Soil sampling was carried out on a subset of 100 plots of two overlapping extinction scenarios on Site A (Extinction scenario A and B). The sampüles were collected from young forest communities composed of 1 - 24 species (see general plot information concerning community compositions of the BEF-China experiment: http://china.befdata.biow.uni-leipzig.de/datasets/71; Species nomenclature following the Flora of China). The species composition covers two overlapping species pools and also includes two commercially relevant tree species monocultures (Extinction scenario E, Pinus massoniana and Cunninghamia lanceolata). Moreover, we also sampled plots without trees. Sampling was carried out in July 2015.

Within the plots, we collected soil samples from 12 random positions which followed a frame around the central 8 x 8 trees of every plot. Soil cores with a diameter of 13 mm were taken from the first ten centimeters of the top layer. These subsamples were pooled plotwise, freshly sieved in 2 mm fraction, homogenized and stored in the fridge for analyses.

Plots of Exctinction scenario A and B (http://china.befdata.biow.uni-leipzig.de/datasets/71). 100 plots were sampled, including non-tree community plots and monocultures of Pinus massoniana and Cunninghammi lanceolata (Extinction Scenario E).

Sampling was carried out throughout July 2015.

No tree species were included in the analyses. However, soil cores were taken from communities comprising the following species (and their mixtures according to (http://china.befdata.biow.uni-leipzig.de/datasets/71): Nyssa sinensis Castanea henryi Quercus fabri Lithocarpus glaber Quercus serrata Koelreuteria bipinnata Choerospondias axillaris Castanopsis sclerophylla Triadica sebifera Liquidambar formosana Sapindus saponaria Schima superba Castanopsis eyrei Cyclobalanopsis myrsinifolia Rhus chinensis Cyclobalanopsis glauca Castanopsis carlesii Melia azedarach Triadica cochinchinensis Diospyros japonica Acer davidii Cinnamomum camphora Daphniphyllum oldhamii Quercus acutissima Pinus massoniana Cunninghamia lanceolata

Nitrogen mineralization rates were determined from 10 g fresh soil extracted with 100 ml 1 M KCl. After reacting with a sodium salicylate solution to a indophenol colour complex (Kandeler and Gerber 1988; Niklaus et al. 2001) NH4+ concentrations of the filtered extracts were measured colorimetrically (Microplate Reader Synergy Mx, BioTek, Winooski, United States). Another subsample of 10 g fresh soil was incubated at 40 °C in glas vials for seven days. To ensure anaerobic condition, the soil was water-logged with 15 ml H2O and vials were sealed with a crimp-cap to reduce oxygen exchange to a minimal level (Niklaus et al. 2001). Under anaerobic conditions, soil microbes only produce NH4+ instead of NO3- (Niklaus et al. 2001). After incubation, we filtered and extracted the samples with 85 ml 1.176 M KCl to ensure similar molarity to the non-incubated samples. Ammonium analyses were done analogous to the non-incubated samples. N-mineralization rates were calculated as the difference of the ammonium concentration in the incubated and non-incubated sample divided by the number of days of incubation (Niklaus et al. 2001).