Flowering plant sampling
During the three sampling campaigns, we recorded all flowering forb species (henceforth plant species) nearby the pan traps, in order to characterize the local plant community, respectively the nectar supply for the pollinators. For this purpose, we placed a circle (r=5m) around the traps with eight equally sized segments (‘pie slices’). For each segment, we recorded the presence of flowering forb species (excluding wind-pollinated forbs Sanguisorba minor and Plantago lanceolata ), in order to have a measure of local abundance for the forbs (ranging from zero to eight). In orientation to Fornoff et al. (2017), we gathered functional flower traits of the recorded forbs that are considered to be relevant for pollinators from Biolflor Database (Klotz et al. 2002) and Jäger (2016): UV radiation [a,b], UV reflectance [numeric 1-6], color [categorical: yellow, red, blue, white, rose, purple, violet], flowering height [continuous] and nectar access [categorical: open, half-open, hidden] (Supplementary information, Appendix 1). We hypothesized that functional diversity of these traits positively affect pollinator diversity (hypothesis 4). Further, we predict that a higher share of yellow flowers attract a higher number of pollinators, since this color is preferred by numerous bee and hoverfly species (Leong and Thorpe 1999, Lunau 2014). A larger flowering height is supposed to attract more pollinators, first, because of the generally high conspicuousness of tall floral units, second, because of the ”effective pollination” hypothesis, according to which tall stature resulting from strong apical dominance attracts greater pollinator visitation, thus allowing larger pollen loads (Donelly et al. 1998).