Global warming has deeply affected coastal and offshore marine ecosystems and is thought to have a significant impact on the fish stocks and their decline in the northwest Atlantic region in Canada. However, it is difficult for scientists and marine biologists to have a clear and precise insight of this impact. Studying fish aggregations and their evolution in time and space may help scientists to have a better understanding of the decline of fish stocks and elaborate potential solutions for it. Fish aggregations are not only 3D spatiotemporal phenomena which need to be represented and managed in 3D but also have fuzzy boundaries which makes it too difficult to clearly identify and delineate them in the space. Although, Geographic Information Systems (GIS) constitute a powerful tool for handling spatial information, they are prone to problems when dealing with 3D dynamic phenomena, especially, when those phenomena have fuzzy boundaries. This paper addresses these two problems at local and regional scales and proposes two new approaches for the representation and visualization of fish aggregations and their evolution through time and space. At the regional scale, the proposed approach combines fuzzy logic methods with spatial raster representation tools within GIS to provide a more realistic representation and visualization of fish aggregations and their evolution in time. At the local scale, we developed an integrated method based on 3D Delaunay triangulation and the 3D alpha-shapes algorithm to carry out the spatial modeling of fish aggregations in a true 3D space. The applications of these methods to the fisheries data reviled several potentials and limitations of the proposed methods which are discussed throughout this paper.