Abstract: In this paper, nested plots were used to investigate the plant species number of subalpine meadow plant communities in the Eastern Qilian Mountains of China. The relationships between plant species frequency and sampling area were analyzed, and the speciesarea curves of the plant communities were fitted with ten selected models. Correlation index (CRI) was then used to test the fitness of the models. The results showed that when the sampling area increased over 16 m², the related increase of species number in four quadrats decreased significantly, and basically maintained at the same level. Furthermore, the species number in each quadrat was more than 75% of the total species number in the quadrat. When the sampling area reached about 32 m², the species number in high-yield quadrats I and II and in low-yield quadrats I and II were basically maintained at a certain level, accounting for 88%, 89%, 84%, and 92% of the total species number in the quadrats, respectively. With a CRI of ≥0.9, eight of the ten models had good fitness. Models S=(b+alnA)^c and S=c/(a+A^-b) were the best, followed by models S=aln(bA+1), S=b+alnA, S=aA^b, and S=c-ae-bA, and only models S=aln(A-1) and S=a(1-e-bA) had bad fitness.

Key words: riparian zone, Luanhe Basin, altitude, human activity., invasive plant