Global Synthesis of Drought Effects on Maize and Wheat Production

PLoS One. 2016 May 25;11(5):e0156362. doi: 10.1371/journal.pone.0156362. eCollection 2016.

Abstract

Drought has been a major cause of agricultural disaster, yet how it affects the vulnerability of maize and wheat production in combination with several co-varying factors (i.e., phenological phases, agro-climatic regions, soil texture) remains unclear. Using a data synthesis approach, this study aims to better characterize the effects of those co-varying factors with drought and to provide critical information on minimizing yield loss. We collected data from peer-reviewed publications between 1980 and 2015 which examined maize and wheat yield responses to drought using field experiments. We performed unweighted analysis using the log response ratio to calculate the bootstrapped confidence limits of yield responses and calculated drought sensitivities with regards to those co-varying factors. Our results showed that yield reduction varied with species, with wheat having lower yield reduction (20.6%) compared to maize (39.3%) at approximately 40% water reduction. Maize was also more sensitive to drought than wheat, particularly during reproductive phase and equally sensitive in the dryland and non-dryland regions. While no yield difference was observed among regions or different soil texture, wheat cultivation in the dryland was more prone to yield loss than in the non-dryland region. Informed by these results, we discuss potential causes and possible approaches that may minimize drought impacts.

Publication types

  • Meta-Analysis

MeSH terms

  • Agriculture / methods
  • Droughts
  • Food Supply
  • Soil / chemistry
  • Stress, Physiological
  • Triticum / growth & development*
  • Zea mays / growth & development*

Substances

  • Soil

Grant support

This research was supported by a postdoctoral fellowship from Schlumberger Foundation (http://www.slb.com/about/foundation.aspx) for SD and USDA grant (2014-51130-22492) for LW and PAJ. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.