<p>Understanding the potential drivers of spatial-temporal patterns in biodiversity has been a central tenet in biogeography and palaeontology for decades. More than 30 hypotheses have been proposed, including null models and theories based on environmental controls, energy, area, speciation/extinction dynamics, time, habitat features, ecological niches and biotic interactions. Yet, no consensus has been reached, and question remains whether a primary cause explains temporal trends and spatial patterns in biodiversity such as the latitudinal biodiversity gradient. Here we combine a macroecological model with global climate simulations to show that the niche-environment interaction may explain changes in global marine biodiversity and associated large-scale spatial patterns during the Phanerozoic (last 541 million years). We show that the niche-environment interaction imposed both a species carrying capacity and spatial constraints on marine biodiversity that defined the latitudinal biodiversity gradient. Although our model suggests that climate modulated the niche-environment interaction, hence spatial biodiversity patterns, it also demonstrates that palaeogeographical evolution imposed changes in shallow-water area, continental fragmentation and the location of landmasses relative to climatic belts and may have constituted the fundamental driver of changes in global marine biodiversity at the geological time scale. Therefore, several mechanisms interacted to balance the niche-environment interaction and drove the trajectory of marine biodiversity during the Phanerozoic.</p>