Summary1 Restored floodplains and backwaters lacking a viable propagule bank, may need flood pulses to facilitate inward dispersal of diaspores. Temporal patterns of hydrochorous plant dispersal are, however, not well known. 2 Diversity and abundance of diaspores dispersed in a water body over 12 months were quantified using a 200 µ m net in order to: (i) test for a relationship between discharge and the number of species and diaspores dispersed; (ii) examine the effect of seed buoyancy and seed release period on the length of the dispersal period; and (iii) test whether diaspores of species that disperse during a similar period of the year are characterized by similar dispersal and dormancy traits. 3 A total 359 188 individuals of 174 vascular species developed from 144 samples, with most (90%) from vegetative diaspores and only 10% from seeds. Mean number of species and diaspores varied between months in parallel with discharge levels. Stepwise multiple regression analysis showed that both seed buoyancy and seed release influenced dispersal periods. 4 In general, species that dispersed most diaspores in spring and summer had nondormant seeds, a shorter seed release period and a shorter seed dispersal period than species whose dormant seeds dispersed in autumn and winter. Vegetative diaspores were dispersed on average over 8 months, indicating their importance to long-distance dispersal. Several species dispersed both generative and vegetative diaspores, often in different seasons. 5 Our results may assist the planning of regenerative processes in riverine wetlands at landscape scales, as dispersal phenology, and discharge rates must be taken into consideration. Vegetative diaspores may be more important than seeds, although the latter may extend the species dispersal period into other seasons. Temporal heterogeneity in diaspore dispersal influences the identity of diaspores reaching restored habitats.