2.2. Attractiveness of the Flower Margin to Natural Enemies and Phytophagous Insects

Visual samplings were performed to estimate the attractiveness of the margins to adult hoverflies. For three minutes per plot (12 minutes per margin), an observer counted the number of adult hoverflies hovering above or touching the flower of the insectary or spontaneous plant species present in a plot. The observations were conducted under favorable weather conditions: sunny to lightly overcast days, with no or low wind speed (0–4.2 m/s) and temperature above 15 °C.

Beating tray samplings (BTS) were conducted to estimate the attractiveness of the margins to natural enemies (except adult hoverflies) and phytophagous insects. BTS consisted of three consecutive beats with the hand. In each plot, four BTS were taken after visual observations had been made. In the flower margin, each BTS corresponded to an insectary plant species, while in the control margin the beatings were made on flowering species whenever possible. In the latter case, the plant species where the BTS was conducted were noted. The arthropods falling as a result of the BTS were collected on a white tray (24 × 35 cm) and classified into the following groups: predators (Araneae, predatory Thysanoptera, predatory Coleoptera, predatory Heteroptera, Syrphidae larvae, Chrysopidae, Cecidomyiidae, Trombidiidae and Forficulidae), parasitoids (both adults and mummies), and phytophagous insects (Aphididae, phytophagous Thysanoptera and phytophagous Heteroptera). Individuals were returned to the margin after visual identification to family or order in the case of spiders (Table S4).

Both visual samplings and BTS were conducted fortnightly from April (petal fall) to September in 2015 and 2016. In both years, the samplings were performed on coincident weeks.