Identifying and protecting essential fish habitats like spawning grounds requires an accurate knowledge of fish spatio-temporal distribution. Data available through commercial declarations provide valuable information covering the whole year and consequently they could prove useful to identify spawning grounds. We developed an integrated framework to infer fish spatial distribution on a monthly time step by combining scientific and commercial data while explicitly considering the preferential sampling of fishermen towards areas of higher biomass. Over the spawning period, we applied a method to identify areas of persistent aggregation of biomass and interpret these as spawning areas. The model is applied to infer monthly maps of three species (sole, whiting, squids) in the Bay of Biscay on a 9-years period. Integrating several commercial fleets in inference provide a good coverage of the study area and improves model predictions. The preferential sampling parameters give insights into the temporal dynamics of the targeting behavior of the different fleets. Persistent aggregation areas reveal consistent with the available literature on spawning grounds, highlighting that our approach allows to identify potential areas of reproduction.