1.Plant demography is known to depend on both spatial dynamics and life history, but how these two factors interact is poorly understood. We conducted a longitudinal study of the wind-pollinated annual plant Mercurialis annua that varies geographically in its sexual system to investigate this interaction. 2.Metapopulation demographic models predict that regular population turnover should be a more common feature of monomorphic than dimorphic populations because males and females cannot found new populations by selfing but hermaphrodites can. We tested the prediction that rates of population turnover would be higher in monomorphic compared to dimorphic regions. 3.We surveyed 356 populations of M. annua along five regional transects in Morocco and the Iberian Peninsula over a three-year period to examine their demography and persistence. Each transect crossed a transition in the sexual system, from a monomorphic region where almost all populations were hermaphroditic to a dimorphic one in which most populations had separate sexes (males with females or hermaphrodites). 4.As predicted, rates of local apparent extinctions (i.e. the disappearance of adult plants) were nearly 50% higher in monomorphic compared to dimorphic regions. Local extinctions appeared to be driven by changes in vegetation cover, with extinctions tending to occur in sites in which perennial cover also declined. This suggests that disturbance is a primary agent of local extinctions. 5.We further examined the influence of regional dynamics on local demographic properties by investigating patterns of spatial autocorrelation in population density across years. We found positive spatial autocorrelations in plant densities within regions for both sexual systems. However, these positive autocorrelations extended over shorter distances in monomorphic regions, perhaps as a result of greater population flux in these regions. 6.Synthesis: Our study shows that population dynamics may be influenced by processes acting at a range of spatial scales: within patches, across patches within sites, and across sites within regions, as well as by life-history variation. In M. annua, regional variation in apparent extinction rates is affected by life history and implicated in regulating the geographical distribution of populations with different sexual systems.