TeaComposition experiment results in Mukhrino field station

The general scheme of TeaComposition project was implemented in the study [Keuskamp et al., 2013]. Two plots (200 m2) (waterlogged graminoid-Sphagnum lawn and relative dry pine-dwarf shrubs-Sphagnum treed bog) differing mainly in water regime and associated vegetation composition were within the raised bog. The third plot was located in well-drained dark coniferous-deciduous mixed forest (see Table A1 in APPENDIX for details of vegetation, plots coordinates and soil properties).
At each site two replicas were established with 32 tea bags from each type buried inside a square 4 m2 for incubation over the period of 3 years with the retrieval time after 3, 12, 24, and 36 months.
Before the incubation, the tea was dried at 70°C for 48 hours and the initial weight was recorded (see Table A4 in APPENDIX for details of installation protocol). The tea was incubated in the aerated acrotelm layer (0-5 cm below the Sphagnum surface) at the two bog sites and under the litter layer (0-5 cm, approximately in organic layer as it is poorly distinguished in the podzolic soils) at the forest site, in June 2016. Additionally to the tea bags, 11 types of native litters (Sphagnum fuscum, S. balticum, Chamaedaphne calyculata, Ledum palustre, Rubus chamaemorus, Carex limosa, Eriophorum vaginatum, Scheuchzeria palustre, Populus tremula, Betula pendula, Pinus sibirica) were packed into the litter bags (mesh size 0.2 mm) and installed in a similar way in October 2016 due to the lack of shed leaves in the beginning of vegetation season. Therefore, these data could not be considered in present publication.
After 3 months of incubation the first 8 bags from each plot were collected, cleaned from earth and roots, dried and weighed.
The calculation of k (decomposition constant), S (stabilization factor) and ar (decomposable fraction of rooibos or green tea) was made as described in [Keuskamp et al., 2013].
The significance of differences between three plots were calculated by Wilcoxon rank sum test [Sharaf et al., 1986] (this test is equivalent to the Mann-Whitney U test) using RANKSUM MATLAB-function.