Network link prediction: Ecological factors shape community structure
Despite the uncertainty in the host repertoires, network link prediction (NLP) models confirmed the influence of the hosts’ evolutionary history on fish-monogenean community structure. The host phylogeny contributed considerably to the performance of the plug-and-play algorithm. However, host-parasite links appear to be mostly predicted by ecological parameters as the basin/basin type–parameter (Table 1) contributed the most (Fig. 5c). Therefore, ecological opportunity might play a major role in the assembly of cichlid-Cichlidogyrus communities similar to neotropical teleost-monogenean communities (Braga et al.2014), and these opportunities are predicted by host presence in rivers and lakes.
The uncovered significance of opportunity is highly relevant for aquaculture and fish conservation efforts. This study is the first to quantify host-pathogen interactions in tilapias, Nile tilapia (Oreochromis niloticus L.) being one of the most widely farmed fish worldwide (FAO 2019). Introductions of infectious diseases can have devastating effects on native ecosystems (Thompson 2013). Concerning tilapia, co-introductions of the tilapia-lake virus have caused significant economic losses (Eyngor et al. 2014; Fathi et al. 2017). Moreover, introductions of tilapias have led to co-introductions of their monogenean parasites in continental Africa (Jorissen et al. 2020), Madagascar (Šimková et al. 2019), Asia (Paperna 1960; Duncan 1973; Wu et al. 2006), Australia (Wilson et al. 2019), and the Americas (Jiménez-García et al. 2001; Azevedo et al. 2006), with occasional spillover to native fishes (Jiménez-García et al. 2001; Šimková et al.2019), albeit with little changes to the respective meta-community structures (Fig. 3). Our results suggest that anthropogenic introductions might promote further host switches in the future. In this context, network predictions could present key tools for understanding and possibly minimising the risk of emerging diseases (Albery et al. 2021).
Our results underline that NLP can verify traditional statistical analyses and provide further insight into ecological and evolutionary mechanisms shaping host-parasite interactions. For instance, we inferred that life style, trophic level, and host size are among the more informative predictors of cichlid-Cichlidogyrus interactions whereas parasite phylogenetic relationships and morphological parameters mostly failed to improved model performance (Fig. 5b). Therefore, host switches might more likely occur between ecologically similar hosts and emerging diseases in aquaculture could be avoided through culturing native fishes (Ju et al. 2020; Nobile et al. 2020). Previous studies on fish parasites have delivered inconclusive results for the role of host and parasite traits on host-parasite community composition. No studies investigated the effects of life style as coded here (Table 1), but host habitat preference can affect parasite communities (Locke et al. 2013). Parasite community composition correlated with the host trophic level in some cases, e.g. for shelf fish off Buenos Aires (Timi et al. 2011), but not in others, e.g. for freshwater fish in Canada (Locke et al. 2013) and marine fish in Finland (Locke et al. 2014). Host sizewas suggested as important predictor for the community composition of ectoparasitic monogeneans (Guégan et al. 1992; Sasal & Morand 1998; Sasal et al. 1999; Šimková et al. 2001; Desdeviseset al. 2002; Morand et al. 2002). However, these correlations might reflect phylogenetic patterns of host size (Poulin 2002) explaining the variable importance of host size here. Lastly, no correlation of attachment or reproductive organ morphologywith community composition was found for species of Cichlidogyrusunlike for species of Dactylogyrus (Šimková et al. 2001; Jarkovský et al. 2004). Instead, the morphology mostly reflects phylogenetic relationships of the parasites (Vignon et al. 2011; Cruz-Laufer et al. 2021b). The results of these studies highlight the challenge of linking host and parasite traits with community composition parameters and generalising observed patterns as sampling biases (Fründ et al. 2016) (Fig. 5a) and character coding (Pavoine et al. 2009) influence the results. NLP provides an accessible path to start uncovering the role of various parameters (Fig. 5c) and predicting undetected interactions (Fig. 5b).