Grain protein concentration and gluten strength are important quality traits in durum wheat.  High grain protein concentration and strong gluten durum semolina produce pasta products with better nutritive value and superior end-use quality.  In general, a minimum semolina protein concentration is required to produce acceptable quality pasta products.  This is because semolina protein concentration alone can account for 30 to 40% of the variability in pasta cooking quality.  Durum cultivars with high protein produce macaroni, spaghetti, and other pasta products with greater cooking firmness and increased tolerance to overcooking.  Similarly, pasta cooking quality improves as protein concentration in the same cultivar increases.  Thus, many pasta manufacturers require semolina suppliers to meet a minimum semolina protein content.

Many studies have indicated that emmer wheat (T. turgidum  L. var. dicoccoides), the wild relative of cultivated durum wheat, contains a large number of accessions that have high grain protein concentrations.  Triticum dicoccoides accessions with grain protein concentrations of 200 to 240 g kg^-1 have been reported.  The T. dicoccoides gene pool represents a new source of genetic variability for grain protein concentration and may provide useful sources of high protein genes for introgression into durum and bread wheat breeding germplasms.

At NDSU’s durum wheat project, the modified pedigree method is used to develop germplasm with high protein concentration. ‘Maier’ (released in 1998) and ‘Dilse’ (released in 2002) are two durum wheat cultivars that were released from the program because of their high protein concentration and gluten strength. Protein concentration is evaluated at the F₅ and subsequent generations while gluten strength is evaluated, starting at the F₃ generation.  Whole grain protein is evaluated using a Technicon InfraAlyzer, where the targeted whole grain protein concentration is 13.5% or higher, at 12% mb.  Visual selection for gluten strength can be practiced starting at the F₂ generation.  Strong gluten is linked with the gene Rg1 for glume color in durum wheat. White glume color is associated with strong gluten, while buff or brown color is associated with weak gluten.  In the F₂ population, only white glume color plants are selected.  Gluten strength at the F₃ and subsequent generations is evaluated by the SDS micro-sedimentation test.  A sedimentation value below 30 mm indicates weak gluten, and a sedimentation value of 35 or higher indicates strong gluten. SDS micro-sedimentation test is effective in distinguishing weak from strong gluten, but is less effective in distinguishing strong from very strong gluten. F₅ and subsequent generations are evaluated for whole grain protein and for semolina protein.  Semolina protein is evaluated using a Technicon InfraAlyzer.  The target protein concentration for semolina is 12.5% or higher, at 14% mb.

Durum lines in the F₆ generation or higher are evaluated for whole grain protein and semolina protein.  Gluten strength is evaluated using SDS micro-sedimentation, mixograph, and wet gluten/gluten index tests. The mixograph procedure used is similar to AACC Method 54-40A (2000), with some modifications as follows.  Semolina (10 gm, 14% mb) was brought to constant water absorption of 58% and mixed for 8 min in the mixograph bowl (spring setting of 8).  Mixograph curves were compared to reference mixograms and scored.  Mixogram scores of 1-3 indicate a weak  gluten, 5-6 strong gluten, and 8 very strong gluten.  Wet gluten/gluten index is determined using AACC Method 38-12A (2000).   A gluten index <5% indicates weak gluten, 40-60% strong gluten, and > 80% very strong gluten.  There is a market for the three gluten strengths. Weak gluten is desirable for crimped or stamped pasta products, strong gluten is desirable for long goods, and very strong gluten is desirable for blending with lower quality semolina and may be advantageous in bread products.

Spaghetti made from durum lines in the F₉ generation or higher also are evaluated for cooking quality. Spaghetti, 10 g, is cooked in 300 ml boiling water for 12 min.  Cooked weight is determined as the weight of 10 g of dry spaghetti after cooking.  Cooking loss (weight of total solids expressed as a percent) is measured by evaporating the cooking water to dryness in a 110 C oven. Firmness of cooked spaghetti is measured using a TA-XT2 texture analyzer.  Cooked firmness is measured by the work required to shear five cooked spaghetti strands.