Interactions between individuals of different species are highly relevant in the potential transmission of shared pathogens in multi-host systems. In recent decades, several technologies to study pathogen transmission have been developed, such as proximity loggers, GPS tracking devices, and/or camera traps. Despite the diversity of methods aimed at detecting contacts, the analysis of transmission risk is often reduced to contact rates and the probability of transmission given contact. However, the latter process is continuous over time and unique for each contact, and it is influenced by the characteristics of the contact and the pathogen's relationship with both the host and the environment. In this study, we utilized a movement-based model that decomposes transmission into contact formation, contact duration, and host characteristics, assigning a unique transmission risk to each contact. We aimed to assess whether this more comprehensive approach reveals disease transmission dynamics that are not detected with more traditional approaches. The model was built from GPS data from two management systems in Spain where animal tuberculosis (TB) circulates: a national park, and an area with extensive free-range pigs and cattle farms. In addition, we assessed the effect of the GPS device sampling rate on the performance of the model. Considering the specific conditions under which each contact occurs (i.e., whether the contact is direct or indirect, its duration, the hosts characteristics, the environmental conditions, etc.) resulted in the identification of different transmission dynamics compared to a model based solely on contact rates. This indicates that not taking these conditions into account may result in misidentifying the key species in disease transmission. The different transmission dynamics identified between both management systems highlight the need to analyze each system independently. We found that temporal intervals greater than 30 minutes in the GPS tracking data resulted in missed interactions, and intervals greater than 2 hours may be insufficient in interaction studies for epidemiological purposes. This study describes a clear and repeatable methodology to study pathogen transmission from GPS data, and provides further insights to understand how TB is maintained in multi-host systems under different management scenarios in Mediterranean environments.

Interactions between individuals of different species are highly relevant in the potential transmission of shared pathogens in multi-host systems. In recent decades, several technologies to study pathogen transmission have been developed, such as proximity loggers, GPS tracking devices, and/or camera traps. Despite the diversity of methods aimed at detecting contacts, the analysis of transmission risk is often reduced to contact rates and the probability of transmission given contact. However, the latter process is continuous over time and unique for each contact, and it is influenced by the characteristics of the contact and the pathogen's relationship with both the host and the environment. In this study, we utilized a movement-based model that decomposes transmission into contact formation, contact duration, and host characteristics, assigning a unique transmission risk to each contact. We aimed to assess whether this more comprehensive approach reveals disease transmission dynamics that are not detected with more traditional approaches. The model was built from GPS data from two management systems in Spain where animal tuberculosis (TB) circulates: a national park, and an area with extensive free-range pigs and cattle farms. In addition, we assessed the effect of the GPS device sampling rate on the performance of the model. Considering the specific conditions under which each contact occurs (i.e., whether the contact is direct or indirect, its duration, the hosts characteristics, the environmental conditions, etc.) resulted in the identification of different transmission dynamics compared to a model based solely on contact rates. This indicates that not taking these conditions into account may result in misidentifying the key species in disease transmission. The different transmission dynamics identified between both management systems highlight the need to analyze each system independently. We found that temporal intervals greater than 30 minutes in the GPS tracking data resulted in missed interactions, and intervals greater than 2 hours may be insufficient in interaction studies for epidemiological purposes. This study describes a clear and repeatable methodology to study pathogen transmission from GPS data, and provides further insights to understand how TB is maintained in multi-host systems under different management scenarios in Mediterranean environments.