Various maize yield losses and their dynamics triggered by drought thresholds based on Copula-Bayesian conditional probabilities

https://doi.org/10.1016/j.agwat.2021.107391Get rights and content

Highlights

  • The vulnerability of maize yield to various drought stresses was assessed.

  • Drought thresholds triggering various maize yield loss were quantified.

  • The average drought thresholds resulting in 30%, 40% and 50% maize losses were −1.06, −1.53 and −2.23.

  • Increasing drought thresholds were mainly caused by precipitation and ET dynam-ics.

Abstract

In the context of global warming and human interventions, the impact of climate on crop yield may change over time. Therefore, assessing the dynamics of drought thresholds that trigger various maize yield losses is critical for food security under climate change. To this end, this study aims to assess the vulnerability of maize to drought stress in three provinces of Northeast China (Heilongjiang, Jilin and Liaoning provinces) and to quantify the drought thresholds that cause different levels of maize yield loss. A Copula-Bayesian conditional probability bivariate model is constructed to combine drought conditions and maize yield. Results indicate that: (1) for the whole study period 1980–2018, drought thresholds that induced different levels of maize yield reduction were significantly different in the three northeastern provinces of China; on average, the drought thresholds that induced 30%, 40% and 50% maize yield losses were −1.06, −1.53 and −2.23 in the three provinces of Northeast China; (2) during the transition from mild to moderate and severe drought, maize vulnerability in Liaoning province gradually exceeded that of Jilin and Heilongjiang province; (3) from 1980 to 1999–2000–2018, the drought thresholds that triggered the same percentage of maize yield reduction increased in all three provinces, suggesting a dramatically increasing trend in the vulnerability of maize yields to drought; (4) the changes in precipitation and evapotranspiration leading to increased drought severity were the main factors inducing drought threshold dynamics in both sub-study periods. The probabilistic assessment of the impact of drought on maize yield is expected to provide useful insights into the mitigation of drought and its effects under climate change.

Introduction

Global climate change has become one of the most studied topics in recent decades due to its adverse impacts on social economy, environment and biology (IPCC, 2014). Climate change is increasing the frequency, severity and duration of droughts in many regions of the world (Liu et al., 2020, Peng et al., 2020, Vogt et al., 2021). Over the past few decades, droughts have caused significant impacts on the agricultural sector, leading to food insecurity especially in developing countries (Chi et al., 2020, Randell et al., 2021). Despite significant research breakthroughs in the technology and crop yield potential, food production and security remain highly dependent on weather and climate change, as temperature and precipitation are key drivers of crop growth (Petritsch and Hasenauer, 2014; Zou et al., 2019). Thus, studying the response of crop yield to climate change is essential to ensure food security (Leng and Hall, 2019, Feng et al., 2019).

As a result, food security issues caused by extreme weather events, such as drought, have generated research and public interest in climate change analysis (Hameed et al., 2019, Guo et al., 2019, Fang et al., 2019, Cui et al., 2019, Zhu et al., 2021, Zhou et al., 2021). A number of studies have explored the effect of drought on crop yield (Liu et al., 2018, Marina et al., 2019, Leng and Hall, 2019; Venkatappa et al., 2021). The mechanism of crop yield reduction caused by drought has been explored based on the knowledge of crop physiology (Shah et al., 2017; Feng et al., 2019; Leng and Hall, 2019). Temperature anomalies were found to cause crop yield changes during the growing season (Deryng et al., 2014). High temperature can reduce production by shortening the growing season and destroying plant cells (Zampieri et al., 2016, Zampieri et al., 2018).

In order to detect and quantify droughts, the standardized precipitation evapotranspiration index (SPEI) was proposed by Vicente-Serrano et al. (2010) to account for the effects of both temperature and precipitation. SPEI has been proved to be an effective drought indicator and been widely used in a range of studies (e.g., Liu et al., 2018). Maize is the crop with the largest planting area and yield in China (Tian et al., 2019). The three provinces of Northeast China (Heilongjiang, Jilin and Liaoning provinces) are an important commodity grain production base, in which maize is one of the most important crops. Therefore, the present study focuses on the investigation of drought impacts on maize yield in these three provinces.

The GAR Special Report on Drought 2021 states that while droughts may develop gradually over several months, they can also trigger sudden and dramatic famines and subsequent food riots when a social tipping point is crossed (Vogt et al., 2021). Tipping points occur in nonlinear dynamic systems and the effects of droughts on the food system are highly nonlinear (Vogt et al., 2021). It is critical to assess the tipping point of drought impacts on food systems, namely to quantify drought thresholds that induce grain yield reduction. Some studies have analyzed drought thresholds and their effects on crops based on field experiments (Liu et al., 2021). However, few studies have explored and quantified, through probabilistic statistics, the drought thresholds that can cause different levels of food yield reduction at the provincial level. Probabilistic methods for comprehensive assessment of drought impacts on crop production are helpful to capture the multivariate characteristics of agricultural drought risk (Ribeiro et al., 2019). It has been widely recognized that copula functions are becoming quite popular among multivariate analysis approaches (e.g. Lee et al., 2013; Li et al., 2015; Ribeiro et al., 2019). The application of Copula theory in agrometeorological research is also relatively new. Madadgar et al. (2017) and Bokusheva et al. (2016) have recently modeled the joint distribution of agricultural crops and drought conditions with copula-model. The advantage of using Copula in multivariate modeling is that the probability distributions of individual variables are not necessarily the same (Nelsen, 1999; Ribeiro et al., 2019). Existing studies have assessed grain yield reduction under different drought levels (e.g. mild, moderate, severe, and extreme drought), but did not accurately quantify the drought thresholds that can cause dynamic food production losses. The impact of climate on yields may vary over time due to global warming (and related extremes) and changes in agricultural practices (Trnka et al., 2016, Zipper et al., 2016, Feng et al., 2021). Accurate estimates of drought thresholds triggering corresponding food yield losses and their dynamics over different periods can help individuals and societies to appropriately predict these impacts and adjust agriculture practices to maximize agricultural yields.

Therefore, the purpose of this study is to assess the probability response of maize yield to drought conditions based on a copula-Bayesian framework in three provinces of Northeast China. A model of multivariate distribution is employed to jointly study the behavior of drought and maize yields. The conditional probability distribution framework for maize yield loss under different drought conditions was then used to quantify the drought thresholds that induce different yield losses. The main objectives of this study are: (1) to evaluate the vulnerability of maize yield in different provinces under drought stress at different periods; (2) to quantify the drought thresholds inducing maize yield loss and their dynamics in different provinces.

Section snippets

Study area and data

This study selects three provinces in Northeast China (118°50′−135°20′E, 38°43′−53°33′N) as the research area, which includes Heilongjiang, Jilin and Liaoning provinces covering an area of 7.87 × 105 km2 with complex topography of plains and mountains (Fig. 1). Located in the northernmost part of China, this region is the country's largest granary, the largest grain output, and an important forestry base in China. The Changbai Mountains, the Greater Khingan Mountains and the Lesser Khingan

Drought index calculation

In terms of meteorological drought indices, the Standardized Precipitation Evapotranspiration Index (SPEI) at growing season (from May to September, average growing season of maize in the three northeastern provinces) at scales of 5 (SPEI5) during 1980–2018 is used in this study. The detailed steps for the calculation of the SPEI are as follows: first, the monthly potential evapotranspiration were estimated based on the climate data (precipitation and temperature) using the method of

Impact of drought on maize yield during the growing season

Fig. 3A shows the variation of maize yield per unit area in the three Provinces of Northeast China during 1980–2018. In general, during this period of time, strong human interventions, such as the improvement of agricultural practice, investment and technological progress, have led to a steady increase in maize yield. Nevertheless, the trend of sharp decline, and significant decline in maize yield were found during severe drought years (as shown in the gray vertical shadow in Fig. 3). Fig. 3B

Changes of meteorological factors and irrigation conditions in three northeastern provinces

To understand the reasons for the different maize yield losses under the influence of drought in different provinces of the Northeast China, precipitation (P), evapotranspiration (E), the difference between precipitation and evapotranspiration (P-E) in different periods, and the aridity index (AI). AI is defined by the ratio of precipitation (P) to reference crop evapotranspiration (ET0) (UNEP, 1992), and is considered an important indicator of climate regionalization. It is widely used in

Conclusions

Based on the Copula-Bayesian Probability Assessment theory, this study developed a bivariate probabilistic framework, which facilitates the quantification of crop vulnerability to drought stress at multiple levels and the identification of drought-vulnerable farmland ecosystems in a more informative way. In the proposed framework, drought conditions are characterized by the SPEI, which can be flexibly replaced by any other yield-related climate variable as part of an effort to understand the

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This study was jointly funded by Supported by the Strategic Priority Research Program of the Chinese Academy of Sciences (XDA28060100), the National Key Research and Development Program of China (2017YFC0405900), and the National Natural Science Foundation of China (51709221).

References (54)

  • X. Huang et al.

    Assessment of the regional agricultural water-land Nexus in China: a green-blue water perspective

    Sci. Total. Environ.

    (2022)
  • G. Leng et al.

    Crop yield sensitivity of global major agricultural countries to droughts and the projected changes in the future

    Sci. Total. Environ.

    (2019)
  • X. Liu et al.

    Thresholds of key disaster-inducing factors and drought simulation in the Xilinguole Grassland

    Ecol. Inform.

    (2021)
  • X.F. Liu et al.

    Drought evolution and its impact on the crop yield in the North China Plain

    J. Hydrol.

    (2018)
  • Y. Liu et al.

    Flash droughts characterization over China: From a perspective of the rapid intensification rate

    Sci. Total. Environ.

    (2020)
  • H. Randell et al.

    Food insecurity and compound environmental shocks in Nepal: implications for a changing climate

    World Dev.

    (2021)
  • A. Ribeiro et al.

    Copula-based agricultural drought risk of rainfed cropping systems

    Agr. Water Manag.

    (2019)
  • L.X. Tian et al.

    Effects of waterlogging stress at different growth stages on the photosynthetic characteristics and grain yield of spring maize (Zea mays L.) Under field conditions

    Agr. Water Manag.

    (2019)
  • M. Venkatappa et al.

    Impacts of droughts and floods on croplands and crop production in Southeast Asia-An application of Google Earth Engine

    Sci. Total. Environ.

    (2021)
  • L. Wang et al.

    Vegetation vulnerability and resistance to hydrometeorological stresses in water- and energy-limited watersheds based on a Bayesian framework

    Catena

    (2021)
  • L. Yu et al.

    Effect of natural factors and management practices on agricultural water use efficiency under drought: a meta-analysis of global drylands

    J. Hydrol.

    (2021)
  • M. Zampieri et al.

    Global assessment of heat wave magnitudes from 1901 to 2010 and implications for the river discharge of the Alps

    Sci. Total. Environ.

    (2016)
  • H. Zhao et al.

    What were the changing trends of the seasonal and annual aridity indexes in northwestern China during 1961–2015?

    Atmos. Res.

    (2019)
  • Z. Zhao et al.

    Changes in spatiotemporal drought characteristics over northeast China from 1960 to 2018 based on the modified nested Copula model

    Sci. Total. Environ.

    (2020)
  • X. Zhu et al.

    Assessing the vulnerability and risk of maize to drought in China based on the AquaCrop model

    Agr. Syst.

    (2021)
  • S. Bhargava et al.

    Drought stress adaptation: metabolic adjustment and regulation of gene expression

    Plant Breed.

    (2013)
  • D. Deryng et al.

    Global crop yield response to extreme heat stress under multiple climate change futures

    Environ. Res. Lett.

    (2014)
  • Cited by (30)

    View all citing articles on Scopus
    View full text