In the last decade, various spatial and temporal methodologies were developed to investigate the processes that drive ecological and evolutionary patterns. However, these methods frequently fail to acknowledge that the observed patterns result from the overlap of different underlying processes. In order to understand how the patterns are formed, we must have recourse to methods that allow us to disentangle these simultaneous processes. Here we develop a hierarchical spatial predictive process (PP) combined with a separable temporal PP to disentangle and describe those overlapping processes in one very frequent setting in ecology and evolution: multilevel spatio-temporally indexed data. We present our methodology through a case study of fisheries discards and investigate for example whether the inclusion of the hierarchical structure and the temporal processes of the system alter the observed spatial patterns. Recently it is recognized that understanding the processes driving discards is essential to sustainably manage and conserve marine resources. The results show that consideration of multiple underlying processes dramatically changes the pattern and characteristics of the discards hot- and coldspots. In the Irish Sea, the inclusion of the hierarchical structure of the system leads to the reduction of the hot- and coldspots. Simultaneously, our model identifies key bi-annual fluctuations in the temporal process which, together with the variance associated at the level of individual fishing trips in the hierarchical structure of the data explained most of the variance driving discards. Whether the hierarchical, spatial and temporal processes are considered together or not can profoundly alter our understanding of what constitutes an appropriate mitigation measure. Misidentification of hotspots can culminate in inappropriate mitigation practices which can sometimes be irreversible. As the proposed method offers a unified approach for understanding the processes that drive observed patterns, many areas in ecology such as conservation and epidemiological studies can benefit from its use, increasing the effectiveness of management plans.