Evidence for a beneficial effect of corridors on species richness and abundance in habitat patches is mixed. Even in a single microecosystem of microarthropods living in moss patches connected by a moss corridor, experiments have had different results (positive and neutral). This paper attempts to provide an explanatory framework for understanding these results. I developed a stochastic individual-based model of the moss—microarthropod microecosystem. Some of the movement parameter values were estimated from two manipulation experiments. Assuming mortality independent of the season, and assuming the corridors merely increase migration rates between patches, only a very weak beneficial effect of corridors was possible in simulations. Incorporating a seasonal pattern to mortality caused some simulated populations to die out, which were then occasionally rescued by migrants from the adjacent patch. Corridors were slightly beneficial if there was little or no immigration from the surrounding matrix. In contrast, corridors were very beneficial in simulations that incorporated lower emigration to the matrix when a corridor was present, even for moderate levels of immigration from the matrix. Thus corridors may reduce the chance of species extinction in patches even when the lifespan of the individuals is long relative to the time-scale in question. The beneficial effect in this case can act via two possible mechanisms: seasonal mortality imposing brief periods of high vulnerability to extinction, and the presence of a corridor reducing the rate of emigration to the matrix by encouraging movement along the corridor. Either one or both of these mechanisms may have operated in the study whereby corridors had a beneficial effect, but not in the study whereby corridors did not have a beneficial effect. This work demonstrates that corridor effectiveness is dependent on the species and landscape in question, and that it is important to understand the mechanisms by which corridors function.