Despite high scientific interest in the ecological and evolutionary drivers of mammal-virus associations, contrasting evidence exists on the characteristics that make wild mammals more subject to acquiring and transmitting viruses. Once research effort and phylogeny are accounted for, our results show that carrier status is best predicted by species body mass, longevity, and interbirth interval. Our models predicted large-sized mammals on both extreme ends of the fast-slow continuum of life-history as being more likely to carry viruses. These results suggest that although mammals on opposite ends of the fast-slow continuum have likely evolved diverging immunological strategies, viruses are able to exploit both strategies for replication and spread. This means the peculiarities of each viral group drive patterns of “carrier preference” reflecting viruses’ replication speed, adaptability and possibly evolutionary history.
Mammalian species with larger body sizes were more likely to be carriers of both general and zoonotic viruses. Usually, large-bodied species have higher dispersal abilities and need large home ranges in order to meet their high metabolic requirements21. The ability to move farther may play a role in increasing exposure to viral infections, as these species are more likely to come into contact with disease hosts (e.g., primary, secondary, or reservoirs) and vectors. Past works have found mammals’ body mass to be related to susceptibility to tick-borne encephalitis virus, yellow fever virus, and Zika virus22, suggesting that large-bodied species might significantly contribute to vector-borne viruses’ transmission. Moreover, home range was a key predictor of wildlife hosts of Rift Valley fever virus23, in agreement with the hypothesis that larger home ranges may enhance exposure to viruses by increasing inter and intraspecific contacts. Large-sized mammals consume large amount of food to meet high energetic requirements24, which makes them more likely to ingest viruses that rely on indirect routes of transmission via fomites (such as contaminated food sources).
In addition to the aforementioned mechanisms, larger body mass of zoonotic carriers may be explained by dynamics that underlie zoonotic viral spillover from wildlife. Humans interact more frequently with large wild mammals through several forms of direct exploitation, such as hunting and wildlife trade25, and hunters typically have a preference towards large-bodied target species26–28. Practices such as wildlife hunting and trade may pose a public health risk because of the often-lacking sanitary controls and inadequate animal manipulation29, which may facilitate viral spillover to people at different stages of the supply chain. Large mammals may also contribute to the emergence of zoonotic viral diseases as amplification hosts (i.e., organisms in which the infectious agent can replicate rapidly and reach high concentrations)30. It is possible that large mammals can develop higher viral loads and release greater quantities of virus, increasing the probability of transmission in case of encounter with humans or other species and acting as amplification hosts17. Traditionally, livestock animals are considered to be amplification hosts for zoonotic viruses31,32, but evidence of wild amplifiers has been found for Old World monkeys as chikungunya virus hosts in Senegal33, great apes and forest antelopes as Zaire Ebola virus hosts in West Africa34, and white-tailed deer as Cache Valley virus hosts in the United States35.
In both general and zoonotic models, mammals placed on the slower end of the life-history continuum (longer longevity and interbirth intervals) were more likely to have carrier status for relatively slow-evolving viral groups associated with chronic or persistent infection, such as RNA retroviruses (phylum Artverviricota) and dsDNA viruses (phyla Peploviricota, Baltimore classes VII and I). To ensure transmission, most dsDNA viruses and retroviruses are required to remain in the host for extended periods of time, resulting in persistent or chronic infections with low or delayed disease severity and virulence36. Because of the extended infectious period, viruses that cause chronic or persistent infection (e.g., Herpesvirus, Lentivirus) are expected to have a greater viral fitness (i.e., the capacity of a virus to produce infectious progeny) in long-lived hosts15. Our results sustain this hypothesis, suggesting that viruses which stay in the host for longer got a selective advantage by co-evolving with (and adapting to) slow-lived species which may carry the infection for prolonged periods of time and increase the probability of transmission.
For general and zoonotic carriers of viruses with linear genomic segments and an envelope (i.e., those in our 6th ecological group, such as Arenaviridae, Hantaviridae, and Peribunyaviridae), carrier status probability was highest for species with faster life-history as represented by short longevity and interbirth interval. These RNA viruses, which include Lassa fever virus and Sin Nombre virus, are considered relatively fast-evolving due to rapid nucleotide substitution rate and ability to reassort homologous genome segments36,37. Our findings suggest that such evolutionary mechanisms may be furtherly favoured in fast reproducing carriers, where population density and generation overlap may increase viral fitness due to the large availability of naïve individuals. Still, we cannot fully exclude that these results were driven by taxa representation in the training set, as bats, moles, shrews and especially rodents—that are considered the main natural hosts of these viruses—display shorter lifespans and faster life-histories. To further explore the adaptive links between viral evolvability and host traits it will be necessary to address critical data gaps on host-virus interactions which currently limit our ability to make inference.
It is important to consider that our study has limitations, associated with the dataset we used. We defined mammal-virus associations as the detection of a virus in a given mammalian species via serology, PCR, or isolation. None of these detection methods necessarily implies the species is an actual viral reservoir, just that the species is susceptible to infection. Thus, it is likely that we included different levels of susceptibility within the term “carrier”, such as: reservoir, natural host, and incidental host. Additionally, we recognise that sampling bias is present in the data, as the observed patterns of viral distribution across host taxa are not an accurate representation of the (largely unknown) mammalian virome38. This limitation affects any other similar study investigating general host-pathogen associations from a limited number of observations. By working above the viral species level using dichotomous outcomes (carrier status), and accounting for species’ virus-related citations (i.e., reporting effort), we mitigated the impact of such bias.
In this work, we identified common patterns along the fast-slow continuum of mammalian viral carriers, highlighting that mammals’ associations with viruses are generalizable only to a certain extent. The eco-evolutionary peculiarities of each viral group drive patterns of compatibility between mammals and viruses, as we partially captured by separately assessing different eco-evolutionary groups of viruses. By including multiple sides of viral diversity, we were able to pick up on trends that went undetected in previous studies where viral richness was analysed as a whole4. We found that larger mammals are more likely to be viral carrier, but both fast- and slow-living species could be exploited by viruses. General and zoonotic carriers’ profiles did not differ substantially in terms of life-histories, suggesting that there are no traits that disproportionately affect susceptibility to zoonotic viruses. Our analysis presents an updated conceptual framework for research at the interface of ecology and health, as underestimating viruses’ functional diversity may lead to neglecting potentially important sources of zoonoses, with the risk of impairing hazard assessments and outbreak preparedness. Using the insights provided by life-history theory may enhance surveillance strategies in areas where species with a higher proneness for carrying viruses are more abundant.