Seabirds share many characteristics setting them apart from other birds. Importantly, they breed more or less obligatorily in local clusters of colonies that can move regularly from site to site, and they routinely exchange breeders. The properties of such metapopulations have only recently begun to be examined, often with models that are occupancy-based (using only colony presence or absence data) and deterministic (using single, empirically determined values for each of several population biology parameters). Some recent models are now frequency-based (using actual population sizes at each site), as well as stochastic (randomly varying critical parameters between biologically realistic limits), yielding better estimates of the behavior of future populations. Using two such models designed to quantify relative risks of population changes under different future scenarios (RAMAS/stage and RAMAS/space), we have examined probable future populations dynamics for three hypothetical seabirds -- an albatross, a cormorant, and a tern. With real parameters and ranges of values we alternatively modelled each species with and without density dependence, as well as with their numbers in a single, large colony, or in many smaller ones, distributed evenly or lognormally. We produced a series of species-typical lines for different population risks over the 50 years we simulated. We call these curves Instantaneous Threat Assessments (ITAs), and their shapes mirror the varying life history characteristics of our three species. We also demonstrated (by a process known as sensitivity analysis) that the most important parameters determining future population fates of all three species were correlation of mean growth rate among colonies; dispersal rate of present and future breeders; subadult survivorship; and the number of subpopulations (=colonies) - in roughly that descending order of importance. In addition, density dependence was found to markedly alter ITA line shape and position, dramatically in the tern. Finally, we show that for each of our three seabirds, a substantial reduction in the risk of the entire population's going to extinction was provided by a metapopulation (i.e. colonial) breeding structure -- thus comfortably confirming what avian ecologists have long known but about which population modellers are somtimes still unsure.