Abstract: Multi-scale networked observation is the basic means to obtain regional ecological information, and it is the source of data for comprehensively and deeply understanding the dynamics of ecosystems and assessing the interrelationships between ecosystems, global changes and human activities. The Chinese Ecosystem Research Network (CERN), based on the field observation stations of important ecosystem types in China, adopts unified protocols to carry out long-term positional observation and research on typical ecosystems such as forests, grasslands, deserts and marshes in China, so as to provide scientific data support for the construction of the national ecological environment. This dataset covers the background, plant community species richness, biomass and habitat information of 84 biological long-term observation plots from 10 forest, 6 desert, 2 grassland and 1 marsh stations of CERN from 1998 to 2010. The dataset has gone through a strict three-tier audit and quality control process, and constructed a biological and habitat dataset with the annual observation data of the biological long-term observation samples as the basic unit. The dataset can provide data support for the distribution of biological resources, the relationship between biological and environmental changes and human activities.

Key words: long-term observation, networking study, vegetation types, soil characteristics, climate change