Changes in litter decomposition rate of dominant plants in a semi-arid steppe across different land-use types: Soil moisture, not home-field advantage, plays a dominant role

Abstract

Litter decomposition is a key process of nutrient cycling in ecosystems, and has been studied in a wide range of grasslands, including the steppes on the Mongolian Plateau. However, information is hardly available on the responses of grassland litter decomposition to different land-use types, especially if the litters produced from grazing or mowing grasslands have home-field advantage in decomposition. In a five-year field experiment examining the effects of land-use types on grassland ecosystems in central Inner Mongolia, we determined the litter decomposition rate and quality indicators of three dominant steppe plants (Leymus chinensis, Stipa krylovii and Cleistogenes squarrosa) produced in grazing, mowing or enclosed grasslands, both in their respective home habitats versus in reciprocally displaced habitats. We found that (i) the litters from the grazing grassland had higher decomposition rate than that from enclosed grassland, especially when all litters were placed in the enclosed grassland; and the decomposition rate was negatively related with the C/N ratio, lignin concentration and lignin/N ratio in litters (P < 0.001). (ii) During the two-year decomposition period, the mass loss rates, and the difference in mass loss rates of grazing litters between in situ and displaced habitats, were the highest in the first summer period (June-November). (iii) Compared to the decomposition in situ, the litters from grazing or mowing grasslands decomposed apparently faster when being displaced in enclosed grassland, whereas the litters from enclosed grassland decomposed significantly slower (P < 0.05) when being displaced in grazing or mowing grasslands. The faster decomposition in enclosed grassland was associated with the better soil moisture. No home-field advantage is shown in the decomposition of steppe plants across land-use types in the studied grassland, possibly due to the non-significant divergence in soil microbial communities over the five-year experimental period. Our results indicate that the effects on litter decomposition of the soil moisture variation between the grazing/mowing and the enclosed grasslands, is greater than the effects of the variation in litter quality and possibly in microbial communities across different land-use types. Further HFA studies to involve the variation in soil microbial communities are warranted.