sampling and length of eels.

In this study, we found a clear relationship between the parasite infection parameters and host size on both a total length and size-class basis. The influence of host length on parasite infection levels has been reported from various studies where larger eels harboured more parasites than smaller ones (Dogiel, Petrushevski & Polyanski, 1958; Moravec, 1985; Conneely & McCarthy, 1986; Molnar, Szekely & Perenyi, 1994; Schabuss et al. 1997; Poulin, 2000; Lefebvre et al. 2002; Audenaert et al. 2003), but various factors like sample size and size range of hosts are also of major importance and not all parasite species exhibit a correlation between intensity of infection and host size (see Poulin (2000) for a critical review). This accumulation of parasites in larger specimens is explained by a longer exposure time of older (=larger) eels to the parasite, a larger host area (intestine size) for parasite establishment, a higher degree of consumption of infected prey with increasing host size and a differential efficiency of the parasite transmission between intermediate and paratenic hosts (Poulin, 2000; Lefebvre et al. 2002). Moravec (1985) assumed that the intermediate host Asellus aquaticus is the main source of A. lucii infection for smaller eels, but their proportion in the food of large eels gradually decreases with increasing proportion of forage fish. Gut analyses showed that eels in Neusiedler See mainly feed on aquatic insects (Odonata, Coleoptera, Ephemeroptera and Trichoptera representing more than 50% of the food items throughout the year) and on Asellus aquaticus especially in spring, whereas planktonic animals and fish are of minor importance (Herzig et al. 1994; Wolfram et al. 2001). Eels in Neusiedler See become piscivorous at a size larger than 50 cm. As there are few eels in Neusiedler See with a body length exceeding 60 cm due to high population density and intensive fishing (Herzig et al. 1994; Wolfram et al. 2001; Wolfram & Mikschi, 2003), Asellus aquaticus seems to be the main source of infection with A. lucii and A. anguillae in this lake and eels larger than 50 cm seem to continue to feed intensively on this isopod. In this study the host length also had significant influence on the mean number of parasites and mean number of parasite species per eel. Norton et al. (2003) and Conneely & McCarthy (1986) reported a significant influence of eel length on the number of parasite species per eel, but no correlation between eel size and number of parasites per eel has been found in their study. The higher number of helminth species and specimens in larger eels in Neusiedler See may indicate a consumption of an increasing amount and variety of intermediate hosts with increasing host size.

Within this study, significant differences in the infection levels of both acanthocephalan species between the seasons of sampling were observed. A. anguillae exhibited significantly higher abundance and mean intensity in spring in both sampling sites. A. lucii recorded in Illmitz showed significantly higher abundance and mean intensity in autumn, whereas in the southern sampling site this parasite species displayed the highest values of abundance and prevalence in spring. Moravec (1985), Conneely & McCarthy (1986) and Kennedy et al. (1998) reported seasonality of parasite infections in eels, but not for acanthocephalans. Kennedy & Moriarty (2002) reported higher abundance of A. lucii in May, July and August, but noted that the increase in abundance is very dependent upon the values found in a single month and alternative samples in the same month can produce very different patterns of dispersion. Kennedy & Moriarty (1987) recorded lowest levels of abundance and prevalence of A. lucii and A. anguillae in March and April with an increase to a peak in mid-summer, a decrease in July and an increase again in autumn. A variety of ecological factors seem to account for the seasonal patterns and between-site variations in eel parasite infections (Conneely & McCarthy, 1986). Eels exhibit marked seasonality in their activity patterns and feeding intensity and temperature-related changes in feeding habits rather than physiological synchronization of parasite and host life-cycles are more likely to be responsible for the increased parasitizing of eels in certain seasons (Conneely & McCarthy, 1986). The gut analyses of the eel in Neusiedler See showed that eels were feeding more intensively in April and June while the feeding intensity decreased in summer and autumn and Asellus aquaticus was consumed intensively especially in spring (Herzig et al. 1994). This feeding behaviour of eel in Neusiedler See could explain the higher infection rate for acanthocephalans and the higher number of helminth species and specimens in spring in the southern sampling area. In Illmitz, however, the maximum infection level of A. lucii was recorded in autumn and it seemed that this parasite utilized eel at a different time of the year than in the southern sampling site. Other ecological factors, like a different population cycle of A. aquaticus or different cycles in recruitment and reproduction of the parasite may influence the transmission of A. lucii in Illmitz.

The significant differences between the 2 sampling sites concerning the infection levels of A. lucii and the increasing infection of eel with A. anguillae in Illmitz are surprising. Both species use Asellus aquaticus as their intermediate host and the 2 sampling sites were in close proximity and were similar in terms of water quality, host size and invertebrate abundance (Wolfram et al. 2001). Neither acanthocephalan species are eel specialists: the preferred final host of A. lucii is the perch (Perca fluviatilis) and that of A. anguillae is the chub (Leuciscus cephalus). As the chub is absent in the Neusiedler See (Herzig et al. 1994; Wolfram et al. 2001), eel might be the preferred definitive host of A. anguillae as in Lough Derg (Kennedy & Moriarty, 1987, 2002).

Results of the fish ecological investigation of the reed belt showed that the fish community of the 2 sampling sites is different. Perch were more abundant in Illmitz than in the southern sampling area (Wolfram et al. 2001). Thus, the higher prevalence, mean intensity and abundance of A. lucii in Illmitz compared to the southern sampling site could be explained by the higher number of A. lucii introduced to the community by perch. Kennedy & Moriarty (2002) observed changes in the relative abundance of A. lucii and A. anguillae in Lough Derg and attributed these changes to differences in infection levels in different parts of the lake and to changes in the movement patterns of eels in the Lough. An increased movement of eels from areas with high abundance of A. anguillae, like in the southern sampling site, into the bay of Illmitz might be a cause for the increasing infection of eels with A. anguillae in this area.

Kennedy (1992) reported competition between A. lucii and A. anguillae in eels such that A. anguillae populations are depressed by its congener, whereas Kennedy & Moriarty (1987) could not detect competitive displacement or resource partitioning in space or time between these two acanthocephalans, but still considered that exploitation competition might occur. In this study it seems that in the southern sampling site, the proportions of the 2 species remained fairly constant over the period and both congeneric species of acanthocephalans co-exist in apparently stable equilibrium in eels. In Illmitz, however, the observed changing proportions and species ratios of A. lucii and A. anguillae and the increasing number of eels infected with both species, seem to reflect the differences in the fish community and eel movements into this sampling site rather than competition as A. lucii is reported to be the stronger competitor (Kennedy, 1992). However, resource partitioning in time might be possible in this sampling site, as indicated by the maximum infection levels of A. lucii in autumn and A. anguilla in spring.