Pathogen transmission is a complex process that scales from allocation of a host's energy to immune defense to contacts between individuals. I'm particularly interested in how heterogeneity in host resources (i.e., food availability) influences wildlife physiology and behavior and how these effects scale up to determine infection spread across landscapes. For many wildlife, interactions between consumers and their resources are being heavily modified by human activities that subsidize animals with novel and abundant food. Both intentional and unintentional forms of "resource provisioning" (e.g., bird feeders and fruit plantations) have been responsible for the emergence of virulent pathogens in wildlife and humans, encouraging a better understanding of how these resource shifts influence infection processes and when and where food supplementation may amplify pathogen transmission. My research combines simple theoretical models, phylogenetic meta-analyses, and field studies of wild bats and birds to better understand how resource provisioning affects pathogen transmission and spillover risks across scales. See below for brief summaries of projects and approaches.
linking consumer–resource and host–pathogen theory
Most epidemiological models ignore bottom-up relationships between hosts and their food resources, but integrating resource dependence into mechanistic models can have important implications for the local and spatial spread of pathogens. Our work has shown that strong resource-dependent immune defense (i.e., host resistance) can prevent pathogen invasion even in the face of resource-improved fecundity, survival, and contact rates. At larger scales, improving a moderate to large proportion of host habitat can allow virulent pathogens to invade a metapopulation and spread across all occupied patches, increasing spillover risk. An extension of this work with Richard Hall and funded by the NSF Population Biology of Infectious Diseases REU Site examined how variation in patch quality affects host colonization between habitat types and assessed the consequences for metapopulation persistence and spatial infection dynamics. Future projects also principally aim to better link these effects of local resource abundance and quality on both within- and between-patch host processes (e.g., using stochastic and spatially explicit modeling frameworks).
resource supplementation and wildlife health
Much of my work on linking infectious disease dynamics and consumer–resource interactions is motivated by the applied example of how human activities accidentally and intentionally subsidize wildlife populations with food. I've used both systematic reviews and meta-analysis to show that substantial variation in decreased or increased infection outcomes from resource provisioning can be attributed to parasite taxonomy and the type of feeding used. More recent work shows that wide-ranging, dietary generalist wildlife species are most prone to experience elevated infection risks with bacteria and viruses when provisioned. I have also organized a symposium on this topic at the 2016 Annual Meeting of the Ecological Society of America and a forthcoming Theme Issue in Philosophical Transactions of the Royal Society B: Biological Sciences. I have also collaborated on an interdisciplinary NSF study of how supplemental food impacts the health of urbanized white ibis in South Florida, both by assessing chronic stress with hematology and by using simple theory and sensitivity analyses to understand processes that allow Salmonella to persist in urbanized ibis flocks. As part of my postdoctoral position with Raina Plowright, I am combining longitudinal data on flying fox colonies with roost-level data on history of food shortages and urbanization to test how deforestation-driven urban habituation of bats around low-quality food resources in Australia affects temporal patterns of Hendra virus shedding and consequently risks of spillover to horses and humans.
vampire bat immunology and infection
The common vampire bat is the primary reservoir host of rabies virus throughout Latin America, with this pathogen being spread on a potentially nightly basis when blood-feeding bats feed. Livestock expansion into rainforest habitats could provide vampire bats with an abundant and stable food source, which could decrease stress from starvation and improve antiviral defenses that help resist lethal rabies infection. My PhD used mark–recapture studies across 10 field sites in Peru and Belize that varied in local livestock abundance to explore the consequences of this diet shift for bat stress, immunity, and bacterial infection. We showed that bats in livestock-dense habitats invest more in their innate rather than adaptive immune response, that these innate immune profiles were strong predictors of bacterial infection, and that these immunological responses to provisioning could explain site-level infection patterns. Other work funded by a NSF DDIG and in collaboration with Ken Field is using RNA-Seq to quantify differential immune gene expression between low- and high-livestock sites and to develop better immunological tools for this reservoir host. We are also using population genetics to infer connectivity between bat colonies and to test if and how between-patch movement is influenced by local livestock density and its landscape distribution. Relationships between resource availability, antiviral defense, and connectivity will then be built into a spatially explicit transmission model of vampire bat rabies using likelihood-based inference and parameter estimation.
persistence and network dynamics of bacteria in bats
Bats are notorious as important reservoirs for zoonotic viruses, yet we understand far less about their role in the spread of zoonotic bacteria and how such pathogens are maintained in wild populations. Our work in the vampire bat system has focused on two bacterial (and potentially zoonotic) pathogens, Bartonella spp. and hemoplasmas. Extensions to our livestock density work have shown both bacterial pathogens are genetically diverse and widespread in vampire bats and display endemic infection dynamics across sites. Analyses of vampire bat saliva samples also support plausible direct transmission of hemoplasmas between bats and to or from prey. I am currently building simple theory with Richard Hall to explore the relative contribution of bat–bat contact and arthropod vectors to transmission and building a time series of these infections within Belize vampire bats to confront models with data. New work is taking a more extensive approach by sampling the diverse Neotropical bat community for these infections and using phylogenetic comparative methods and network analyses to identify traits that make particular species more infected and central to transmission.
distribution and impact of mercury exposure on bats
Mercury is a widespread contaminant that has neurotoxic effects on many wildlife taxa. Such impacts have been minimally studied in bats, which could be especially prone to negative fitness consequences of chronic mercury exposure owing to their slow life histories and ubiquitousness in contaminated anthropogenic environments. Long-term mercury exposure could also have distinct effects on bat immune defense, as bats may have components of their immune systems that are unique among mammals. Our work in the vampire bat system has shown that even concentrations of mercury considered sublethal correlate with shifts in innate immunity. Related work has quantified mercury concentrations across a wide range of other Neotropical bats and is now testing if this immune sensitivity of bats to mercury is a widespread phenomenon throughout Chiroptera and its implications for susceptibility to infection. I am also involved in assessing whether Australian flying foxes are exposed to mercury in urban habitats where they forgo nomadism.