Larval and phenological traits predict insect community response to mowing regime manipulations.

For the restoration of biodiversity in agricultural grasslands, it is essential to understand how management acts as an ecological filter on the resident species. Mowing constitutes such a filter: only species that possess functional traits enabling them to withstand its consequences can persist in the community. We investigated how the timing of mowing modulates this filtering effect for insects. We predicted that two traits drive species responses. Species with larval development within the meadow vegetation will suffer more from mowing than species whose larvae develop in or on the ground, or outside the meadows, while species with a later phenology should benefit from later mowing. We conducted a five-year experiment, replicated at 12 sites across the Swiss lowlands, applying three different mowing regimes to low-intensity hay meadows: (1) first cut of the year not earlier than 15 June (control regime); (2) the first cut delayed until 15 July; and (3) leaving an uncut grass refuge on 10-20% of the meadow area (after earliest first cut on 15 June). Before the first cut in years 4 or 5, we sampled larvae of Lepidoptera and sawflies, and adults of moths, parasitoid wasps, wild bees, hoverflies, ground beetles, and rove beetles. Overall, before the first cut of the year, abundances of species with vegetation-dwelling larvae were higher in meadows with delayed mowing or an uncut grass refuge, with some taxon-specific variation. In contrast, species whose larval development is independent of the meadow vegetation showed no differences in abundance between mowing regimes. Species richness did not differ among regimes. For species with vegetation-dwelling larvae, a fourth-corner analysis showed an association between early phenology and the control regime. No associations were found for the other functional groups. Our results show that slight modifications of mowing regimes, easily implementable in agri-environmental policy schemes, can boost invertebrate abundance, potentially benefitting insectivorous vertebrates.

Specialisation and diversity of multiple trophic groups are promoted by different forest features.

While forest management strongly influences biodiversity, it remains unclear how the structural and compositional changes caused by management affect different community dimensions (e.g. richness, specialisation, abundance or completeness) and how this differs between taxa. We assessed the effects of nine forest features (representing stand structure, heterogeneity and tree composition) on thirteen above- and belowground trophic groups of plants, animals, fungi and bacteria in 150 temperate forest plots differing in their management type. Canopy cover decreased light resources, which increased community specialisation but reduced overall diversity and abundance. Features increasing resource types and diversifying microhabitats (admixing of oaks and conifers) were important and mostly affected richness. Belowground groups responded differently to those aboveground and had weaker responses to most forest features. Our results show that we need to consider forest features rather than broad management types and highlight the importance of considering several groups and community dimensions to better inform conservation.