Discussion
Results suggest that in this fish-Gyrodactylus system, the presence of additional species leads to a dilution effect, but only in the presence of a low competent host. A significant decrease in mean abundance of G. kobayashii on goldfish was detected in silver crucian carp group, a low competent host, but not significantly in the presence of non-competent hosts. However, no dilution effect was observed in the four-species treatment that included silver crucian carp.
Species with low competence play key roles in the dilution effect (Thieltges et al. 2008). While dilution effect occurred commonly in communities with higher diversity (Ostfeld & Keesing 2000; Johnson & Thieltges 2010; Keesing & Ostfeld 2021a), the important determinant of parasite transmission was the specific composition of the host community which affected the strength of the dilution effect (Ostfeld & Keesing 2000; Logiudice et al. 2008; Johnson et al. 2013; Salkeld et al. 2013). Variation in host quality is a condition necessary for a reduction in pathogen transmission (Keesing & Ostfeld 2021a). In the Lyme disease system, low-quality hosts in high-diversity communities deflected ticks away from the most competent hosts (Keesinget al. 2009; Keesing et al. 2010). The presence of low competence treefrogs (Hyla versicolor ) reduced Ribeiroia ondatrae abundance in toads (Bufo americanus ) through encounter reduction (Johnson et al. 2008). In a field investigation, the presence of the invasive round goby (Neogobius melanostomus ), itself infected by the eyefluke Diplostomum spp. at low infection levels, sharply reduced the prevalence and abundance of the eyeflukes in native fishes from the St Lawrence River, Canada (Gendron & Marcogliese 2017). In our study, a dilution effect was not observed in the community with the highest species richness, but only in that with sufficient numbers and/or density of silver crucian carp, a low-competence host. Therefore, results further confirmed that the specific composition in a community, including species with low competence, was a necessary condition leading to the dilution effect, rather than simply diversity. However, abundance and/or density of the low competent host also appears important, as the mix-species assemblage also contained silver crucian carp, but at lower abundance and density (Thieltges et al. 2008). The relative abundance of low-competence hosts compared to competent hosts is an important component for the dilution effect (Ostfeld & Keesing 2012). Curiously, a higher density of all potential hosts (competent plus low competence) did not enhance parasite transmission, despite the fact that density of primary host play a roles in transmission of gyrodactylids (Johnson et al. 2011; Jin et al. 2022).
Incompetent hosts (decoys) can interfere with parasite transmission and distract free-living parasite stages away from suitable hosts, leading to wasted transmission events (Johnson & Thieltges 2010). Abundance of trematode (Himasthla elongata ) in cockle hosts (Cerastoderma edule ) was significantly reduced by the presence of incompetent hosts (Crepidula fornicata , Mya arenaria andCrassostrea gigas ) that acted as decoy hosts for cercariae (Thieltges et al. 2008). Interestingly, the presence of incompetent hosts did not result in a dilution effect in our experiments, unlike other systems (Johnson & Thieltges 2010). Decoy hosts may not divert infective parasites away from goldfish in this system, likely because the parasites are host specialists. This lends support to the idea that the dilution effect is weaker with host specialists, with little or no wasted transmission (Ostfeld & Keesing 2012).
Several mechanisms have been proposed for how biodiversity reduces disease risk, such as by reducing encounters between hosts and parasites, reducing establishment after encounters occurred, or decreasing numbers of susceptible hosts (Keesing et al. 2006). Addition of low competence treefrogs significantly reduced infection ofR. ondatrae in toads by limiting transmission of the cercariae to toads likely via a strong immune defense against them in treefrogs, a form of encounter reduction (Johnson et al. 2008; Johnson & Hartson 2009). In our experimental system, transmission to goldfish likely was reduced by the presence of sufficient numbers of low competent hosts via encounter reduction, as the infective gyrodactylids may have been deflected away from the goldfish to the crucian carp, where they then only established in low numbers. A dilution effect was observed not only at the population level on the goldfish hosts (population level) but also at the host community level on the goldfish and crucian carp combined (Buss & Hua 2023).
The presence of low competence or incompetent hosts may influence the behavior and activity of a competent host (Johnson & Thieltges 2010; Keesing et al. 2010). For pathogens transmitted via host contact, theoretical models supported the DEH under the condition that additional host species mixed well with the focal hosts (Keesing et al.2006; Keesing & Ostfeld 2021b), promoting encounter reduction between infected and susceptible hosts. In the present study, observations of schooling behavior showed that the silver crucian carp mixed well with the goldfish. Thus, the low competent host was in close proximity to goldfish, potentially attracting infective gyrodactylids away from susceptible goldfish, leading to encounter reduction. Furthermore, the average adjacency distance among goldfish significantly increased, and the cumulative contact time significantly decreased following the addition of the silver crucian carp. The higher adjacency distance and lower contact time may contribute to a reduction in transmission of gyrodactylids between goldfish. However, changes in schooling behavior alone are insufficient to cause a dilution effect, as they were also observed in the presence of non-competent hosts. This may in part be due to the fact that the incompetent hosts did not mix with goldfish. The lack of contact between the host species may have also served to prevent wasted transmission events. Host behavior has been shown to be important in the transmission of other gyrodactylids. The frequency of social contacts through shoaling behavior in the guppy determined the outbreak of epidemics in G. turnbulli (Johnson et al. 2011). Indeed, more than one mechanism contributing to a dilution effect may be at play at the same time in a particular system (Johnson & Thieltges 2010).
Although no dilution effect was detected in the presence of the two incompetent hosts, the higher mean abundance on goldfish in the grass carp group may result from reinfection of detached gyrodactylids in the water column due to spatial distribution of goldfish in the whole tank (Jin et al. 2022).
Our experiment demonstrated that the dilution effect depends more on specific community composition than biodiversity in this host-parasite system. In this system, low competent hosts in sufficient number or density are required for a dilution effect. The primary mechanism in this case appears to be encounter reduction. Changes in host behavior in the presence of a low-competence host also may contribute to a dilution effect. In this host-parasite system, the presence of incompetent hosts did not lead to a dilution effect, supporting the idea that the dilution effect may not be as strong with host specialists. The occurrence of a dilution effect in the presence of the silver crucian carp schooling with the goldfish, but the absence of a dilution effect in the presence of additional non-competent hosts that segregated from the goldfish suggested that host schooling behavior plays a key role in restricting the dilution effect.