Get access

Prediction of the dry-milling performance of maize hybrids through hardness-associated properties

Authors


Corresponding author Massimo Blandino,

Department of Agriculture, Forestry and Land Management,

University of Turin,

Via Leonardo da Vinci 44, 10095 Grugliasco (TO), Italy.

Tel: +39-011-6708895; fax +39-011-6708798.

E-mail address: massimo.blandino@unito.it

Abstract

Background

The hardness of kernels determines the dry-milling processing performance of maize hybrids. The identification of the best maize hybrids for the dry-milling process requires a limited number of simple, practical and reliable tests that are able to predict the potential grit yield.

Results

A total of 119 samples from different genetic and environmental backgrounds, collected over 3 years, were analysed for coarse/fine ratio (C/F), floating test (FLT), protein content (PC), kernel sphericity (S), total milling energy (TME) and test weight (TW). The total grit yield (TGY) of each sample was obtained through a micromilling procedure based on the manual separation of kernel endosperm followed by grinding and sieving under standard operational conditions. The TGY was used to establish the capability of the tests to predict the dry-milling aptitude. Single and multiple linear regression analyses were performed to establish equations for the prediction of TGY using C/F, FLT, PC, S, TME and TW as independent variables. The analyses were performed on three data sets, clustered year by year of the sample collection and analysis, and the resulting average coefficients of determination (R2) were compared by analysis of variance. C/F, FLT, TME and, to a lesser extent, TW appeared to be easy-to-use independent descriptors of maize dry-milling. An improved model prediction ability was observed when different combinations of a few physical and chemical properties were used as input variables. However, the models in which three or more variables were used did not lead to any significant improvement in TGY prediction compared with the smaller models.

Conclusion

This study contributes towards establishing the best predictor of maize kernel aptitude to dry-milling processes. Of all considered data sets, a milling evaluation factor (C/F or TME) coupled with kernel density (measured by means of the FLT) showed the best predictive ability for dry-milled product yields.

Ancillary