Fishers exploit populations that are heterogeneously distributed in space and time without full knowledge of species distributions and with fishing gear that is not fully selective. The ability to change catch composition is limited by species mix at a particular location and time and the capture characteristics of the fishing gear. Models capturing the dynamics of the fisheries (‘fleet dynamics’ models) are often simplistic due to a lack of knowledge of the processes driving these catches, which occur both at large and small scales. We develop a simulation framework to investigate the importance of scaling on the interactions between fish populations and fisheries dynamics. The framework provides i) a realistic but tractable biological model of fish populations in space and time, including daily population processes (mortality, growth and recruitment) and population movement implemented as a combination of diffusive density-dependent processes and migrations; and, ii) a realistic fishing simulation model to capture how fishers may exploit heterogeneously distributed fish populations with different values and uncertain knowledge about the underlying spatial processes. We generate a model system where we investigate the consequences of scaling and data aggregation and validate this simulated data against data collected on fisheries operating in the Celtic Sea. The simulation allows a more in-depth understanding of factors important when using fisheries-dependent data to develop spatial management measures, from the micro- to the large-scale and individual to population processes, not otherwise possible due to the limitations in ‘real-world’ spatiotemporal data on fish distributions.