1. Intensifying agricultural production in sustainable ways is pivotal to increasing food production while reducing environmental impacts. Ecological intensification is based on managing organisms that provide services underlying crop production to simultaneously intensify agricultural production and increase biodiversity. However, few studies address the interactions and trade-offs between biodiversity, multiple ecosystem services and crop production.

2. We experimentally quantified the effect of uncultivated field margins, a prominent practice of ecological intensification, on agricultural production, biodiversity, as well as on multiple ecosystem services and disservices, in an intensive Mediterranean agro-ecosystem. We used a split-plot design and sampled butterflies, rodent and arthropod pests, arthropod natural enemies (both parasitoids and predators), weeds, damage to crop and crop yield in different distances into the field in three tomato and eleven wheat crops along the growing season.

3. Field margins increased natural enemy densities, reduced pest-damage to crop and consequently increased yield in tomato crops. Notably, we found that pest control by one predator species was dominant in the field centre, while parasitoid natural enemies were confined to the field edges. Pest control was more prominent in the late crop-stage compared to early season sampling and field margins increased weed control in tomato crops by reducing weed cover.

4. Field margins increased natural enemy densities in wheat at the beginning of the season, but effects on arthropod pests were inconsistent. Field margins slightly increased weed cover, but had no impact on rodent densities and total yield.

5. Butterfly abundance, but not richness, was positively affected by vegetated field margins.

6. Synthesis and applications: Our results highlight the importance of a mechanistic and holistic approach to promoting ecological intensification. Although field margins provided pest and weed control in the highly intensive tomato crop, they increased weed cover in wheat, which could potentially restrict yields at the field scale. Farmers' guidelines should therefore consider the interactive effects of multiple services on a variety of crops. Moreover, biodiversity components that do not provide crop production services should be independently targeted (e.g. by sowing plants that provide food resources).

The methods are described in detail in the manuscript and supplementary material in Segre et al. (2020) J. Appl. Ecol. 

List of files in the database:

PlotList.csv - sampling design and plot attributes of the tomato and wheat experiment.

VegetationCompositionMtx.csv - vegetation composition in field margins in the experiment plots (species-sites matrix format).

Butterfly.csv - butterfly sampling data.

Rodents.csv - rodents captures data.

Weeds.csv - weed sampling data.

Yield - yield sampling data.

TomatoFruitDamage.csv and TomatoLeafDamage.csv - damage to tomato crop sampling data.

TomatoVacuumData.csv and WheatVacuumData.csv - vegetation dwelling arthropods sampling data in both experiments.

The file MetaData.csv contains a list of column names and descriptions for all files.