Colossoma macropomum is an ecologically and economically important fish distributed throughout the major tributaries of the Amazon River. C. macropomum require a suite of habitat types for different life stages making them potentially susceptible to the impacts of habitat fragmentation and alteration. As a means of better understanding the potential impacts of development, baseline data on connectivity and patterns of gene flow in species from relatively undisturbed habitat will be of value to monitor potential ecosystem impacts of anthropogenic habitat alteration on native fish communities. We used 13 single sequence repeat markers to determine if fine-scale structuring could be detected at the landscape scale at the Pacaya Samiria National Reserve, Perú. We also applied a model testing approach to evaluate the strength of different migration models, including panmixia, stepping stone and isolation models. Bayesian clustering detected a single genetic grouping across 131 fish. However, a comparison of marginal likelihoods for alternative migration models across PSNR supported a stepping stone model, rather than panmixia (Probability ~1.0). These results demonstrate that even in highly migratory fish with limited genetic structure, the effects of anthropogenic aquatic habitat alterations can be explored using genetic data.