The literature on the impacts of heterogeneity and selection in population analysis has been limited largely to the conventional unistate perspective in which only decrements are considered; and their temporal (selectivity) impacts on independent subpopulations examined. In this paper, the focus is on the evolutionary dynamics of (multistate) multiregional populations whose interdependent subpopulations can experience increments as well as decrements. It is shown that in such instances migration rates that are not true occurrence-exposure rates are ambiguous, because they depend on the relative weightings existing during the initial model-fitting period. Net migration rates, lifetime migration rates, and return migration proportions all are imperfect measures of migration propensities, and their use as input measures to an analysis should be avoided whenever possible.