Trap nests are a well-established tool to monitor cavity-nesting Hymenoptera and to assess parasitism rates. Despite only 5-10% of all non-parasitoid solitary bee and wasp species are colonizing trap nests, it has been show that the cavity-nesting part of the Hymenoptera community is representative for the overall Hymenoptera community in terms of species richness and abundance. In the BEF-China project, we used standardized trap-nests in 40 plots in Site A and 48 plots in Site B. We included all VIPs and additional plots to balance sampling to at least 8 replicates per species richness level (except for 24 species), on Site A eight monocultures were dropped due to logistical limitations. Per plot, two wooden posts with two trap nests (Site B four trap nests), each consisting of a 22 cm long, 12.5 cm diameter plastic tube filled with dried Arundo donax L. (Poaceae) internodes of varying diameter (2-20 mm) were exposed. Trap nests were set up approximately 1.5 m above the ground and controlled monthly for nests that are easily distinguishable by the characteristic nest sealing pluck. Internodes containing nests were replaced with new internodes of the same diameter. Nests were taken to the laboratory, opened and reared in glass test tubes closed with cotton wool till hatching. We found a diverse set of bee, wasp, and parasitoid (and other enemy) species. Individual species differ markedly in their nesting and feeding ecology.

Trap nests (eight traps per plot; four traps each attached to a wooden post; located at two diagonal corners of the core area, close to xxyy: 0707 and 1212) were used to collect solitary bees and wasps in the Main experiment. Abundance and species richness of bees and wasps allow to test to which extent woody plant species richness affects the abundance, species composition and diversity of, e.g. predators (which might profit from higher structural diversity or better resource availability in more plant diverse plots) or pollinators (which might profit from more diverse flower resources). Traps consisted of plastic tubes (diameter 12.5 cm, length 22 cm) that were filled with dried internodes of the reed Arundo donax L. (2-20 mm diameter). Trap nests were exposed from August 2014 to August 2015. All trap nests were checked monthly for nests of solitary bees and wasps. These were then reared in the laboratory till hatching. Taxa were first sorted to morphspecies and then identified to species level with the help of taxonomic experts whenever possible.

Main experiment Site A, all VIPs and additional plots to balance sampling to at least 8 replicates per species richness level (except for 24 species)

September 2014 - July 2015

Solitary cavity-nesting Hymenoptera

The data allow to test to what extent woody plant species richness affects the abundance, species composition and diversity of, e.g. predators (which might profit from higher structural diversity or better resource availability in more plant diverse plots) or pollinators (which might profit from more diverse flower resources). Individual numbers presented are brood cell numbers, i.e. the number of brood cells provisioned by within one reed internode, which can be multiple internodes for the same species within one trap. The identity and number of higher species 1 and 2 are given for each nest. If not empty then the host species was attacked by the higher species with a complete loss of the offspring in the attacked brood cell. Therefore, to calculate the number of host species hatched from the internode, the sum of the higher species needs to be subtracted from the number of brood cells of the host species. Hosts and higher species are provided to construct interaction networks and calculate parasitism rates, which we test in response to plant species richness, succession and host occurrence.