Logo image
Back
FROM SPILLOVER TO SUSTAINED TRANSMISSION: MODELING THE MULTIDIMENSIONAL ECO-EPIDEMIOLOGY OF ZOONOTIC DISEASES
Dissertation   Open access

FROM SPILLOVER TO SUSTAINED TRANSMISSION: MODELING THE MULTIDIMENSIONAL ECO-EPIDEMIOLOGY OF ZOONOTIC DISEASES

Katie K. Tseng
Doctor of Philosophy (PhD), Washington State University
12/2025
DOI:
https://doi.org/10.7273/000008354
pdf
Tseng_Dissertation_ProQuestFinal5.66 MBDownloadView
Open Access CC BY-ND V4.0
xlsx
Supplementary Data 2.12.39 MBDownloadView
Spreadsheet The predicted probailities from the link prediction models Open Access CC BY-ND V4.0
xlsx
Supplementary Data 2.245.17 kBDownloadView
Spreadsheet The predictive functions of associated proteins for genes ranked by their loading contributions to influential principal components Open Access CC BY-ND V4.0
xlsx
Supplementary Data 2.333.40 kBDownloadView
Spreadsheet The importance and ranks of mammal traits and viral features incorporated in the host exposure model, susceptible host model, and link prediciton models Open Access CC BY-ND V4.0
xlsx
Supplementary Data 2.422.79 kBDownloadView
Spreadsheet Host-virus associations for link prediciton based on Orthopoxvirus genomes extracted from NCBI Open Access CC BY-ND V4.0

Abstract

Eco-epidemiology Modeling Zoonoses Ecology Global Health Infectious Diseases
Over half of known infectious diseases are zoonoses, diseases transmitted between animals and humans that contribute to an estimated 2.7 million human deaths per year. The global burden of zoonoses is driven not just by the spillover of pathogens from wildlife or domestic animals to humans, but by the ability of pathogens to sustain onward transmission after emergence. In this dissertation, I investigate the multidimensional factors (i.e., the interacting host, pathogen, and environmental dimensions of zoonotic ßrisk) that enable sustained zoonotic transmission through an eco-epidemiological framework, drawing on diverse modeling approaches to predict risk and identify key drivers of pathogen persistence. In epidemic contexts, bidirectional transmission of zoonotic pathogens can perpetuate spillover-spillback cycles, leading to larger-scale outbreaks, the establishment of new animal reservoirs, and the spread of disease beyond endemic regions. However, for many emerging zoonoses, the host range that defines the hazard landscape remains poorly characterized. To address this gap, we used orthopoxviruses (OPVs) as a case study, given their broad and variable host range as shaped in part by the evolutionary loss of accessory genes involved in immune evasion. Using boosted regression trees, we developed a link prediction model that integrates host traits with viral genomic features to predict compatibility between mammal genera and OPV species. The link prediction model outperformed a traditional model trained on host traits alone (AUC = 0.98 vs. 0.886), predicting a more taxonomically diverse set of potential OPV hosts and revealing more distinct geographic hotspots of predicted host species overlap. With this model, we demonstrate how combining ecological and molecular data can improve predictions of host-virus compatibility and potentially strengthen targeted wildlife surveillance efforts. In endemic contexts, the persistent circulation of pathogens within host and vector populations are maintained through complex ecological interactions that shape both vector ecology and when and where contact between humans and vectors occurs. However, the behavioral pathways that shape exposure to infected vectors are frequently overlooked, particularly in vector-borne disease research. To examine how network-structured human behaviors contribute to endemic disease persistence, we reconstructed the social network in a rural region of the Argentine Chaco endemic for Chagas disease, where frequent residential mobility has been shown to interrupt sustained exposure to vector infestation. By integrating social network metrics into spatial models of house infestation and child T. cruzi infection, we found that human social connectivity increased the odds of both outcomes, but only in interaction with mobility. While moving alone was generally protective, mobile households that maintained social connections faced higher odds of infestation and infection even after adjusting for spatial proximity, suggesting that social networks may facilitate passive vector transport or sustained exposure through social visiting. Together, these studies demonstrate how integrating biological, ecological, and social processes in zoonotic risk assessment, can help identify key drivers of sustained transmission across complex ecological and social landscapes.

Metrics

7 File views/ downloads
12 Record Views

Details

Logo image