Abundance counts for ant species visiting nutrient baits. Nutrient baits were six falcon tubes per replicate filled with different nutrient solutions, with four replicates per plot. After three hours placement in the field, falcon tubes were closed and ant individuals contained were counted and identified to (morpho)species level.

The MultiTroph core plots combine plots from both the random and non-random extinction scenarios of the BEF-China main experiment (tree plots only; tree+shrub plots were excluded). At Site B, the complete set of MultiTroph core plots was selected, together with 32 random extinction plots (VIP) from Site A, resulting in a total of 182 plots (see comments for details on omitted plots). Data was collected by performing a standardized nutrient baiting experiment (Peters et al. 2014, Journal of Biogeography 41(12), 2245-2255). A set of six different nutrient baits containing sugar (C), salt (S), phosphate (P), amino acid (A), water (W) and oil (L) was used. All nutrient solutions, except for oil, which was pure rapeseed oil, were diluted with purified water. The purified water (W) also served as negative control. Each bait consisted of 8 mL of nutrient solution placed in a 15 mL centrifuge tube (1.5 cm diameter). The solution was sealed with cotton wool to prevent leakage, while still allowing it to saturate the cotton. Each set of six different nutrient solutions served as one bait station, with four replicates per plot placed at the center of the edges of the 10 x 10 inner tree positions within a plot (East, North, South, West), each positioned between four adjacent tree positions (tp1–tp4). After three hours, baits including attending ants were collected by closing the tube lids.

Ants (Formicidae)