Risk analysis of maize yield losses in mainland China at the county level

Xuan Li,Yingjie Wu,Shibo Fang,Yongchao Zhu,Dong Wu

doi:10.1038/s41598-020-67763-3

Abstract

Food security in China is under additional stress due to climate change. The risk analysis of maize yield losses is crucial for sustainable agricultural production and climate change impact assessment. It is difficult to quantify this risk because of the constraints on the high-resolution data available. Moreover, the current results lack spatial comparability due to the area effect. These challenges were addressed by using long-term county-level maize yield and planting area data from 1981 to 2010. We analyzed the spatial distribution of maize yield loss risks in mainland China. A new comprehensive yield loss risk index was established by combining the reduction rate, coefficient of variation, and probability of yield reduction after removing the area effect. A total of 823 counties were divided into areas of lowest, low, moderate, high, and highest risk. High risk in maize production occurred in Heilongjiang and Jilin Provinces, the eastern part of Inner Mongolia, the eastern part of Gansu-Xinjiang, west of the Loess Plateau, and the western part of the Xinjiang Uygur Autonomous Region. Most counties in Northeast China were at high risk, while the Loess Plateau, middle and lower reaches of the Yangtze River and Gansu-Xinjiang were at low risk.

Highlights

Estimated by the method proposed in IPCC 5th assessment reports, is determined by three indicators: exposure degree, sensitivity, and adaptive c apacity[24]
The planting scale of a county affects the yield loss risk result when field-based or farm-based observed yields are aggregated by county
This study aims to provide high-resolution information on the spatial distribution of yield loss risk based on a new comprehensive risk index, which was established by combining the reduction rate, coefficient of variation (CV), and probability of yield reduction after removing the area effect

Summary

Introduction

Estimated by the method proposed in IPCC 5th assessment reports, is determined by three indicators: exposure degree, sensitivity, and adaptive c apacity[24]. Assessments are often conducted at a coarse resolution, such as at the provincial or district level[6], or at a high resolution over small regions[27,28], which has significant limitations because of the difficulty in accessing yield data at high spatial resolution These studies focused on only the economic responses to sown area size or used the percentage of the affected areas to planting areas to show the degree of damage/exposure[24,28,29,30], without considering the area effects (see the paragraph for further explanation). The provinces in western China, such as Qinghai Province, the Tibet Autonomous Region, and the Xinjiang Uygur Autonomous Region (Xinjiang), have relatively small crop-planting areas in most counties[31], where the estimated yield variation or loss risk could be higher than those of the counties with large planting areas. The results may be crucial for agricultural decision-support systems and climate change assessments

Objectives

Findings

Discussion

Conclusion