Search

Abstract Phosphorus (P) is an indispensable resource for plant growth, and under the background with P consumption increase in crop system, the agriculture and crop production may be affected through the changes of phosphorus use efficiency (PUE) due to the human interference and climate change. Therefore, it is of great importance to investigate the responses of PUE to human interference and climate change from the perspectives of historical simulation and future projection. In this study, the changes of PUE from 2001 to 2015 in the middle and lower reaches of the Yangtze River (MLRYR) of China were evaluated by means of empirical PUE-model based on the substance flow analysis (SFA) method. Their responses to future human interference and climate scenarios during 2016–2030 were investigated by driving PUE-model with three neural networks (BP, GA-BP and PSO-BP) as well as the multiple downscaling climatic data (HadGEM2-ES, BCC-CSM1.1 (m) and GFDL-ESM2M) with the help of a statistical downscaling method (SDSM). The results showed that the PUE values of the three crops (rice, wheat and maize) were 53.61%, 36.22%, and 32.56%, respectively, during the period from 2001 to 2015 and 54.11%, 38.92% and 32.78%, respectively, during the period from 2016 to 2030. Consequently, the PUE showed an increasing trend from history to future. However, the overall PUE level was not high, with a value of not more than 60%, and notably for wheat and maize, the PUE was less than 40%. Meanwhile, the precipitation affected the leaching of P because increasing in precipitation may result in a relatively high increase in the probability of low-PUE. Moreover, great potential exists for enhancing the PUE in this region, and measures to improve PUE management practices have been proposed, such as, raising the crop yield, keeping the arable area and balancing the fertilization, etc. The present study provides a beneficial reference to comprehensively understand the impacts of human interference and climate change on the PUE and further improves regional sustainable strategies for future P resource management. Highlights • Evaluating the responses of PUE to human interference and climate change from historical simulation and future projections. • Predicting the future PUE by driving PUE-model with the multiple neural networks and GCMs. • Besides the rice, the PUE values of wheat and maize were less than 40% during 2001–2030. [ABSTRACT FROM AUTHOR]