Potential Parasite Transmission in Multi-Host Networks Based on Parasite Sharing

Epidemiological networks are commonly used to explore dynamics of parasite transmission among individuals in a population of a given host species. However, many parasites infect multiple host species, and thus multi-host networks may offer a better framework for investigating parasite dynamics. We investigated the factors that influence parasite sharing – and thus potential transmission pathways – among rodent hosts in Southeast Asia. We focused on differences between networks of a single host species and networks that involve multiple host species. In host-parasite networks, modularity (the extent to which the network is divided into subgroups of rodents that interact with similar parasites) was higher in the multi-species than in the single-species networks. This suggests that phylogeny affects patterns of parasite sharing, which was confirmed in analyses showing that it predicted affiliation of individuals to modules. We then constructed “potential transmission networks” based on the host-parasite networks, in which edges depict the similarity between a pair of individuals in the parasites they share. The centrality of individuals in these networks differed between multi- and single-species networks, with species identity and individual characteristics influencing their position in the networks. Simulations further revealed that parasite dynamics differed between multi- and single-species networks. We conclude that multi-host networks based on parasite sharing can provide new insights into the potential for transmission among hosts in an ecological community. In addition, the factors that determine the nature of parasite sharing (i.e. structure of the host-parasite network) may impact transmission patterns.

traps, distanced 1 to 5 km from one other were set over four days. The traps were evenly distributed among four habitat types: forest (natural forest and tree plantations); non-flooded upland (shrub, orchards and upland agriculture); lowland flooded areas (rice paddies); and peridomestic locations (houses and immediate surrounding areas).
Helminths survey for each rodent was conducted following [1]. Briefly, trapped rodents were euthanized and dissected. The stomach, small intestine and large intestine were separated and examined for helminth infection under a stereo-microscope. The collected helminths (Table   S2) were preserved in 70% alcohol and identified according to general helminth identification keys as referenced in [1,3].
Across the three localities, the three multi-species networks had 27-40 individuals from 2-4 rodent species infected by 6-10 helminth taxa. The single-species networks had 5-23 individuals infected by 2-7 helminths. Helminth richness (number of helminth taxa infecting an individual rodent) ranged between 1 and 4. When averaged across individuals within each network, mean helminth richness ranged between 1 and 2.13. The prevalence of each helminth in each rodent species is indicated in Table S3.
We built a phylogenetic tree (Fig. S2) based on molecular data of the cytochrome b mitochondrial gene. We compiled cytochrome b sequences from the NCBI gene bank and used a maximum likelihood analysis with the GTR+G+I substitution model of molecular evolution with the aid of the function 'phymltest' in the R package 'ape' [4]. To ensure that our results were not affected by the way we constructed the tree, we re-ran analyzes with a tree from [5], but that did not include Mus cervicolor. The results were qualitatively the same. Table S2. Information on helminth taxa used in this study. Data are from [3,6]. Helminths in the table are gastrointestinal parasites transmitted via fecal-oral pathways. All helminths were identified to species level (four as unique morpho-species). We included one helminth with direct mode of transmission (Syphacia muris) because our preliminary work indicated that removing this helminths did not change our results and this helminths was very common.

Controlling for network size and connectance in the analysis of network modularity.
The value of the modularity function M may be affected by network size or connectance. In our case, in each locality, the multi-species network was larger than each of the single-species networks and its connectance was lower (see Table 1  in Buriram was not significantly modular (see Table 1 in main text).

Multiple regression on distance matrices
We give an additional description of the multiple regression on distance matrices (MRM) using a graphical visualization (Fig. S4). Cell values in the explanatory matrix E2 are differences between pairs of individuals in a continuous characteristic (e.g. patristic distances or body mass). This is represented by the shade of brown (stronger the shade the larger the difference). E2 thus represents a categorical variable.

Construction of sub-TPNs.
Our goal was to compare TGI between a multi-species TPN and a single-species TPN (with >10 individuals) within the same locality. It is inappropriate, however, to compare networks of different sizes and connectance (i.e. the number of realized interactions divided by the number of possible ones). To control for different size and connectance while comparing multi-species TPNs to their respective single-species TPNs (within the same locality) we built 250 multi- Rattus tanezumi for 38%. These proportions were kept at each of the 250 sub-networks.
We also kept the connectance of the sub-TPNs constant to that of the original TPN. The connectance of the single-species TPN was always higher than that of the multi-species TPN ( Table 1 in main text). Therefore, we built 250 single-species sub-TPNs by randomly removing edges from the original one to adjust for the connectance of the original multi-species network.
The result was a set of 250 multi-species sub-TPNs and a set of 250 single-species sub-TPNs of equal size and connectance. For each of these 500 sub-TPNs we generated a distribution of 250 TGI values by randomly selecting individuals as starting points. We used the distribution of 250 mean TGI values (averaged for each sub-TPN) to examine differences between the single-and multi-species TPNs.