1431

DETECTION OF GENETICALLY MODIFIED GRAINS BY NEAR-INFRARED SPECTROSCOPY CHARLES R. HURBURGH, JR., GLEN R. RIPPKE, CRAIG HEITHOFF, SYLVIE A. ROUSSEL, CONNIE L. HARDY, 1563 Food Sciences Building, Ames, Iowa 50011-1061. Email: [email protected] Genetic modifications (referred to as GMOs), such as insect or herbicide resistance, are causing increased consumer concern in international grain markets. Sales of guaranteed non-GMO grain are increasing, causing grain buyers to isolate certain grain varieties for these markets. Grain buyers need a method of product identification that is faster and simpler than chemical methods used by laboratories. The Iowa State University Grain Quality Laboratory has successfully identified Roundup Ready TM (RR) soybean samples (a GMO) with near-infrared spectroscopy (NIR) technology. Sample sets of 1998-crop RR and non-RR samples were used to develop discriminate analysis calibrations for three models of nearinfrared analyzers. Figure 1 shows the spectral differences between RR and non-RR samples. There was a general offset plus specific effects in wavelengths around 910 nm and around 1000 nm. These correspond to C-H overtones, which supports other published reports that the RR modifications may alter fiber structure. The average nutritional values (protein, oil, fiber, and saturated fats) were not significantly different between RR and non-RR samples. The Foss Grainspec TM and Foss/Tecator Infratec TM near-infrared transmittance instruments correctly identified 99.0% and 89.7% as being either RR or non-RR, respectively. These are the two models most often used by grain buyers. On an independent validation set of 20 non-RR and 19 RR samples, the Infratec instrument correctly identified 95% of the non-RR samples and 84% of the RR samples. The calibrations were field-tested in working Infratec units during harvest 1999. An update of the calibrations, using a much larger database and alternate mathematics, was done. Final statistics will be presented. 3.7-

3.5

3.2

3.1

3.0

.................................................

...................................

............

oo

WJ,~,'el~tth(am) Figure 1. Near-lnfnlrlid Triutlmllian¢.o Spectra of Roundup Ready and Non-Roundup Ready Soyburls