The rate and distribution of ecological functions is modified by how species respond to the composition of landscapes. Extensive loss of habitats has led to habitat restoration becoming an important management tool, however, it is not clear where restoration sites should be located in heterogeneous landscapes to maximise outcomes for ecological functions. We used restored oyster reefs, and the guild of predators associated with them, as a model system to test whether, and how ecological functioning is modified by the spatial context of restoration sites in marine landscapes (i.e. seascapes). We measured predation rates and surveyed predators using videoed deployments of ‘squidpops’ (dried squid tethered using fishing line) at multiple restored oyster reefs and nearby control sites in Queensland, Australia. Sites were located in different spatial contexts in a seascape composed of a mosaic of habitat types. Predation rates at restored oyster reefs were double those at control sites. Seascape context was important in modifying these predation rates; consumption near reefs was significantly lower when reefs were close to seagrass and mangroves. By contrast, higher rates were observed on reefs surrounded by non-vegetated seafloor, far from seagrass and mangroves. In addition, the distance over which predation extendeds into the surrounding unvegetated areas was greater on reefs father from vegetation. Strategically placing restoration sites in heterogeneous landscapes can maximise the effects of habitat restoration for ecosystem functioning and modify the distance over which these effects extent into surrounding seascape.