Wilson-Aggarwal, J.; Goodwin, C.; Lechenne, M.; Swan, G.; Sidouin, M.; Silk, M.; Moundai, T.; Ozella, L.; Tizzoni, M.; Cattuto, C.; McDonald, R.

Transmission of infection is affected by the spatial-temporal dynamics of host contacts. Domestic dogs Canis familiaris share pathogens with humans and wildlife, and managing dog-mediated diseases is a priority for public health and conservation interests. We combined proximity sensors and GPS tracking to analyse spatial-temporal variation in contact among free-ranging dogs in six villages in rural Chad, during both the wet and dry seasons. We investigated dyadic interactions between dogs from different villages, the same village but different households and the same household. We assessed variation in (i) the probability of individuals having had contact, (ii) the hourly frequency of contact and (iii) contact durations. Our results highlight clear seasonal and hourly patterns in contact behaviour. Contacts between dogs from different villages were rare, short in duration, primarily between male-female dyads and predominantly occurred within villages and during the dry season. Contact between dogs in the same village peaked at dawn and dusk. Sex differences were most pronounced in the wet season, where males from different households had the highest hourly contact probabilities, followed by male-female dyads. For all dogs, contact durations were longer in the dry season, but showed little hourly variation. Contact patterns were not equal in space, and the probability of individuals having had contact was less than 5% when dwellings were more than 500m apart. Spatially, the probability of contact was lowest outside the village, but this increased in the dry season and peaked in the morning hours. Contact durations were notably longer outside the village, where they increased in duration for between-household dyads in the dry season. At a coarse temporal scale, variation in dog contacts within and among households, and rarely between villages, may underpin seasonal variation in the incidence of dog-mediated diseases. Variation at finer temporal (hourly) and spatial scales (around households, within and outside villages) highlights the importance of routine behaviours and space use in determining patterns of contact between dogs. Practitioners should consider behavioural heterogeneities, such as those reported here, when using strategic models to support disease management decisions.