Exploring effective sampling design for monitoring soil organic carbon in degraded Tibetan grasslands

JOURNAL OF ENVIRONMENTAL MANAGEMENT

卷:173

Article

Chang, Xiaofeng¹  Bao, Xiaoying⁵  Wang, Shiping^3,4,6,7  Zhu, Xiaoxue^2,5  Luo, Caiyun²  Zhang, Zhenhua²  Wilkes, Andreas⁸ 

[1] Northwest A&F Univ, Inst Soil & Water Conservat, State Key Lab Soil Eros & Dryland Farming Loess P, 26 Xinong Rd, Yangling 712100, Peoples R China

[2] Chinese Acad Sci, Northwest Inst Plateau Biol, Key Lab Adaptat & Evolut Plateau Biota, Haibei Alpine Meadow Ecosyst Res Stn, 23 Xinning Rd, Xining 810008, Peoples R China

[3] Chinese Acad Sci, Inst Tibetan Plateau Res, Key Lab Alpine Ecol & Biodivers, 16 Lincui Rd, Beijing 100101, Peoples R China

[4] CAS Ctr Excellence Tibetan Plateau Earth Sci, Beijing 100101, Peoples R China

[5] Univ Chinese Acad Sci, Beijing 100049, Peoples R China

[6] Tibet Univ, Naqu Integrated Observat & Res Stn Ecol & Environ, Lasa 850012, Peoples R China

[7] Chinese Acad Sci, Inst Tibetan Plateau Res, Lasa 850012, Peoples R China

[8] World Agroforestry Ctr ICRAF, East & Cent Asia Program, 12 Zhongguancun Nan St, Beijing 100081, Peoples R China

关键词: Soil monitoring;    Effective sample size;    Sampling interval;    Grassland degradation;   

DOI  :  10.1016/j.jenvman.2016.03.010

来源: Elsevier

【 摘 要 】

The effects of climate change and human activities on grassland degradation and soil carbon stocks have become a focus of both research and policy. However, lack of research on appropriate sampling design prevents accurate assessment of soil carbon stocks and stock changes at community and regional scales. Here, we conducted an intensive survey with 1196 sampling sites over an area of 190 km(2) of degraded alpine meadow. Compared to lightly degraded meadow, soil organic carbon (SOC) stocks in moderately, heavily and extremely degraded meadow were reduced by 11.0%, 13.5% and 17.9%, respectively. Our field survey sampling design was overly intensive to estimate SOC status with a tolerable uncertainty of 10%. Power analysis showed that the optimal sampling density to achieve the desired accuracy would be 2, 3, 5 and 7 sites per 10 km2 for lightly, moderately, heavily and extremely degraded meadows, respectively. If a subsequent paired sampling design with the optimum sample size were performed, assuming stock change rates, predicted by experimental and modeling results, we estimate that about 5-10 years would be necessary to detect expected trends in SOC in the top 20 cm soil layer. Our results highlight the utility of conducting preliminary surveys to estimate the appropriate sampling density and avoid wasting resources due to over-sampling, and to estimate the sampling interval required to detect an expected sequestration rate. Future studies will be needed to evaluate spatial and temporal patterns of SOC variability. (C) 2016 Published by Elsevier Ltd.

【 授权许可】

Free