2.2. Experimental design
In August 2018, an area of Salix cupularis with basically the same growth condition was selected to delineate a test sample site for exogenous N and P addition. The soil basic physicochemical properties, site conditions and initial values of the chemical composition of litter in the test plots are shown in Table 1. The topography and climate types in western China are complex, and the atmospheric N deposition is lower compared with other regions, with an annual N deposition of 3.62 g·m-2·year-1 in western Sichuan (Yang et al., 2018). A According to the N deposition and N, P ecological stoichiometric ratio, control group without fertilization, three fertilization types (N, P, and NP) and three fertilization gradients were set up in this experiment for a total of 10 treatments: low N (N1, 3 g·m-2·year-1), medium N (N2, 6 g·m-2·year-1) and high N (N3, 9 g·m-2·year-1); low P (P1, 1 g·m-2·year-1), medium P (P2, 3 g·m-2·year-1), and high P (P3, 6 g·m-2·year-1); low NP (N1+P1), medium NP (NP2, N2+P2) and high NP (NP3, N3+P3). Urea (CO(NH2)2, N: 46.67%) and sodium dihydrogen phosphate (NaH2PO4, P: 44.65%) were used as fertilizers. A completely randomized design was used to randomly select three Salix cupularis for each treatment in test plot, for a total of 30 Salix cupularis. A 1m×1m sample square was randomly laid within the vertical projection of the canopy of the thickets, and the litter was collected, removed from the thickets, and brought back to the laboratory for drying and shredding. A portion of the shredded litter was used for initial nutrient concentration analysis, and three samples (50g) of each treatment were accurately weighed, mixed with each addition treatment, and packed into 10cm×15cm nylon mesh decomposition bags. The nylon mesh bags were randomly placed in the 1m×1m sample square of the thickets, sealed with sealing clips to ensure that soil animals could not enter, and the four corners of each litter bag were fixed with stainless steel wire ties.