RESUMEN
In many locations, the highest proportion of roaming dogs and cats, might have owners. The prevention of roaming in owned dogs and cats is a key intervention to reduce the number of unsupervised animals in public spaces. Sterilization is an important population management intervention but it is unclear if, apart from its effects on birth rates and animal behavior, it also affects the roaming status of owned dogs and cats. We formulated a directed acyclic graph to represent a causal hypothesis regarding the effect of sex, age and sterilization on roaming behavior in owned dogs and cats. The causal hypothesis was tested using survey data and Bayesian multilevel logistic regressions. The odds of roaming were lower in sterilized dogs and cats but the credible interval (CI) supported the effect of sterilization only for dogs. The odds of roaming were higher in both male dogs and cats and the CI supported the effect of sex. The CI did not support the effect of age on roaming, irrespective of species. If the observed association between sterilization and roaming was causal, then sterilization may contribute to a reduction in roaming. Studies on other populations and the exploration of other potential determinants of roaming are required to better understand the causal relationship between sterilization and roaming.