We assessed the relative importance of bottomÁup and topÁdown processes in structuring an eelgrass community in Sweden, a system impacted both by eutrophication and overfishing. Using artificial seagrass as substrate, we manipulated nutrient levels and predator abundance in a full-factorial cage-experiment.The results revealed a seagrass community dominated by strong top-down processes controlling the aggregate biomass of mesograzers and macroalgae. In the absence of predators the large amphipod Gammarus locusta became very abundant resulting in a leaf community with low biomass of algae and smaller mobile fauna. One enclosed gobid fish predator reduced the abundance of adult G. locusta by 90%, causing a three to six times increase in the biomass of algae, smaller mesograzers and meiofauna. Numerous small predators in uncaged habitats reduced the biomass of G. locusta and other mesograzers by 95% in comparison to the fish treatment, further increasing the biomass of epiphytic algae and meiofauna. Although water column nutrient enrichment caused a temporal bloom of the filamentous macroalgae Ulva spp., no significant nutrient-effects were found on the algal community at the end of the experiment. The only lasting nutrient-effect was a significant increase in the biomass of G. locusta, but only in the absence of ambient predators.These results demonstrate that mesograzers can respond to enhanced food supply, increase their biomass and control the algal growth when predation rates are low. However, in the assessed system, high predation rates appear to make mesograzers functionally extinct, causing a community-wide trophic cascade that promotes the growth of ephemeral algae. This topÁdown effect could penetrate down, despite a complex food-web because the interaction strength in the community was strongly skewed towards two functionally dominant algal and grazer species that were vulnerable to consumption. These results indicate that overexploitation of gadoid fish may be linked to increased macroalgal blooms and loss of eelgrass in the area through a trophic cascade affecting the abundance of mesograzers.

Using a combination of empirical and model studies we tested whether European shore crab larvae (Carcinus maenas) from environments with different tidal regimes in the North Sea area have different swimming behaviors, and whether this affects connectivity and settlement success of larvae. Laboratory studies demonstrated the presence of an inherited tidal migration rhythm in newly hatched crab larvae from the mesotidal Danish Wadden Sea, and field studies showed that postlarvae swam in surface water almost exclusively during flood tides, suggesting that larvae use selective tidal stream transport to control the dispersal process. In contrast, shore crab larvae from microtidal Skagerrak displayed a nocturnal vertical migration behavior that appeared to switch to a diurnal behavior at the end of the postlarval phase, indicating an adaptation to avoid visual predators and to use wind‐driven transport to reach shallow settlement areas. A biophysical model showed that tidal‐migrating larvae in the Wadden Sea had two times higher settlement success than larvae with a diel behavior. However, no differences in settlement success were found between the two larval behaviors in microtidal Skagerrak, where lower fitness is suggested for tidal‐migrating larvae due to higher predation mortality from visual predators. We suggest that the differences in inherited larval behavior in larvae from meso‐ and microtidal regions reflect local adaptations maintained through natural selection of successful recruits. Consistent with recent population genetic studies, modeled connectivity of shore crabs indicated an oceanographic dispersal barrier to gene flow in Eastern Wadden Sea that may facilitate such adaptations.