The majority of human emerging infectious diseases (EIDs) are zoonotic, with viruses originating in wild mammals of particular concern (e.g. HIV, Ebola, SARS)1–3. Understanding patterns of viral diversity in wildlife and determinants of successful cross-species transmission, or spillover, are therefore key goals for pandemic surveillance programs4. However, few analytical tools exist to identify which host species likely harbor the next human virus, or which viruses can cross species boundaries5–7. Here we conduct the most comprehensive analysis yet of mammalian host-virus relationships and show that both the total number of viruses that infect a given species, and the proportion likely to be zoonotic are predictable. After controlling for research effort, the proportion of zoonotic viruses per species is predicted by phylogenetic relatedness to humans, host taxonomy, and human population within a species range – which may reflect human-wildlife contact. We demonstrate for the first time that bats harbor a significantly higher proportion of zoonotic viruses than all other mammalian orders. We identify the taxa and geographic regions with the largest estimated number of ‘missing viruses’ and ‘missing zoonoses’ and therefore of highest value for future surveillance. We then show that phylogenetic host breadth and other viral traits are significant predictors of zoonotic potential, providing a novel framework to assess if a newly discovered mammalian virus could infect people.

In this paper, we discuss an extension to two popular approaches to modeling complex structures in ecological data: the generalized additive model (GAM) and the hierarchical model (HGLM). The hierarchical GAM (HGAM), allows modeling of nonlinear functional relationships between covariates and outcomes where the shape of the function itself varies between different grouping levels. We describe the theoretical connection between HGAMs, HGLMs, and GAMs, explain how to model different assumptions about the degree of intergroup variability in functional response, and show how HGAMs can be readily fitted using existing GAM software, the mgcv package in R. We also discuss computational and statistical issues with fitting these models, and demonstrate how to fit HGAMs on example data. All code and data used to generate this paper are available at: github.com/eric-pedersen/mixed-effect-gams.

No abstract

Bacillus anthracis is a spore-forming, Gram-positive bacterium responsible for anthrax, an acute infection that most significantly affects grazing livestock and wild ungulates, but also poses a threat to human health. The geographic extent of B. anthracis is poorly understood, despite multi-decade research on anthrax epizootic and epidemic dynamics; many countries have limited or inadequate surveillance systems, even within known endemic regions. Here, we compile a global occurrence dataset of human, livestock and wildlife anthrax outbreaks. With these records, we use boosted regression trees to produce a map of the global distribution of B. anthracis as a proxy for anthrax risk. We estimate that 1.83 billion people (95% credible interval (CI): 0.59-4.16 billion) live within regions of anthrax risk, but most of that population faces little occupational exposure. More informatively, a global total of 63.8 million poor livestock keepers (95% CI: 17.5-168.6 million) and 1.1 billion livestock (95% CI: 0.4-2.3 billion) live within vulnerable regions. Human and livestock vulnerability are both concentrated in rural rainfed systems throughout arid and temperate land across Eurasia, Africa and North America. We conclude by mapping where anthrax risk could disrupt sensitive conservation efforts for wild ungulates that coincide with anthrax-prone landscapes.

The realization that complex systems such as ecological communities can collapse or shift regimes suddenly and without rapid external forcing poses a serious challenge to our understanding and management of the natural world. The potential to identify early warning signals that would allow researchers and managers to predict such events before they happen has therefore been an invaluable discovery that offers a way forward in spite of such seemingly unpredictable behavior. Research into early warning signals has demonstrated that it is possible to define and detect such early warning signals in advance of a transition in certain contexts. Here we describe the pattern emerging as research continues to explore just how far we can generalize these results. A core of examples emerges that shares three properties: the phenomenon of rapid regime shifts, a pattern of 'critical slowing down' that can be used to detect the approaching shift, and a mechanism of bifurcation driving the sudden change. As research has expanded beyond these core examples, it is becoming clear that not all systems that show regime shifts exhibit critical slowing down, or vice versa. Even when systems exhibit critical slowing down, statistical detection is a challenge. We review the literature that explores these edge cases and highlight the need for (a) new early warning behaviors that can be used in cases where rapid shifts do not exhibit critical slowing down, (b) the development of methods to identify which behavior might be an appropriate signal when encountering a novel system; bearing in mind that a positive indication for some systems is a negative indication in others, and (c) statistical methods that can distinguish between signatures of early warning behaviors and noise.