Area monitoring by remote sensing and spatiotemporal variation of rubber plantations in Xishuangbanna

Abstract: As one of the major commercial crops in Xishuangbanna, Southwest region of China, rubber plantations have been expanding rapidly in the recent decade, which mainly occupied agricultural lands and secondary forest covers. These land use and land cover changes have negative influences on regional ecosystem and ecological processes. Based on the RS and GIS technology, the growth process of rubber plantations in Xishuangbanna was explored in this article and a method of judging the forest age of rubber plantations was put forward in the study area. Firstly, by using the multi-temporal Landsat images (16-days interval from January to April in 2014), time series changes of NDVI, NIR, Bright and Wetness of rubber plantations and other vegetation covers were constructed. It was found that the period from late February to mid-April was the optimal time for extracting rubber plantations. NDVI and NIR had better effects on the discrimination of rubber plantations and other vegetations. Then, annual Landsat images were used to construct NDVI time series changes of newly planted rubber trees and mature rubber trees, and determine the NDVI thresholds of each growth periods of rubber plantations. Because of the limitation of accessibility of annual Landsat images, only data of 14 consecutive years (2000-2014) were collected in this research. Each age stage of the rubber plantations was judged by these ways. Finally, from the planting area, the spatial distribution, the structure of forest age and the conversion of land use types between the rubber plantations and nature forest, the spatiotemporal changes of rubber plantations in Xishuangbanna during1990-2014 were studied. The results showed that: 1) Rubber plantations in Xishuangbanna can be divided into three classes based on different forest ages: young rubber plantations with forest age ranging from 0 to 5 years; mature rubber plantations with forest age ranging from ＞5 to 13 years; old rubber plantations with forest age above or equal to 13 years. 2) During 1990-2014, the area of rubber plantations increased, young mature and old rubber plantations converted from each other, the average area and density of rubber plantations in the nature reserves was far below that in the non nature reserves in Xishuangbanna. 3) From spatial center to the periphery, the spatial distributions of the rubber plantations in 1990 were from old rubber plantations to mature or young rubber plantations, in 2000 were from old rubber plantations to mature rubber plantations, then to young rubber plantations, in 2010 and 2014 were both from young rubber plantations to old rubber plantations, then to mature rubber plantations. This trend was more evident in 2014 than 2010. 4) The expansion of rubber plantations occupied the nature forests in Xishuangbanna. The conversion between the rubber plantation and natural forests in altitude from 0 to 800 m was most intense, followed by which in altitude from 800 to 1 000 m, and the competition in altitude above 1 000 m was the weakest. Tropical monsoon forest was most intensively interfered by rubber plantation expansion in natural forest.