Ecological communities are being impacted by climate change at increasing rates, including an increase in the intensity of disturbances, impacting communities through the destruction and isolation of biota. Meta-communities allow dispersal between communities and habitat patches and are associated with increased population resilience. Recently, meta-community research has expanded to incorporate a food web component. Meta-food webs allow an exploration of how spatial connectivity and dispersal contribute to food web stability. Examining how network topology interacts with food webs is an important step, especially in streams which have dendritic networks that are known to directly influence community structure. We modeled food webs in single patches and in multi-patch linear, lattice, and dendritic networks with and without disturbance. Disturbance models were run for ramp and pulse disturbances at mild and severe intensities for each network topology. Network robustness and mean patch robustness were calculated at each timestep. Multi-patch topologies had a greater network robustness and lower patch robustness than single patches. Disturbance caused a greater decrease in patch robustness and network robustness in the dendritic topology than linear or lattice topologies. However, after the disturbance was over, patch robustness recovered most in dendritic topologies. Network topology and connectivity play a critical role in the resistance and resilience of communities during and after disturbance. Our results highlight the potential capacity for streams to recover from disturbance, which will become more important in freshwater systems as hydrologic disturbance increases. These results have substantial implications for metacommunity, meta-food web, and stream research, as dendritic topologies are not equivalent to the more commonly studied lattice and linear topologies and are likely to have different relationships with dispersal, disturbance, and community structure.