Response of wheat and maize growth-yields to meteorological and agricultural droughts based on standardized precipitation evapotranspiration indexes and soil moisture deficit indexes

Abstract

Drought is a natural hazard that may decrease agricultural production. To investigate crop growth and yield responses to drought conditions are vital for drought prevention during crop growth periods. This study aims to analyze the impacts of meteorological and agricultural droughts on wheat/maize yields from multiple perspectives and to select the key parameters which describe the best relationship between crop yield and drought indices. Using standardized precipitation evapotranspiration index (SPEI) and soil moisture deficit index (SMDI) at 1- to 9-month timescales, the drought characteristics of different crop growth periods at the selected 98 sites in different subregions were analyzed. DSSAT-CERES-Wheat/Maize models were used to simulate the leaf area index (LAI), biomass and yield of spring wheat, spring and summer maize over 1961 − 2018. The relationships between yield related factors and SPEI/SMDI0–10 at different timescales were investigated using Pearson correlation. The key timescale and growth period which showed the best correlations between crop yield/growth and SPEI/SMDI were determined and used to obtain the yield/growth equations using multivariable linear regression. The results showed that: (1) The temporal variations of SPEI and SMDI0–10 differed with different timescales, months and subregions. DSSAT-CERES generally performed well in simulating growth and yields of wheat and maize over 1961–2018. (2) For spring wheat, the correlations of yield and SMDI were highest at 3-month timescale in July, at 5-month timescale in July and at 3-month timescale in June in subregions I, II and IV, respectively. For spring maize, in subregion I, yield correlated with 1-month SPEI in June best), while yield was correlated largest with 4-month SPEI in August in subregion III and IV. For summer maize, the best correlations occurred in August between yield and 4-month SPEI. Therefore, different crop had varying key parameters for drought prevention measures. (3) The multivariable linear equations described yield/growth vs. drought indices relationship well for different crops. The results are referable for providing measures for agricultural production practice under drought.