Table 1.
Summary of all disease systems and subsets by pathogen type.
Table 2.
Summary of all disease systems and subsets by reservoir type.
Table 3.
Summary of all disease systems and subsets by reservoir category.
Table 4.
Summary of disease system reservoirs by major animal taxa.
Table 5.
Summary of mammalian reservoirs by orders represented.
Fig 1.
Expected versus observed representations of mammalian orders among reservoirs.
The expected representations (gray bars) assume that orders are found among reservoir mammals in the same proportion as they are among all mammals (n = 5,416) listed by Wilson and Reeder [22]. Observed bars (black) show the number of unique mammalian species in each order that were identified as reservoir hosts in this study. (A) shows the results for all systems (n = 330) in which 178 mammalian species were found to be reservoir hosts. (B) shows results for the human target subset of systems (n = 261) in which 155 mammalian species were found to be reservoir hosts.
Fig 2.
Kernel density plots comparing mean mass-corrected residual trait values for all mammalian species versus mammalian reservoir species.
We conducted imputation tests using mass-corrected residual trait values that were generated by regressing six key life history traits on adult body mass (g). To generate an expected distribution of mean residual trait values, we randomly generated 1,000 sets of mammalian species that had the same taxonomic representation (at the order level) as the unique mammalian reservoir species we identified. By calculating the mean residual trait values for these random sets of species, we were able to generate the mean mass-corrected residual trait distributions displayed in grey. The 2.5th and 97.5th percentiles of each distribution are represented with dashed vertical grey lines. For comparison, the observed mean mass-corrected residual trait value of the identified mammalian reservoir species is shown as a vertical black line.
Table 6.
Loadings of mass-corrected life history trait residuals on the first two principal component axes.
Fig 3.
Principal component scores for mammalian species.
We conducted principal components analysis using mass-corrected residual values for six key life history traits: gestation length (days), litter size, neonate body mass (g), interbirth interval (days), weaning age (days), and age at sexual maturity (days). Plotted in black are the first two PC scores for the 178 unique mammalian species we identified as reservoirs. Scores for all other mammal species are shown in grey.
Fig 4.
Kernel density plots comparing mean principal component scores for all mammalian species versus mammalian reservoir species.
PCA was conducted on six key residual life history traits: gestation length (days), litter size, neonate body mass (g), interbirth interval (days), weaning age (days), and age at sexual maturity (days). We then used scores from the first two PC axes as variables of interest in imputation tests. To generate an expected distribution of mean PC scores, we randomly generated 1,000 sets of mammalian species that had the same taxonomic representation (at the order level) as the unique mammalian reservoir species we identified. By calculating the mean PC scores for these random sets of species, we were able to generate the mean PC score distributions displayed in grey. The 2.5th and 97.5th percentiles of each distribution are represented with dashed vertical grey lines. For comparison, the observed mean PC score of the identified mammalian reservoir species is shown as a vertical black line.