New strategies for a reliable assessment of baking quality of wheat - Rethinking the current indicator protein content

New strategies for a reliable assessment of baking quality of wheat are required since the protein content is a widely discussed indicator in this respect. In this study, the indicator value of the protein content for baking quality was evaluated using almost 600 wheat samples representing different quality and varieties. NIRS calibrations were developed to estimate bread volume at different levels of processing. Bread volume increased with increasing protein content in a wide curvilinear relationship with an R2 of 0.59. At protein contents above 12% representing the relevant range for trading, the R2 dropped to 0.15. Different wheat varieties showed differential responses (i.e. linear and non-linear) in bread volume to increasing protein content. The performance of NIRS calibrations for bread volume were best at the processing level of dough (R2 0.81) and good at the level of whole grains (R2 0.70). Hence, the protein content is not a reliable indicator for baking quality on its own. Bread volume was predicted better by NIRS calibrations than by protein content. A strategy is presented paying attention to the variety specific interplay between baking quality and protein content with implications for the variety registration system, trading, breeding and nitrogen fertilisation.