We sampled trophobioses from 29 May to 31 July 2014 and 11 September to 8 October on site A and B dayly alternating between sites (for data on Site B see dataset 567). We selected in total 150 plots from tree diversity level 1, 2, 4, 8 and 16 at both field sites. As the abundance of a tree species in a plot decreases with increasing tree diversity, we increased the number of sampled trees per plot accordingly, summing up to a total of 10 359 trees which were sampled once each. On each tree, 20 randomly selected young leaves and the attached branch sections were visually inspected for the occurrence of sap-sucking Hemiptera and honeydew-collecting ants. Voucher specimens of each trophobiosis were collected with soft insect forceps and stored in 99% ethanol. Hemiptera were grouped into morphospecies and further identified in the lab. Ants were first assorted to genera with Bolton (1994) and then identified to species level whenever possible with primary taxonomic literature and the Antweb Database (www.antweb.org). In a few cases Hemiptera and ants could only be identified to morphospecies, which we also refer to as species.

BEF China Main Experiment, Site A, near Xingangshan, Jiangxi

Mai 2014 - July 2014 (Spring) and September 2014 - October (Autumn)

Sap-sucking Hemiptera and honeydew-collecting ants

The data unit of all data is the number of occurrence of trophobiosis, i.e. every interaction between a Hemiptera and an ant species observed on an individual tree was only counted once, regardless of the number of individuals involved. Number of individual Ants and Hemiptera are the sum of the same trophobioses found on one tree.