| dc.description.abstract | Livestock movement is an intrinsic part of animal husbandry (i.e., breeding, maintenance, slaughter of livestock), significantly impacting global economies and public health due to its potential for contagious disease spread. This dissertation focuses on two key areas: understanding uncertainties in cattle movement networks and their correlation with epidemic modelling, and introducing a methodology that characterises mobility patterns as flows while considering spatial aspects to realistically capture uncertainties and associated risks.
There are an estimated one billion cattle heads in the world used for various products. Delivering these animal goods efficiently leads to stress in the system by increasing the number of animals kept, traded, and their possible contacts. This situation poses significant economic threats due to highly contagious diseases like brucellosis and foot-and-mouth disease, which spread easily through contact. While some countries have precise monitoring through electronic tags (e.g., Australia, Canada), many major producers lack such tracking systems (e.g., Mexico, USA), enabling unregulated cattle trade and introducing uncertainty into official movement data. These uncertainties directly impact epidemic behaviour estimation, leading us to model three common uncertainties in cattle movement networks and evaluate them through simulations in synthetic and real networks from Minas Gerais, Brazil.
We introduce a general methodology to complement the benefits of network analyses. Our proposed methodology converts movement networks into vector-field representations, significantly accelerates simulations and enhances understanding of risk areas, trade predictability, and interest points like sinks and sources. While demonstrating its effectiveness in cattle trading, this approach is flexible and can be applied to various areas ( if we model the phenomena as cases of origin-destination (O-D) flows), including human mobility. | en_GB |
