Evaluation of Near Infra‐red Reflectance Spectroscopy for Predicting Grain and Stover Quality Traits in Maize

Abstract

AbstractThe prediction of grain and stover quality parameters in maize {Zea mays L.) by near infra‐red reflectance spectroscopy (NIRS) was studied. A total of 110 grain and 135 stover samples originating from different genotypes and environments were assayed. Calibration equations for content of crude protein (CP), crude fat (CF), starch (ST), and water soluble carbohydrates (WSC) in grain were obtained by multiple linear regression of known manual values on NIRS data from the odd numbered samples. Calibrations for CP, acid detergent fibre (ADF), in vitro digestible organic matter according to the Tilley & Terry (IVDOM‐T & T) and the gas production (IVDOM‐Gp) method, respectively, and metabolizable energy (ME) in stover were developed analogously. Equations were validated with the evennumbered .samples and for ME additionally with the 1584 stover samples from an experiment with 66 F1 hybrids tested in six environments. The coefficients of multiple determination (R2) of the prediction equations ranged from 0.80 for IVDOM‐Gp and ME in stover to 0.94 for CP in grain. Standard errors of calibration (SEC) and prediction (SEP) were in most cases not higher than commonly reported for conventional manual assays. With regard to the correct ranking of hybrids, prediction equations for ME applied well to stover samples from other environments with one exception. We concluded that NIRS can evaluate the quality traits investigated to a similar degree to that of conventional methods of analysis. Since NIRS is simple and safe to operate and allows rapid screening of several quality traits simultaneously, it should be particularly attractive for breeding purposes.