Trophic cascades focusing on direct (consumptive) effects have been well studied in marine ecosystems. However, less attention has been given to cascades involving indirect interactions embedded in distinct habitats. We focused on the interactions between a non-indigenous predator (the green crab Carcinus maenas), a consumer (the mud crab Panopeus herbstii), and a prey species (juvenile eastern oysters Crassostrea virginica). These interactions were studied in 3 small yet distinct habitats of increasing complexity: bare sediments, patches of blue mussels Mytilus edulis (BM), and a unique habitat consisting of giant Irish moss Chondrus crispus and mussels combined (IMBM). In the field, green crab predation rates on mud crabs were estimated in each of the abovementioned habitats using tethering experiments. The results showed that green crab foraging was most effective in bare sediments and least effective in IMBM, i.e. the least and most complex habitats, respectively. Trials conducted in the laboratory with mud crabs foraging over oysters showed a similar outcome: oyster mortality rates declined with increased habitat complexity. However, when trials were conducted in the presence of a green crab, this pattern reversed, and oyster mortality was lowest in bare sediments. Mud crab behavior was consistent with these results: in the presence of a green crab, mud crabs were less active and spent more time sheltering, whereas in its absence, the opposite pattern was observed, especially in bare sediments. These behaviorally driven indirect interactions are dependent on the presence of a non-indigenous predator and mediated by the type of habitat.