Impacts of urbanization on soil organic carbon stocks in the northeast coastal agricultural areas of China

Shuai Wang; Kabindra Adhikari; Qianlai Zhuang; Hanlong Gu; Xinxin Jin

doi:10.1016/j.scitotenv.2020.137814

Impacts of urbanization on soil organic carbon stocks in the northeast coastal agricultural areas of China

Sci Total Environ. 2020 Jun 15:721:137814. doi: 10.1016/j.scitotenv.2020.137814. Epub 2020 Mar 9.

Authors

Shuai Wang¹, Kabindra Adhikari², Qianlai Zhuang³, Hanlong Gu⁴, Xinxin Jin⁵

Affiliations

¹ College of Land and Environment, Shenyang Agricultural University, Shenyang 110866, Liaoning Province, China; Key Laboratory of Ecosystem Network Observation and Modeling, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China; Department of Earth, Atmospheric, and Planetary Sciences, Purdue University, West Lafayette, IN, USA.
² USDA-ARS, Grassland, Soil and Water Research Laboratory, Temple, TX 76502, USA.
³ Department of Earth, Atmospheric, and Planetary Sciences, Purdue University, West Lafayette, IN, USA.
⁴ College of Land and Environment, Shenyang Agricultural University, Shenyang 110866, Liaoning Province, China. Electronic address: allenguhan@126.com.
⁵ College of Land and Environment, Shenyang Agricultural University, Shenyang 110866, Liaoning Province, China. Electronic address: jinxinxin0218@163.com.

PMID: 32197288
DOI: 10.1016/j.scitotenv.2020.137814

Abstract

Dynamic changes in soil organic carbon pools have significant impacts on regional and global carbon balance. Due to rapid development in urbanized areas, the land use changes dramatically, impacting soil organic carbon (SOC) stocks in topsoil. This study aimed to document the impacts of urbanization on SOC stocks in a rapidly urbanized area from northeastern China. A total of 12 auxiliary variables were as SOC predictors including elevation, slope aspect, slope gradient, topographic wetness index, Landsat TM band3, Landsat TM band4, Landsat TM5, and normalized difference vegetation index. Urban-specific variables including population (POP), gross domestic product (GDP), distance to the socio-economic center, and distance to the roads are also considered. A set of 523 (in 1990) and 847 (in 2015) top soil samples with SOC measurement were collected. Two random forest (RF) models, one with all auxiliary variables except urban-specific variable (MA) and the other with all auxiliary variables (MB) were used to map the spatial distribution of SOC stocks in the two periods. Ten-fold cross-validation was conducted to evaluate the performance of RF models. We find that the full auxiliary variables model had a better performance for the both periods. POP and GDP were key auxiliary variables affecting spatial variability of SOC stocks in 2015. Over a 25-year period, SOC stocks decreased from 2.77 ± 1.09 kg m^-2 to 2.16 ± 0.93 kg m^-2, resulting in 3.78 Tg SOC loss in this region. Rapid urbanization led to drastic land- use change, which was the main reason for the decrease of SOC stocks. Additionally, urban-specific variables should be used as the main auxiliary variables when predicting SOC stocks in the areas that experience rapid urbanization. We believe that accurate prediction and mapping of SOC stocks will help manage land use and facilitate soil quality assessment so as to increase soil carbon sequestration in the region.

Keywords: Auxiliary variables; Random forest; Soil organic carbon stocks; Spatial variability; Urbanization.