Use of Stem Reserve Carbohydrates to Predict Yield in Wheat.

Soluble stem carbohydrates in wheat (Triticum aestivum L.) are an important component of drought response, the ability to yield well under low water. Previous research indicates they are a reserve that can be remobilized during periods of drought to assist in grain fill. However, their role in the physiological response of Pacific Northwest wheat germplasm to drought has not been assessed. Assaying for stem reserve carbohydrates is both time consuming and involves dangerous chemicals. Near infrared spectroscopy (NIR) is a fast and inexpensive alternative for indirect measurement of soluble carbohydrates, but it requires an accurate statistical model to calibrate the spectral data with laboratory measurements. A set of field experiments was conducted over two years at five sites to test differences in the stem reserve carbohydrate concentration of five spring wheat cultivars grown in varying intensities of drought. The stem carbohydrates were estimated using NIR spectroscopy calibrated to a lab-measured value. The purpose of this experiment was two-fold: (1) to build a reliable and robust calibration model between NIR spectra and lab-measured stem carbohydrates; and (2) to determine if stem carbohydrates can be used as a proxy for yield rankings in drought environments in the Pacific Northwest. A highly predictive model was created using best subset selection of multiple linear regression (r²= 0.95), but that model was less accurate in predicting soluble carbohydrates in separate, independent experiments (r²= 0.12-0.66). The cultivar rankings of soluble carbohydrates were consistent in the drier environments and matched the cultivar rankings for yield. These data provide preliminary evidence that soluble stem carbohydrates do play a role predicting the drought response of PNW wheat germplasm and can be accurately predicted by NIR spectroscopy, although more samples are needed to build robust calibration.