This data set set belongs to the SP5 litter decomposition experiment that was structured in three parts. First we determined single-species litter decomposition constants of 26 woody species, second litter decomposition of Schima superba was assessed in all 27 CSPs to reveal inter-plot variability using a common standard litter substrate, and third litter diversity was manipulated in 27 plot-specific litter mixtures with species richness ranging from 7 - 17 species. In this dataset remaining mass, nitrogen and carbon concentration für each bag and time step are reported. Please find detailed information below.

Litter bags (15 cm x 15 cm) were made of nylon with 1 mm mesh size and filled with 5 g of dried leaf litter. The decomposition experiment was started in spring 2009. After carefully removing loose forest floor litter, bags were slightly anchored in 5 cm spacing at the ground with 10 cm long metal stakes puncturing the hem. Because an organic humus layer (Oh) was often absent, litter bags were placed directly on the mineral soil in most cases. Gaps between litter bags were refilled with the removed forest floor litter. Replicated litter bags were retrieved in six time steps after 92, 146, 205, 272, 333 and 383 days of decomposition in the field. Because CSPs were spread over the entire GNNR, litter bag retrieval took about seven days to finish. Thus, the given time intervals represent the mean number of days calculated for each sampling period after the start of the experiment. After retrieval, remaining litter was carefully cleaned from adherent mineral soil, ingrown roots and fecal matter, dried at 60 °C until constant mass and weighed after cooling down. We used three complementary approaches to study leaf litter decomposition (s. decription below). (i) Decomposition of single-species leaf litter ("single_species" We determined litter decomposition rate of 26 sub-canopy and canopy tree species occurring in the CSPs under the same microclimatic and edaphic conditions. For each species, 12 litter bags were filled with dried leaf litter, and two replicated bags per species were retrieved at each time step. Owing to the large coverage (several square meters of forest floor) by litter bags, we decided not to conduct the experiment within a selected CSP but rather outside on undisturbed forest floor providing enough space. Hence, litter bags were placed on the ground a few metres next to one plot (CSP 9) in two blocks containing one replicated bag for each time step, respectively. Random arrangement of the 156 bags in each block ensured that species-specific decomposition rates were not affected by spatial variation of microclimatic conditions. (ii) Decomposition of Schima superba leaf litter ("common_substrate") Leaf litter collected from one of the most dominant and abundant tree species (Schima superba) was used to study plot-specific litter decomposition along the tree diversity and successional forest gradient using a total number of 324 litter bags. 12 bags filled with Schima superba leaf litter were placed in every plot with 2 bags to be sampled at each of the 6 retrieval events. Bags were randomly positioned within an area of 1 m by 2 m in the central subplot of the corresponding CSP. (iii) Decomposition of mixed-species leaf litter ("mixed_litter") In the litter mixture approach, the number of litter species per bag equalled the number of species found in the different CSPs based on 24 species for which sufficient quantities of litter could be collected. Hence, species number in plot-specific litter mixtures ranged from 7 to 17 with 0.294 g to 0.714 g of leaf litter per species contained in each bag, respectively (Appendix Table 4). Litter bag species richness was significantly related to plot species richness (Appendix Fig. 10). For each plot, 12 bags filled with a plot-specific mixture were prepared and placed next to the Schima superba litter bags in each CSP covering together an area of about 2 m2. After retrieval, separation into single species was not possible due to the large number of species involved and high fragmentation during decomposition. However, we managed to differentiate between broad-leaved and coniferous litter (Pinus massoninana, Pinus taiwanensis and Cunninghamia lanceolata). Dry mass was determined separately for broad and needle-leaved litter.

CSPs of the BEF China Experiment, Gutianshan Nature Reserve 29°08-29°17'N 118°02'-118°11'E

from spring 2009 one year

Leaf litter from trees

1) For Castanopsis tibetana the second replicated litter bag could not be relocated for time step 92 days (here only one litter bag retrieved) but was rediscovered at time step 146 days (here 3 litter bags retrieved). 2) In some bags caterpillars were discovered that consumed a large fraction of litter material (e.g. Castanopsis tibetana). As we used a mesh size if 1 mm, caterpillars hatched from eggs but could not escape anymore from the litter bags, thus reflecting unnatural conditions. For calculating k-values those litter bags with abnormal mass loss were excluded.

This data set lists all measurements for each litter bag. Calculated decomposition constants (k-values) for each species can be found in the k-values dataset "Leaf litter decomposition in Gutianshan: decomposition rates (k-values)"