Background and aims

Most litter decomposition experiments are early term (less than three years) studies. Unfortunately, early and late terms have different decomposition rates due to different chemical compositions and climatic conditions. Therefore, long-term (more than 5 years) experiments are needed to improve our understanding of decomposition dynamics and unique impacts of different litters.

Methods

A 90-month field trial was conducted using a litterbag methodology employing 330 litter samples from four tree species along a climatic gradient in Gongga Mountain, on the Qinghai-Tibetan Plateau, China. The remaining litter mass (%) was fitted, and models were evaluated using R² and Akaike Information Criterion (AIC). Decomposition rates of litters from the same site and climatic gradient were compared based on application of the optimization model.

Results

The dual negative exponential function (DNEF) model of each litter had the highest R² and lowest AIC value of the different models tested. The decomposition rate of all litters at 2250m was higher than that at 3000m. P. purdomii and B.utilis had higher decomposition rates compared to R. faberi and L.cleistocarpus.

Conclusions

The DNEF model best described the long-term decomposition pattern of the different litters and temperature was the controlling factor of decomposition along an elevation gradient on Gongga Mountain. Higher decomposition rates of P. purdomii and B.utilis were related to their deciduous traits and lower initial C:N.