Ecological Indicators (Oct 2023)
Multifunctional trade-off/synergy relationship of cultivated land in Guangdong: A long time series analysis from 2010 to 2030
Abstract
With the development of human society, the cultivated land resources' utilization has become more diversified, and the multifunction of cultivated land is critical to meet the human demand. However, the studies on the multifunction of cultivated land are mostly based on the data of single time node or discontinuous years, which the results may be influenced by abnormal function years. To assess the multifunction of cultivated land and its trade-off/synergy relationships accurately and objectively, this study uses continuous long time series data to analyze multifunctionality of cultivated land. Firstly, the Long short-term memory (LSTM) method was applied to predict the data of 2021–2030 based on the data from 2010 to 2020. Then, the Theil-Sen Median method was used to analyze the spatiotemporal evolution trends of the multifunction of cultivated land from 2010 to 2030. Secondly, based on the trends, the long time series cultivated land function trade-off/synergy degree (LCFTD) was constructed to explore the stable trade-off/synergy relationships among functions. Finally, the cultivated land resources were divided into zones based on the spectral clustering. This study analyzed the multifunction of cultivated land in Guangdong from 2010 to 2030. (1) The multifunction of cultivated land had spatially heterogeneous and temporally fluctuating. (2) The production-ecological, ecological-life, and production-life functions were all dominated by synergy during 2010–2020. Without changing the current utilization state of cultivated land resource, the trade-off among functions will increase in the future (2021–2030). (3) The cultivated land was divided into five type zones: stable development, production core, ecological conservation, comprehensive lifting, and fallow renewal. The LCFTD can better avoid the influence of abnormal function years, and the long time series data can also predict the future development trend reliably. This study idea can help to formulate long time cultivated land management policies and achieve the goal of sustainable development of cultivated land resources.