For endangered species that are hunted, the establishment of mixed conservation areas including both hunting zones and sanctuaries to complement translocation actions (i.e. reinforcement) can improve both hunting yields and population sustainability. However, the effects of this type of management on the demography of the exploited species are not well understood. We used multi-event capture-recapture modelling in a population of captive-bred houbara bustards Chlamydotis undulata translocated into a mixed conservation area in Morocco. The specific management practice of our system (hunting regime varying in time and space) led to a quasi-experimental situation that allowed the differentiation of natural' from hunting-induced' mortality and movement between areas. The analysis uncovered strong asymmetries in both movement and survival that were not only due to direct hunting effects. Firstly, movement probabilities were higher from the sanctuary to the hunting areas than vice versa, even in years without hunting. Secondly, in addition to a direct effect of hunting on mortality in hunting areas, our results uncovered permanent differences in both areas (even outside the hunting period). Overall, our results were consistent with predictions under a source-sink dynamic model but illustrated that mixed conservation areas should not merely be treated as homogeneous systems with spatially heterogeneous hunting pressure but rather as fully heterogeneous systems. The patterns observed may be related to (1) the choice and design of hunting and sanctuary areas by managers, which might not be neutral with respect to habitat quality, or (2) indirect consequences of hunting via an effect on local growth rate and density.