Resources Science ›› 2016, Vol. 38 ›› Issue (4): 714-727.doi: 10.18402/resci.2016.04.13

• Orginal Article • Previous Articles     Next Articles

Drivers of regional agricultural land changes based on spatial autocorrelation in the Pearl River Delta,China

CAO Qiwen1(), WU Jiansheng1,2, TONG De1(), ZHANG Xiaona3, LU Zhiqiang1, SI Menglin1   

  1. 1. Key Laboratory of Urban Habitant Environment Science and Technology,School of Urban Planning and Design,Peking University,Shenzhen 518055,China
    2. Laboratory of Earth Surface Processes of Ministry of Education,College of Urban and Environmental Sciences,Peking University,Beijing 100871,China
    3. School of Management,Tianjin Polytechnic University,Tianjin 300387,China
  • Received:2015-08-17 Revised:2015-12-27 Online:2016-04-25 Published:2016-04-25


Land use/cover change(LUCC)is at the core of global change and agricultural land change is important in the study of LUCC. With socio-economic development,pressure on agri-cultural land protection has been increasing in China. To develop sustainable land use policy, we need to understand regional agricultural land changes and driving forces. Using land use monitoring data,basic geographic spatial data and statistical yearbooks,we analyzed characteristic of agricultural land change in the Pearl River Delta Area,China. Traditional Logistic modeling and AutoLogistic modeling which bring in spatial autocorrelation were compared to investigate the drivers of agricultural land change at regional scale from the perspectives of natural,socio-economic,spatial distance and spatial autocorrelation of land use characteristics. We found that cultivated land and forest in this region decreased from 2000 to 2010,leading to further fragmentation. Cultivated land has become a main source of expansion of construction land. And,both spatial autocorrelation of agricultural land property and land development intensity are important driving forces of agricultural land change. As for other driving forces,cultivated land change was mainly affected by factors such as ‘change of per kilometer GDP',‘distance to nearest railway',‘change of total population density',‘tendency rate of annual sunshine hours'. Forest changes were mainly affected by ‘change of per kilometer GDP',‘change of total population density',‘slope' and ‘distance to nearest road'. Compared with traditional Logistic modeling,the AutoLogistic model is more suited to study driving forces of regional agricultural land change.

Key words: agricultural land change, driving forces, spatial autocorrelation, Logistic model, Autologistic model, Pearl River Delta Area