2. Here, we tested the effect of dyke-slotting to reconnect a floodplain on the trait response of mollusc communities over 6 years, based on a random stratified sampling design (plots within areas) with before and after control impact. Trait characteristics included flood resistance, drought resistance and resilience from which community-weighted mean trait values were derived. FD and its components (alpha, beta, gamma) were calculated with Rao's quadratic entropy.3. Flood duration in the restored area increased from 4-13 to 6-17 weeks/year after reconnection, similarly to the reference area (4-10 to 6-14 weeks). Hence, the re-connection by dyke-slotting had no substantial effect on flood duration (due to seepage water).
4.Despite that, dyke-slotting triggered weak but significant shifts in the trait compositions of the restored floodplain mollusc communities. Traits reflecting species' resilience, crushing resistance and drought resistance increased in abundance, while traits characteristically found in more stable habitats decreased (e.g. crossfertilisation, diet specialisation). Overall, trait composition of the restored area progressively moved to that found in the active floodplain.
5.While there were some significant temporal changes in FD within areas (reference, restored), there were no significant effects of dyke-slotting on any of the FD indices.