Whole crop maize yield modeling based on regional climatic data considering cultivar maturity grouping

Abstract

The sustainable supply of whole crop maize (WCM, Zea mays L.), as the domestic high-quality forage source, is causing great concern among the related parties in the Republic of Korea. Many new cultivars were introduced or developed in recent decades. This study was conducted to construct the WCM weather-crop yield prediction model considering cultivar maturity as well as to evaluate the effects of local climatic factors on yield. Data on the nationwide adaptability tests of WCM cultivars and the meteorological data were collected and merged into a dataset (n = 386, 22 years) after data cleansing. Three climatic variables, including the accumulation values of growing degree days, precipitation, and sunshine hours from seeding to harvesting, were generated. Then, the dataset was split into two sub datasets considering cultivar maturity. Subsequently, the models, including the three climatic variables and the cultivated location, were constructed for both sub datasets. The finesses and accuracy of the models were confirmed by residual diagnostics and 3-fold cross-validation. The accumulated temperature, sunshine time, and precipitation were found to significantly affect the WCM yield variance, while the precipitation factor caused stresses to the yield, which indicates water management is important for WCM cultivation.