Résumé
Dispersal through space or time (via dormancy) determines gene flow and influences demography. Because of their functional similarities, a covariation between dispersal and dormancy is expected. Dispersal and dormancy are anatomically linked in plants, because they both depend on attributes of the seed, albeit this anatomical association is rarely considered when analyzing interactions between dispersal and dormancy. In this paper, we investigate the extent to which dispersal and dormancy can be expected to correlate and how each might influence adaptation to novel environments such as those brought on by climate change. We review the theoretical and empirical literature on the subject with a focus on seed plants. We find that although a negative correlation between dispersal and dormancy has been theoretically anticipated, several models predict deviations from this expectation under scenarios of environmental heterogeneity. The empirical evidence does not support any specific covariation pattern, likely because the interaction between dispersal and dormancy is affected by multiple environmental and developmental constraints. From a climate change perspective, the effects of dispersal and dormancy on population structure are not equivalent: dormancy-mediated gene flow is intrinsically asymmetric (from the past towards the future) whereas spatial dispersal is not necessarily directional. As a result, selection on traits linked to dormancy and dispersal might differ qualitatively. In particular, gene flow through dormancy can only be adaptive if future environmental conditions are similar to those of the past, or if it contributes to novel allelic combinations. We conclude that, in spite of a long tradition of research, we are unable to anticipate a universal relationship between dispersal and dormancy. More work is needed to predict the relative contributions of spatial dispersal and dormancy to gene flow and adaptation to novel environments.