summary The coexistence of many different species has invented biologists at all times. Diversity and composition of communities are influenced by disturbances and heterogeneity of environmental conditions. Although in nature the spatial distribution of environmental conditions is usually self-correlated, this aspect is rarely taken into account in models studying the coexistence of species. In this work, we therefore discussed, using numerical simulations, the coexistence of species and their characteristics within a self-correlated environment. In order to take this spatial element into account, we have developed a metacommunity model (a set of communities linked by species dispersion) that is geographically explicit. In this model, species compete with each other to establish themselves in a limited number of places in a heterogeneous environment. The species are characterised by six lines: niche optimum, niche width, dispersion capacity, competitiveness, investment in reproduction and survival rates. We were particularly interested in the influence of spatial self-correlation and disruption on species diversity and favoured traits in the metacommunity. We have shown that spatial self-correlation can have antagonistic effects on diversity, depending on the level of disturbance considered. The influence of spatial self-correlation on average dispersion capacity in the metacommunity also depends on disturbance and survival rates. Our results have also shown that many species with different degrees of specialisation (i.e. different niche widths) can coexist. However, specialist species are favoured in the absence of disturbances and when the dispersion is unlimited. On the other hand, a high level of disturbance selects more generalist species, combined with low competitiveness. Spatial self-correlation of the environment, interacting with the intensity of disturbance, thus significantly influences coexistence and species characteristics. These characteristics are in turn often involved in important processes, such as the functioning of ecosystems, the ability of species to respond to invasions, habitat fragmentation or climate change. This work has led to a better understanding of the mechanisms responsible for coexistence and species characteristics, which is crucial to predicting the future of natural communities in a changing environment. AbstractUnderstanding how many different species can coexist in nature is a fundamental and long-standing question in ecology. Community diversity and composition are known to be influenced by heterogeneity in environmental conditions and disturbance. Though in nature the spatial distribution of environmental conditions is frequently self-related, this aspect is considered in models investigating species coexistence. In this work, we have thus addressed questions relating to species coexistence and composition in spatial linked territories, with a numerical simulations approach.To take into account this spatial aspect, we developed a spatial explanation of metacommunity (a set of municipal linked by dispersal of species). In this model, species are largely equivalent, and compete for space in a heterogeneous environment. Species are characterised by six life-history traits: Niche optimal, niche breadth, dispersal, competitiveness, reproductive investment and survival rate. We were particularly affected in the influence of environmental space autocorrelation and disturbance on species diversity and on the traits of the species favoured in the metacommunity. We showed that space autocorrelation can have antagonistic effects on diversity dependent on disturbance rate. Similar, space autocorrelation interacted with disturbance rate and survival rate to shape the mean dispersal ability observed in the metacommunity. Our results also revealed that many species with various deficiencies of specialisation (i.e. different niche breadths) can exist together. However specialist species were favoured in the absence of disturbance, and when dispersal was unlimited. In contrast, high disturbance rate selected for more generalist species, associated with low competitive ability.The spatial structure of the environment, together with disturbance and species traits, thus strongly impacts species diversity and, more significant, species composition. Species composition is known to affect several important metacommunity properties such as ecosystem functioning, resistance and reaction to invasion, to habitat fragmentation and to climate changes. This work reduced a better understanding of the mechanisms responsible for species composition, which is of crucial importance to predict the fate of natural metals in changing foodstuffs