*2.5. Data Analysis*

The content of acid hydrolysable components (%) was presented as % of litter dry mass, and the remaining acid hydrolysable components (*AHCr*, % of original; Equation (1)) were calculated as follows:

$$AHC\_{\ell}(\%) = (AHC\_{\ell} \times M\_{\ell}) / (AHC\_{\bullet} \times M\_{\bullet}) \times 100\tag{1}$$

where *AHCC*0 and *AHCt* are the initial content of acid hyrolysable components and their contents at time *t* (*t* = 1, 2, ... , 12), respectively; and *M*0 and *Mt* are the initial litter dry mass and the masses at time *t* (*t* = 1, 2, . . . , 12).

We first examined the overall effects of variations in snow depth and litter species over time via variance analysis (ANOVA). Differences between deep and shallow snow plots were assessed by paired *t*-tests, or by Wilcoxon signed-rank tests if sample sizes were unequal in MATLAB R2012a (MathWorks Inc., Natick, MA, USA). A path analysis was used to distinguish how variation in snow depth affects acid hydrolysable components during litter decomposition by altering the environment (litter temperature and gravimetric water content), litter quality (e.g., C/N ratio) and soil microbes (e.g., microbial biomass C; MBC) in AMOS 22.0 (IBM SPSS, Chicago, IL, USA). We also compared the relationships between litter mass remaining and the remaining acid hydrolysable components for each litter species.
