Abstract: Taking the bare patches in Horqin desertified grassland as test objects, seed traps were installed to investigate the amount and species composition of the seeds diffused on the patches during growth season, and artificial seeding was conducted to monitor the seedling emergence on the patches, aimed to understand the relative limitation between seed availability and seedling emergence on the vegetation restoration on bare patches. Of the seeds diffused on the bare patches, an amount of 71-623 grains·m^-2 came from three psammophyte species, and of 73-410 grains·m^-2 came from nine non-psammophyte species. The seedling emergence of psammophytes (9.3-22.3 plants·m^-2) was higher than that of non-psammophytes (0-3.4 plants·m^-2), and the emergence rate of psammophytes Agriophyllum squarrosum, Artemisia wudanica, and Caragana microphylla was 12.2%-34.1%, 4.5%-10.3%, and 0-0.5%, respectively, whereas that of non-psammophyte Setaria viridis was 0-6% and of the rest non-psammophytes was null. The seedling emergence of psammophytes on bare patches (9.3-22.3 plants·m^-2) was higher than that on vegetation-covered patches (7.7 plants·m^-2), while the seedling emergence of non-psammophytes on bare patches (0-3.4 plants·m^-2) was lower than that on vegetation-covered patches (4.5 plants·m^-2). This study suggested that during the vegetation restoration of bare patches in degraded grassland, the limitation of psammophytes seed availability was more important than that of seedling emergence, but for non-psammophytes, the importance of seed availability limitation was equivalent to that of seedling emergence limitation.

Key words: Rhizomatous grasses, Milky ripe stage, Photosynthetic rate, Leaf temperature