ABSTRACTAimChanges in landscape configuration significantly impact ecosystems and the services they provide, including disease regulation for both humans and wildlife. Land use conversion usually favors disturbed‐adapted species, which are often known reservoirs of zoonotic parasites, thereby potentially escalating spillover events (i.e., the transmission of parasites to new hosts, including humans). Here we aimed to investigate how alterations in landscape use and configuration influence the distribution and co‐occurrence of potential hosts of zoonotic and epizootic parasites.LocationBrazilian Atlantic Forest.Time PeriodData collection spanned from 1997 to 2019.Major taxa studied small mammals.MethodsWe integrated ecological network metrics and joint distribution models while accounting for phylogenetic relationships and functional traits to answer two main questions: (1) do small mammal species considered central hosts in the transmission of parasites exhibit a higher probability of occurrence in landscapes with reduced native vegetation areas? (2) Do small mammal hosts that share a higher number of parasites have higher co‐occurrence probabilities?ResultsOur results demonstrated that species identified as significant hosts in our centrality network analysis displayed an increased probability of occurrence in landscapes that are both more fragmented and have a higher proportion of farming areas, hence fewer native vegetation areas. Regarding the relationship between species co‐occurrence and parasite sharing, our findings indicated that most strong co‐occurrences were prevalent within groups with higher parasite fauna similarity, but not all species sharing parasites had a higher probability of co‐occurring.ConclusionsHere we highlight the effects of landscape conversion on small mammal species, including how different configurations of land use can influence both central and non‐central host occurrences. Besides, our results also indicate that parasite transmission may be overestimated when the co‐occurrence probability of potential host species is not considered. We highly recommend incorporating co‐occurrence data to estimate zoonotic risk.