Spatial and temporal variability in factors influencing mangrove establishment and survival affects the distribution of mangrove, particularly near their latitudinal limit, where mangrove expansion into saltmarsh is conspicuous. In this paper the spatial variability in mangrove distribution and variability in factors influencing mangrove establishment and survival during the Quaternary period are reviewed, focussing on research at latitudinal limits in Australia and mainland USA. Despite similarities in the response of mangrove to some drivers, the expression of these drivers is both spatially and temporally variable, demonstrating the need for analyses of mangrove-saltmarsh dynamics to move beyond generalisations and incorporate regional and local-scale specificity. We propose i) that precursory recognition that ‘correlation does not mean causation’ is inadequate and assumptions, caveats, and limitations should be clearly articulated in correlative studies; ii) experimental design in manipulative experiments must also articulate the spatial and temporal scale to which the analysis is relevant; and iii) analyses that draw from a range of methods will provide greater confidence. Integrated research programs that transect spatial and temporal scales and incorporate a range of techniques are essential to improve projections. Mangrove-saltmarsh distribution research should move beyond simple models that assume equilibrium between realised and fundamental niches.