Registration of three soft red winter wheat germplasm lines with exceptional milling and cookie baking performance

1 School of Plant and Soil Environmental Sciences, Virginia Tech, Blacksburg, VA 24060, USA 2 Texas A&M AgriLife Research & Extension Center, Uvalde, TX 78801, USA 3 Eastern Virginia Agricultural Research and Extension Center, Virginia Tech, Warsaw, VA 22572, USA 4 Dep. of Crop Science, North Carolina State Univ., Raleigh, NC 27695, USA 5 USDA-ARS Soft Wheat Quality Laboratory, Wooster, OH 44691, USA 6 USDA-ARS Cereal Crops Research Unit, Fargo, ND 58102, USA


INTRODUCTION
End-use quality of soft red winter wheat (SRWW, Triticum aestivum, L.) germplasm is generally only considered by breeders once pure lines have been selected and tested in yield trials and sufficient grain is available for quality testing. Other than choosing parents and crosses to make, SRWW breeders focus most of their initial efforts on line selection and evaluation of pure lines for grain yield, test weight, and disease resistance, particularly Fusarium pressure required to crush the wheat kernel increases so does starch damage, which negatively influences cookie diameter. Kernels containing higher levels of protein are generally harder in texture and require more pressure to crush during milling, thus resulting in lower break flour extraction and small cookie diameters (Smith et al., 2011).
Breeding for end-use quality per se, solely on the basis of phenotypic data for multiple quality traits, is difficult due to the quantitative nature of these traits and lack of reliable high-throughput markers to aid in selection of parents and progeny on the basis of quality. Quantitative trait loci (QTL) for wheat milling and baking performance have been reported on all 21 chromosomes (Breseghello & Sorrells, 2006;Cabrera et al., 2015;Li, Song, Zhou, Branlard, & Jia, 2009;Souza et al., 2012). However, markers associated with most of these QTL are not routinely deployed in breeding programs, and thus genetic gains for improved quality remain stagnant (Kiszonas & Morris, 2017). A complementary approach to marker assisted or genomic selection is to introduce superior end-use quality transgressive segregates from breeding populations with the aid of predictive genetic markers to help facilitate genetic improvement of traits that influence milling and baking performance. The release of VA11DH-P46xTrib-28 (Reg. no. GP-1048, PI 691656), VA11DH-P46xTrib-99 (Reg. no. GP-1049, PI 691657), and VA11DH-P46xTrib-103 (Reg. no. GP-1050, PI 691658) germplasm lines is intended to provide SRWW breeders with breeding material that has superior end-use quality in an effort to help improve milling and baking performance in future SRWW cultivars.

Development of lines
These SRWW germplasm lines were identified as transgressive segregates in a doubled haploid (DH) population designed to map QTL associated with milling and baking quality performance. The doubled haploid lines were derived from the cross Pioneer '26R46' (PI 612154) by 'Tribute' (PI 632689) (Griffey et al., 2005

Field test evaluations
Complete management information is included in Table 3. Each DH line, the parents, and adapted checks were grown in yield plots comprised of seven rows spaced 15.24 cm apart with a total harvested area of 1.5 by 2.7 m. All seeds were coated with Gaucho-XT (Imidacloprid, Bayer Crop Science) insecticide and Raxil MD (tebuconazole and metalaxyl, Bayer Crop Science) fungicide seed treatments. Fall nutrient management and spring N applications were based on standard local management practices (Brann, Holshouser, & Mullins, 2000) and recommendations from the Virginia Cooperative Extension Soil Testing Service. Plots were managed for optimum quality in consideration of local conditions, including foliar fungicide applications (

End-use quality evaluations
All grain samples were thoroughly air aspirated prior to testing to remove any dust or chaff. Test weight (kg hl −1 ) of wheat grain was determined according to Method 55-10.01 (American Association of Cereal Chemists, 1999). Kernel weight (mg), hardness, and diameter (mm) were determined using the Single Kernel Characterization System (SKCS, Perten Instruments) according to Method 55-31.01 (American Association of Cereal Chemists, 1999) as the averages of 300 kernels. Wheat grain protein (g kg −1 ) and moisture (%) for tempering were estimated using a DA 7200 near-infrared analyzer (Perten Instruments). Wheat grain was tempered to 15% moisture for 24 h and milled using a modified Quadrumat Senior milling system to determine milling quality and to obtain flour. Flour yield (g kg −1 ) was the fraction of milled products that comprises the break and reduction flour after milling and calculated as [(grain weight − bran weight)/grain weight] × 100. Softness equivalence (g kg −1 ) was calculated as (break flour/total flour) × 100 (Finney & Andrews, 1986). Flour protein was estimated using a near-infrared analyzer (Unity Spectra-Star). Lactic acid and Na 2 CO 3 solvent retention capacities of the flour were determined TA B L E 1 Parental and doubled haploid (DH) line genotypes for major genes and quantitative trait loci associated with end-use quality in SRW wheat

Statistical analysis
An augmented single replicate test design was used, with eight checks including both parents replicated throughout the test at least twice; each site-year was considered a block. This was done because of limited seed sources, the cost of measuring end-use quality, and because milling and baking traits are highly heritable, with consistent results between site-years (Guttieri & Souza, 2003;Guttieri, Bowen, Gannon, O'Brien, & Souza, 2001;Souza et al., 2012). Therefore, numerous replicates are not needed to determine significant differences. The augmented complete block design (ACBD-R) with R for Windows Version 4.0 was used to analyze the three site-years individually (Table 4) and as multiyear environments (Table 2) (Rodriquez, Alvarado, Pacheco, & Burgueno, 2018). Best linear unbiased predictors (BLUP) of each DH line and the checks were generated along with least significant differences (p = .05) for use in line comparisons.

CHARACTERISTICS
Pioneer 26R46 is considered to have very good milling and baking quality. It was consistently the top-performing check or parent in this test. An individual line was considered to have exceptional milling and baking quality if it consistently performs similar to or better than Pioneer 26R46.

Milling and baking performance
Flour yield (g kg −1 ) and cookie diameter (cm) are two primary quality parameters considered and used routinely by breeders in identification of breeding lines with better end-use quality. Cookie diameter and flour yield are positively correlated, while both are negatively correlated with flour protein (g kg −1 ) and solvent retention capacity traits (Knott, Van Sanford, & Souza, 2009;Souza et al., 2012). This is due in part to the endosperm being released more easily from the bran during milling in wheat cultivars that have higher flour extraction. Starch granules suffer less damage during milling if the endosperm is more easily disintegrated into fine flour particles, as observed in wheat grain of low kernel hardness. Wheat cultivars having lower break flour extraction exhibit greater starch damage due to the extra shearing and crushing of endosperm particles during milling (Smith et al., 2011). Sodium carbonate solvent retention capacity (g kg −1 ) is a measure of damaged starch, while lactic acid solvent retention capacity (g kg −1 ) is a measure of gluten strength, both of which tend to increase as kernel hardness and protein increase. desire for pastry products (Gwirtz, Willyard, & McFall, 2007).   (Table 4). The multi-environment combined BLUP results indicate that all three DH lines are similar to Pioneer 26R46 for flour protein and both solvent retention capacities (lactic acid and NA 2 CO 3 ). Line VA11DH-P46xTrib-103 was higher (p < .05) than Pioneer 26R46 for flour yield and cookie diameter. Line VA11DH-P46xTrib-103 was higher (p < .05) for softness equivalence and cookie diameter than Pioneer 26R46.

Fusarium head blight
Data for reaction to FHB was previously collected on the three DH lines and both parents across five locations during 2013 and 2014: Arkansas (Fayetteville and Newport), Kentucky (Lexington), North Carolina (Kinston), and Virginia (Blacksburg). All three lines appear to fit within the normal distribution of progeny, being more susceptible than Tribute and less than or as susceptible as Pioneer 26R46 for FHB severity, Fusarium damage kernels, and deoxynivalenol (DON) content. The average FHB severity, Fusarium damage kernels, and DON concentration are shown in

Seed purification and increase
Sixty-four individual plants of each DH line were planted, vernalized for 7 wk, transplanted into larger pots, grown to maturity in a greenhouse, and harvested to produce a pure seed source. Variant or off-type plants were discarded prior to harvest, and seed from the remaining plants deemed to be uniform and true to type was bulked to form the breeder seed.

AVAILABILITY
Small seed quantities of all three lines will be made available upon request for breeding purposes and can be obtained from the Small Grains Breeding and Genetics group at Virginia Tech. Seed of lines VA11DH-P46xTrib-28, VA11DH-P46xTrib-99, and VA11DH-P46xTrib-103 has been deposited with the USDA National Plant Germplasm System, where it will be available 5 yr after publication of this article.

C O N F L I C T O F I N T E R E S T
The authors declare no conflict of interest.