We applied a novel framework based on network theory and a concept of modularity that estimates congruence between trait-based ( = functional) co-occurrence networks, thus allowing the inference of co-occurrence patterns and the determination of the predominant mechanism of community assembly. The aim was to investigate the relationships between species co-occurrence and trait similarity in flea communities at various scales (compound communities: across regions within a biogeographic realm or across sampling sites within a geographic region; component communities: across sampling sites within a geographic region; and infracommunities: within a sampling site). We found that compound communities within biogeographic realms were assembled via environmental or host-associated filtering. In contrast, functional and spatial/host-associated co-occurrence networks, at the scale of regional compound communities, mostly indicated either stochastic processes or the lack of dominance of any deterministic process. Analyses of congruence between functional and either spatial (for component communities) or host-associated (for infracommunities) co-occurrence networks demonstrated that assembly rules in these communities varied among host species. In component communities, stochastic processes prevailed, whereas environmental filtering was indicated in 4 and limiting similarity/competition in 9 of 31 communities. Limiting similarity/competition processes dominated in infracommunities, followed by stochastic mechanisms. We conclude that assembly processes in parasite communities are scale-dependent, with different mechanisms acting at different scales.

We studied the relationships between functional alpha and beta diversities of fleas and their small mammalian hosts in 4 biogeographic realms (the Afrotropics, the Nearctic, the Neotropics and the Palearctic), considering 3 components of alpha diversity (functional richness, divergence and regularity). We asked whether (a) flea alpha and beta diversities are driven by host alpha and beta diversities; (b) the variation in the off-host environment affects variation in flea alpha and beta diversities; and (c) the pattern of the relationship between flea and host alpha or beta diversities differs between geographic realms. We analysed alpha diversity using modified phylogenetic generalized least squares and beta diversity using modified phylogenetic generalized dissimilarity modelling. In all realms, flea functional richness and regularity increased with an increase in host functional richness and regularity, respectively, whereas flea functional divergence correlated positively with host functional divergence in the Nearctic only. Environmental effects on the components of flea alpha diversity were found only in the Holarctic realms. Host functional beta diversity was invariantly the best predictor of flea functional beta diversity in all realms, whereas the effects of environmental variables on flea functional beta diversity were much weaker and differed between realms. We conclude that flea functional diversity is mostly driven by host functional diversity, whereas the environmental effects on flea functional diversity vary (a) geographically and (b) between components of functional alpha diversity.

We investigated phylogenetic patterns in flea assemblages from 80 regions in 6 biogeographic realms and asked whether (a) flea phylogenetic turnover is driven by host phylogenetic turnover, environmental dissimilarity or geographic distance; (b) the relative importance of these drivers differs between realms; and (c) the environmental drivers of flea phylogenetic turnover are similar to those of host phylogenetic turnover. We also asked whether the phylogenetic originality of a flea species correlates with the degree of its host specificity and whether the phylogenetic originality of a host species correlates with the diversity of its flea assemblages. We found that host phylogenetic turnover was the best predictor of flea phylogenetic turnover in all realms, whereas the effect of the environment was weaker. Environmental predictors of flea phylogenetic turnover differed between realms. The importance of spatial distances as a predictor of the phylogenetic dissimilarity between regional assemblages varied between realms. The responses of host turnover differed from those of fleas. In 4 of the 6 realms, geographic distances were substantially better predictors of host phylogenetic turnover than environmental gradients. We also found no general relationship between flea phylogenetic originality and its host specificity in terms of either host species richness or host phylogenetic diversity. We conclude that flea phylogenetic turnover is determined mainly by the phylogenetic turnover of their hosts rather than by environmental gradients. Phylogenetic patterns in fleas are manifested at the level of regional assemblages rather than at the level of individual species.

We used data on the species composition of regional assemblages of fleas and their small mammalian hosts from 6 biogeographic realms and applied a novel method of step-down factor analyses (SDFA) and cluster analyses to identify biogeographic (across the entire globe) and ecological (within a realm across the main terrestrial biomes) clusters of these assemblages. We found that, at the global scale, the clusters of regional assemblage loadings on SDFA axes reflected well the assemblage distribution, according to the biogeographic realms to which they belong. At the global scale, the cluster topology, corresponding to the biogeographic realms, was similar between flea and host assemblages, but the topology of subtrees within realm-specific clusters substantially differed between fleas and hosts. At the scale of biogeographic realms, the distribution of regional flea and host assemblages did not correspond to the predominant biome types. Assemblages with similar loadings on SDFA axes were often situated in different biomes and vice versa. The across-biome, within-realm distributions of flea vs host assemblages suggested weak congruence between these distributions. Our results indicate that dispersal is a predominant mechanism of flea and host community assembly across large regions.

We revisited the role of dissimilarity of host assemblages in shaping dissimilarity of flea assemblages using a non-linear approach. Generalized dissimilarity models (GDMs) were applied using data from regional surveys of fleas parasitic on small mammals in four biogeographical realms. We compared (1) model fit, (2) the relative effects of host compositional and phylogenetic turnover and geographic distance on flea compositional and phylogenetic turnover, and (3) the rate of flea turnover along gradients of host turnover and geographic distance with those from earlier application of a linear approach. GDMs outperformed linear models in explaining variation in flea species turnover and host dissimilarity was the best predictor of flea dissimilarity, irrespective of scale. The shape of the relationships between flea compositional turnovers along host compositional turnover was similar in all realms, whereas turnover along geographic distance differed among realms. In contrast, the rate of flea phylogenetic turnover along gradients of host phylogenetic turnover differed among realms, whereas flea phylogenetic turnover did not depend on geographic distance in any realm. We demonstrated that a non-linear approach (a) explained spatial variation in parasite community composition better than and (b) revealed patterns that were obscured by earlier linear analyses.