关键词: 云冷杉林, 落叶松林, 物种组成, 径级结构, 空间分布

Abstract: Xishuangbanna, the largest rainforest region in China, is one of the most important biodiversity conservation areas in the world. Studying the temporal and spatial variations of vegetation in Xishuangbanna and its relationship with climate factors can provide scientific supports for environmental management. This study, based on MODISNDVI dataset, aims to assess the spatialtemporal dynamics and spatial autocorrelation distribution pattern of NDVI and its response to thermalmoisture factors in Xishuangbanna from 2001 to 2015. Spatial autocorrelation analysis, linearregression analysis and partial correlation analysis were carried out. From 2001 to 2015, annual NDVI showed a slowly upward trend, with a rate of 0.11%·10 a^-1 and the most pronounced enhancement in 2011-2015. NDVI had a “season inversion” phenomenon, i.e., the lowest NDVI value occurred in summer. In the past 15 years, mean NDVI of Xishuangbanna was 0.875, with a pattern of “high in the east and low in the west”. The NDVI threshold within 0.85-0.9 accounted for the largest proportion of the area. From 2001 to 2015, the Global Moran’s I of NDVI was greater than 0.65, showing a significant positive spatial autocorrelation, but the degree of spatial agglomeration of NDVI had weakened and tended to be fragmented. In 2013, the spatial heterogeneity of NDVI was the strongest. There were differences in vegetation change across different climate types. In the north tropical zone, the areas with improvement and degradation accounted for 28.74% and 9.40%, respectively. In the south subtropical and mid-subtropical zones, the deteriorated area was slightly higher than the improved area. Temperature was the dominant factor driving vegetation growth in Xishuangbanna, which was positively correlated with NDVI from December to February of the following year. Precipitation from June to November was significantly negatively correlated with NDVI.

Key words: spruce-fir forest, larch forest, species composition, size class structure, spatial distribution.