Viral zoonotic risk is homogenous among taxonomic orders of mammalian and avian reservoir hosts

Abstract

The notion that certain animal groups disproportionately maintain and transmit viruses to humans due to broad-scale differences in ecology, life history, and physiology currently influences global health surveillance and research in disease ecology, virology, and immunology. To directly test whether such "special reservoirs" of zoonoses exist, we used literature searches to construct the largest existing dataset of virus-reservoir relationships, consisting of the avian and mammalian reservoir hosts of 415 RNA and DNA viruses along with their histories of human infection. Reservoir host effects on the propensity of viruses to have been reported as infecting humans were rare and when present were restricted to one or two viral families. The data instead support a largely host-neutral explanation for the distribution of human-infecting viruses across the animal orders studied. After controlling for higher baseline viral richness in mammals versus birds, the observed number of zoonoses per animal order increased as a function of their species richness. Animal orders of established importance as zoonotic reservoirs including bats and rodents were unexceptional, maintaining numbers of zoonoses that closely matched expectations for mammalian groups of their size. Our findings show that variation in the frequency of zoonoses among animal orders can be explained without invoking special ecological or immunological relationships between hosts and viruses, pointing to a need to reconsider current approaches aimed at finding and predicting novel zoonoses.

Keywords: generalized additive model; infectious disease; reservoir; surveillance.

Fig. 1.

Species richness and diversity of viruses associated with major reservoir host groups. (A) The distribution of virus families across mammalian and avian reservoir orders. Each rectangle represents a reservoir–virus family combination, with size corresponding to the number of virus species linked to that reservoir and color indicating the proportion of these viruses which are zoonotic. Viral families are abbreviated as follows: Ade = Adenoviridae, Are = Arenaviridae, Art = Arteriviridae, Asf = Asfarviridae, Ast = Astroviridae, Cal = Caliciviridae, Cir = Circoviridae, Cor = Coronaviridae, Fil = Filoviridae, Fla = Flaviviridae, Hepa = Hepadnaviridae, Hepe = Hepeviridae, Her = Herpesviridae, Nai = Nairoviridae, Ort = Orthomyxoviridae, Pap = Papillomaviridae, Para = Paramyxoviridae, Parv = Parvoviridae, Per = Peribunyaviridae, Phe = Phenuiviridae, Pic = Picornaviridae, Pne = Pneumoviridae, Pol = Polyomaviridae, Pox = Poxviridae, Reo = Reoviridae, Ret = Retroviridae, Rha = Rhabdoviridae, Tob = Tobaniviridae, and Tog = Togaviridae. (B) The taxonomic diversity of viruses maintained by each reservoir, at decreasing levels of sensitivity to rare lineages as the q-parameter increases. (C) The number and proportion of virus species associated with each reservoir which are zoonotic; error bars show 95% binomial CIs calculated using the Wilson method (31).

Fig. 2.

Reservoir host and virus predictors of zoonotic propensity. (A) Top 15 models ranked by AIC, along with the top models not containing a virus-family specific effect (ranked 23rd to 28th), and the top models not containing an effect for virus publication count (ranked 177th to 182nd). Rows represent individual models and columns represent variables. Cells are shaded according to the proportion of deviance explained by each effect; effects not present in particular models are indicated in white. The final three columns represent different versions of a potential “special reservoir effect” and were not allowed to cooccur in the same model. (B–F) Effects present in the top model. Lines indicate the predicted effect of each variable, when keeping all other variables at either their median observed value (when numeric) or their most common value (when categorical). Shaded regions indicate the 95% CIs of predictions, while points indicate partial residuals after accounting for all variables in the model except the one on the x axis. Effects whose 95% CI cross zero over the entire range of the predictor variable are shaded in gray. Phylogenetic distance (E) was measured as cophenetic distances, which describe the total evolutionary distance from each group to primates. Note that only the subset of virus families which include significant effects (i.e., those showing no overlap with 0) are illustrated in F (see SI Appendix, Fig. S2 for all families).

Fig. 3.

Relationship between the number of virus species and the number of zoonotic species maintained by each reservoir group. The line shows a linear regression fit, with its 95% CI indicated by the shading.

Fig. 4.

Factors predicting the number of zoonotic virus species across animal orders. (A) Models for all possible variable combinations ranked by AIC. Each row represents a model, while columns represent variables. Filled cells and white cells indicate variable inclusion and absence, respectively. The top four model are color-coded, with colors reused in all other panels to identify the respective models. (B) Coefficient estimates for the top four models; points indicate the maximum likelihood estimate and lines show 95% CIs. All variables were scaled by dividing them by 2 times their SD, meaning coefficients are directly comparable as effect sizes. (C and D) Partial effect plots for variables in the top model. Lines and shading indicate the partial effects and 95% CIs, with points showing partial residuals. (E) Predicted number of zoonotic viruses for each reservoir group when using the top model (blue in A; see SI Appendix, Fig. S5 for other top models).