Recent studies have concluded that release from native soil pathogens may explain invasion of exotic plant species. However, release from soil enemies does not explain all plant invasions. The invasion of Ammophila arenaria (marram grass or European beach grass) in California provides an illustrative example for which the enemy release hypothesis has been refuted. To explore the possible role of plant–soil community interactions in this invasion, we developed a mathematical model. First, we analyzed the role of plant–soil community interactions in the succession of A. arenaria in its native range (north-western Europe). Then, we used our model to explore for California how alternative plant–soil community interactions may generate the same effect as if A. arenaria were released from soil enemies. This analysis was carried out by construction of a 'recovery plane' that discriminates between plant competition and plant–soil community interactions. Our model shows that in California, the accumulation of local pathogens by A. arenaria could result in exclusion of native plant species. Moreover, this mechanism could trigger the rate and spatial pattern of invasive spread generally observed in nature. We propose that our 'accumulation of local pathogens' hypothesis could serve as an alternative explanation for the enemy release hypothesis to be considered in further experimental studies on invasive plant species.