Dynamic response of agricultural productivity to landscape structure changes and its policy implications of Chinese farmland conservation

Abstract

Farmland landscape systems are crucial to sustainable agriculture, but they can readily be disturbed by non-agricultural activities. However, the conclusion that agricultural landscape system would affect farmland productivity remains uncertain and lacks quantitative evidence. To construct a healthy farmland ecosystem to avoid farmland fragmentation and support a sustainable agricultural production, we built a structural classification model for the farmland landscape system to analyse spatiotemporal evolutionary trends in China's farmland landscape system at the pixel level. Then, long-term grain yield monitored by stations and potential productivity of farmland calculated using a global agro-ecological zone (GAEZ) model were applied to explore the relationship between agriculture production and farmland landscape system variations. The results showed that China's farmland landscape tends to be fragmented. Model analyses showed that the fragmentation of farmland will cause an average 55.33 % cumulative decline in potential farmland productivity and that fragmentation of core farmland landscape will cause a 0.85 % decline in potential productivity. Based on these findings, we offer a farmland protection strategy based on the spatial optimum allocation of natural and human landscapes.