Infectious disease spread in connected communities

Ecohydrology and epidemiology share a deep bond in infectious disease modelling. The former focuses on the interaction among species and their water-controlled environment  (i.e., the study of water controls on the biota). The latter revolves around specific host-pathogen relationships and the pathogen's propagation in space and time. Combining these disciplines is crucial when considering some disease-causing water-based pathogens, such as Vibrio cholerae and Opisthorchis viverrini. The importance of a spatially explicit framework to better understand the spread of airborne diseases was also asserted during the COVID-19 pandemic. While many countries still struggle to fight disease transmission, comprehensive knowledge regarding the spread of these diseases in space is still lacking. To this end, this Thesis bridges ecohydrological and epidemiological concepts to advance towards a more thorough understanding of the mechanisms regulating the spread of the pathogens mentioned above and diseases in space and time. Throughout the development of this Thesis, these mechanisms have been pinpointed to epidemiological indicators such as reproduction numbers and epidemicity indices derived from the concept of reactivity in ecological dynamics. These metrics often result from algebraic analyses based on the eco-epidemiological models, which are hereby showcased both in continuous and discrete time. Specifically, models in continuous time, here adapted to water-based diseases, are built on sets of coupled ordinary differential equations that consider any relevant hydrological forcing. A model in discrete time, derived from a suitable discretization of an integro-differential model in continuous time, is applied to an airborne disease, COVID-19. Appropriate calibration algorithms, based on either a Markov Chain Monte Carlo Bayesian framework or on particle filtering techniques, are implemented to calibrate the models on data on the 2010s Haitian cholera outbreak, the endemic transmission of O. viverrini along the Mekong River, and the COVID-19 pandemic in Italy. Where relevant, human actions, such as vaccinations or non-pharmaceutical interventions, are embedded in the model to account for the reduced transmission.The experiments on the two considered water-based pathogens show that including human mobility and riverine transmission into any relevant model is essential to correctly capture the pathogen's spread and, therefore, design appropriate containment measures that, among other things, also target spatial transmission. In addition, a new framework for the computation of effective reproduction numbers based on epidemiological data and mobility fluxes indicates that including the latter into renewal equations may often produce different values of the epidemiological indicators. This suggests that failing to include spatial transmission may misrepresent the actual epidemiological situation.The implications of these results are diverse. On the one hand, they suggest that embedding spatial connectivity into the epidemiological models substantially helps design containment measures to curb the spread of the disease. On the other hand, owing to the more precise nature of the epidemiological metrics computed within a spatially explicit framework, so-computed reproduction numbers and epidemicity indices can improve our surveillance systems and function as early-warning indicators that may anticipate future outbreaks.