Wind-induced turbulence can strongly impact ecological processes in shallow lake ecosystems. The creation of shelter against wind can be expected to affect both primary producers and herbivores in aquatic food webs. Shelter may benefit particular primary producers more than others by changing relative resource availabilities for different primary producers. Herbivore community compositions may be affected either directly or indirectly as a consequence of changes in their food quantity and quality that, in turn, may affect the transfer efficiency between primary producers and herbivores. A reduction in trophic transfer efficiency resulting from wind-induced turbulence potentially can lead to declines of higher trophic levels, but is generally understudied. Here, we focus on the impact of wind on aquatic primary producers and trophic transfer efficiency. We hypothesised that reducing wind-induced turbulence will stimulate higher trophic production in shallow lakes. However, the multitude of impacts of wind-induced turbulence on aquatic food webs make it challenging to predict the direction of change when creating sheltered conditions. We tested our hypothesis in the shallow waters of a newly constructed archipelago named Marker Wadden in lake Markermeer in the Netherlands. Lake Markermeer has experienced declining numbers of birds and fish. These declines have been related to wind-induced sediment resuspension that potentially limits primary production and trophic transfer efficiency. Marker Wadden is a large-scale restoration project that aims to add sheltered and heterogeneous habitat to the otherwise mostly homogeneous lake, thus targeting the potential problems associated with wind-induced turbulence. We executed a 2-month manipulative field mesocosm experiment in the shallow waters of Marker Wadden to study the effect of reduced wind-induced turbulence (i.e., shelter) on aquatic food webs. Specifically, we studied the effects on primary producers, trophic transfer efficiency between phytoplankton and zooplankton (using zooplankton biomass divided by phytoplankton Chl a as a proxy), and benthic fauna. The experiment consisted of three treatments: no shelter, shelter without macrophytes and shelter with submerged macrophytes (Myriophyllum spicatum) present at the start of the experiment. Our results clearly showed that under unsheltered conditions phytoplankton was the dominant primary producer, whereas in sheltered conditions submerged macrophytes became dominant. Interestingly, submerged macrophytes appeared rapidly in the sheltered treatment where first no macrophytes were visibly present; hence, at the end of the experiment, there was little difference among the sheltered treatments with and without initial presence of submerged macrophytes. Despite that phytoplankton concentrations were 23-fold higher under the unsheltered conditions, this did not result in higher zooplankton biomass. This can be explained by a five-fold greater trophic transfer efficiency between phytoplankton and zooplankton under the sheltered conditions. Furthermore, under the sheltered conditions the Gastropoda density reached 746 individuals m−2, whereas no Gastropoda were found under the no shelter treatment. These findings indicate that for shallow lakes that are negatively affected by wind-induced turbulence, measures aimed at ameliorating this stressor can be effective in facilitating submerged macrophyte recovery, increasing Gastropoda densities and restoring trophic transfer efficiency between phytoplankton and zooplankton. Ultimately, this may support higher trophic levels such as fish and water birds by increasing their food availability in shallow lake ecosystems.