[Evaluation and selection of species diversity index under grazing disturbance in alpine mea-dow]

The research selected the plots of six grazing intensities in an alpine meadow in north-eastern Tibet Plateau in four years (2012-2015) and studied the relation between ten species diversity indexes, including two measured indexes (Richness and Abundance) in field and two indexes of each dominance, evenness, richness, integrated indexes, and grazing intensity as well as grazing time aiming at scientific selection of biodiversity index under grazing disturbance. The results indicated that the abundance was a better index than importance value to calculate biodiversity level because it was more sensitive to grazing disturbance. Dominance indexes, including Berger-Parker and Dominance, were not sensitive to grazing intensity and grazing time because they could not clarify the effect of grazing disturbance on dominant species in plant community. Evenness indexes, including Equitability and Evenness, had not relation with grazing intensity, however, the evenness index had a negative correlation with grazing time and it was not influenced by occasional species as well as the variation coefficient of species abundance. Hereby, the evenness index could be chosen for studying evenness change at temporal scale. Richness indexes, including Menhinick and Margalef, had no relation with grazing time, however, the Margalef index had a positive correlation with grazing intensity and the index was not influenced by occasional species. Integrated index, including Shannon and Simpson indexes, had no relation with grazing intensity, however, the Shannon index had a significant positive correlation with species richness and abundance and the index was not influenced by occasional species, and it significantly increased along grazing time. Hereby, Shannon index could be used as an index of studying plant species diversity in long-term. In ten diversity indexes, only the measured indexes in field, including richness and abundance, were signifi-cantly negatively correlated with grazing intensity, and positively correlated with grazing time, and the two indexes were not influenced by occasional species. Hereby, the combination of species richness and abundance mea-sured in field could be considered as the most important indexes for studying plant species diversity under grazing disturbance. Besides, the selection of biodiversity indexes must consider the spatial-temporal feature of grazing, diversity components and research purpose.