averaged, on a 130 g kg Ϫ1 seed moisture basis, a seed yield of 2401 kg ha Ϫ1 , a seed protein content of 354.1 g Soybean [Glycine max (L.) Merr.] seed protein is negatively correkg Ϫ1 , and a seed oil content of 185.6g kg Ϫ1 (Hurburgh, lated with seed oil and often with yield. Our goal was to examine the 2001). These statistics indicated that 67.25 kg of the crop basis for these correlations at a quantitative genetic locus (QTL) level.could be processed into 19.7 kg of 48% protein meal Seventy-six F 5 -derived recombinant inbred lines (RILs) from the matproduct and 4.90 kg of oil product.ing of the high-protein (480 g protein per kg seed) G. max accession PI 437088A with the high-yield cultivar Asgrow A3733 (420 g kg Ϫ1 Soybean protein and oil contents in various regions seed protein content) were evaluated in six irrigation treatments (i.e., of the USA can deviate significantly from the foregoing 100, 80, 60, 40, 20, and 0% replacement of weekly evapotranspiration national averages. Some of that geographic variability loss) of a two-replicate experiment conducted for 2 yr. The RILs were arises from meteorological events that, in any given genotyped with 329 random amplified polymorphic DNAs (RAPDs), season, randomly affect some regions but not others. simple sequence repeats (SSRs), and four other markers, creatingFor example, high temperatures during soybean seed a 2943-centimorgan (cM) genetic map of 35 linkage group (LGs) that, development can elevate seed oil (Howell and Cartter, on the basis of SSR homology, aligned with the 20 known soybean 1958), while severe drought can depress seed protein LGs. The phenotypic regression of RIL protein and oil on yield re-(Specht et al., 2001). Other geographic variability arises vealed respective linear coefficients of Ϫ2.6 and ϩ1.6 percentage from region-specific climatic parameters. Notably, seed points per kg ha Ϫ1 (a protein/oil exchange ratio of Ϫ1.6). A seed protein is typically lower in the northwestern than in protein, oil, and yield QTL mapped close to RAPD marker OPAW13a southeastern soybean-growing states. In 2001, that rein a small LG-I interval that was flanked by the SSR markers Satt496 gional difference spanned 3 percentage points, the and Satt239. The additive effects of the PI 437088A allele on seed protein, oil, and yield were a respective ϩ1.0 and Ϫ0.6 percentage largest ever observed in 17 yr of survey data (Hurburgh, points (a protein/oil ratio of Ϫ1.6) and Ϫ154 kg ha Ϫ1 . Given that the 2001). Processors may not be able to derive the valuable genetic-based protein/oil exchange ratio of 1.6 is smaller than the 2.0 48% soybean meal if the soybean seed has too low of a calorific-based oil/protein ratio, one might expect the remaining 0.4 protein content. To offset this geographical disadvantage, units of carbon and/or energy to be made available for other seed breeders developing high yielding cultivars adapted to dry matter. However, yield almost invariably falls when seed protein northern and western production regions must practice is genetically ...

Soybean [Glycine max (L.) Merr.] yield, when regressed on water needed to replenish 0 to 100% seasonal evapotranspiration (ET), generates an estimate of season‐specific water‐use efficiency (WUE). The impact of unpredictable water deficits might be lessened if high‐yielding genotypes had a smaller beta. Our objective was to determine the genetic basis of beta and carbon isotope discrimination (CID), a theorized indicator of transpiration efficiency (TE). A ‘Minsoy’ × ‘Noir 1’ population of 236 recombinant inbred lines (RILs), genotyped at 665 loci, was evaluated in six water treatments (100, 80, 60, 40, 20, and 0% ET) for 2 yr. Water stress was mild in 1994, but high temperatures and no rainfall in 1995 led to a drought so severe that the 100% ET treatment required 41 cm of irrigation. The 1995 yield‐to‐water regression was highly linear (28 kg ha−1 cm−1). Genotype × water (G × W) interaction was due to genotypic heterogeneity in beta The CID vs. beta correlation was low (r = 0.26), so selection for better leaf TE may not improve crop WUE. Selection of low beta (less sensitivity to drought) will be difficult, given the yield beta vs. yield correlation (r = 0.71). The major quantitative trait loci (QTL) for yield beta, yield, and CID were coincident with maturity and/or determinancy QTLs, except for a CID QTL in linkage group U09‐C2, but it had no effect on beta Genetic improvement of soybean yield performance under drought would be better achieved by coupling a high‐yield grand mean with a high‐ (not low‐) yield beta

Soybean improvement via plant breeding has been critical for the success of the crop. The objective of this study was to quantify genetic change in yield and other traits that occurred during the past 80 yr of North American soybean breeding in Maturity Groups (MGs) II, III, and IV. Historic sets of 60 MG II, 59 MG III, and 49 MG IV soybean cultivars, released from 1923 to 2008, were evaluated in field trials conducted in 17 U.S. states and one Canadian province during 2010 to 2011. Averaged over 27 MG II and MG IV and 26 MG III site-years of data, the estimated rates of yield improvement during the 80 yr were 23 kg ha -1 yr -1 for MGs II and III, and 20 kg ha -1 yr -1 for MG IV cultivars. However, a two-segment linear regression model provided a better fit to the data and indicated that the average current rate of genetic yield gain across MGs is 29 kg ha -1 yr -1 . Modern cultivars yielded more than old cultivars in all environments, but particularly in high-yielding environments. New cultivars in the historic sets used in this study are shorter in height, mature later, lodge less, and have seeds with less protein and greater oil concentration. Given that on-farm soybean yields in the United States are also increasing at a rate of 29 kg ha -1 yr -1 , it can be inferred that continual release of greater-yielding cultivars has been a substantive driver of the U.S. onfarm realized yield increases.

BackgroundAdvances in genotyping technology, such as genotyping by sequencing (GBS), are making genomic prediction more attractive to reduce breeding cycle times and costs associated with phenotyping. Genomic prediction and selection has been studied in several crop species, but no reports exist in soybean. The objectives of this study were (i) evaluate prospects for genomic selection using GBS in a typical soybean breeding program and (ii) evaluate the effect of GBS marker selection and imputation on genomic prediction accuracy. To achieve these objectives, a set of soybean lines sampled from the University of Nebraska Soybean Breeding Program were genotyped using GBS and evaluated for yield and other agronomic traits at multiple Nebraska locations.ResultsGenotyping by sequencing scored 16,502 single nucleotide polymorphisms (SNPs) with minor-allele frequency (MAF) > 0.05 and percentage of missing values ≤ 5% on 301 elite soybean breeding lines. When SNPs with up to 80% missing values were included, 52,349 SNPs were scored. Prediction accuracy for grain yield, assessed using cross validation, was estimated to be 0.64, indicating good potential for using genomic selection for grain yield in soybean. Filtering SNPs based on missing data percentage had little to no effect on prediction accuracy, especially when random forest imputation was used to impute missing values. The highest accuracies were observed when random forest imputation was used on all SNPs, but differences were not significant. A standard additive G-BLUP model was robust; modeling additive-by-additive epistasis did not provide any improvement in prediction accuracy. The effect of training population size on accuracy began to plateau around 100, but accuracy steadily climbed until the largest possible size was used in this analysis. Including only SNPs with MAF > 0.30 provided higher accuracies when training populations were smaller.ConclusionsUsing GBS for genomic prediction in soybean holds good potential to expedite genetic gain. Our results suggest that standard additive G-BLUP models can be used on unfiltered, imputed GBS data without loss in accuracy.

Soybean [Glycine max (L.) Merr.] seeds contain high levels of protein and oil useful for human consumption. Increasing emphasis in breeding programs to produce soybeans with specific protein or oil content for specialty markets demands that more efficient manipulation of these traits be achieved. The objective of this study was to evaluate eight different soybean populations from the midwestern USA for genetic markers linked to seed protein and oil content. The populations were derived from the breeding programs at the Univ. of Minnesota, the Univ. of Nebraska, and Purdue Univ.‐USDA‐ARS. Each population consisted of between 69 and 100 individuals and was mapped with 21 to 85 restriction fragment length polymorphism markers. The F2‐derived populations were grown in field tests in 1992, 1993, and 1994 in the state in which they originated. Single factor analysis of variance was used to detect significant associations between markers and traits. Environmentally stable and environmentally sensitive quantitative trait loci (QTL) were identified for both protein and oil contents in all eight populations. The identified QTL were sensitive to both environment and genetic background although some common QTL were identified in multiple populations across several years. The results show that a number of QTL affect these traits and that markers could potentially be used in breeding programs designed to alter the seed protein and oil content.

The sensitivity of soybean [Glycine max (L.) Merr.] main stem node accrual to ambient temperature has been documented in greenhouse‐grown plants but not with field‐grown plants in the north‐central United States. Biweekly V‐node and R‐stage, stem node number, internode length, and other traits were quantified in an irrigated split‐plot, four‐replicate, randomized complete block experiment conducted in Lincoln, NE, in 2003–2004. Main plots were early‐, mid‐, late‐May, and mid‐June sowing dates. Subplots were 14 cultivars of maturity groups 3.0 to 3.9. Node appearance was surprisingly linear from V1 to R5, despite the large increase in daily temperature from early May (10–15°C) to July (20–25°C). The 2003 and 2004 May planting date regressions exhibited near‐identical slopes of 0.27 node d−1 (i.e., one node every 3.7 d). Cold‐induced delays in germination and emergence did delay the V1 date (relative to planting date), so the primary effect of temperature was the V1 start date of linearity in node appearance. With one exception, earlier sowings led to more nodes (earlier V1 start dates) but also resulted in shorter internodes at nodes 3 to 9 (cooler coincident temperatures), thereby generating a curved response of plant height to delayed plantings. Delaying planting after 1 May led to significant linear seed yield declines of 17 kg ha−1 d−1 in 2003 and 43 kg ha−1 d−1 in 2004, denoting the importance of early planting for capturing the yield potential available in soybean production, when moisture supply is not limiting.

Sclerotinia stem rot [caused by Sclerotinia sclerotiorum (Lib.) de Bary] is considered the second most important cause of yield loss in soybean [Glycine max (L.) Merr]. Soybean cultivars show variability in susceptibility, but no complete resistance to the disease has been reported and little information on the genetics of resistance is available. The objective of this study was to identify putative quantitative trait loci (QTLs) associated with Sclerotinia stem rot resistance in soybean. Recombinant inbred lines (RILs) from five populations were developed by crossing Williams 82, a susceptible cultivar, with five cultivars exhibiting partial resistance: Corsoy 79, Dassel, DSR173, S19‐90, and Vinton 81. The F2 to F5 generations were advanced by single seed descent. Parental polymorphism was tested with 507 simple sequence repeat (SSR) primers from the integrated linkage map of soybean, and primers were selected for progeny screening in the five populations on the basis of polymorphism and distribution in the genome. Five hundred RILs, consisting of 100 F5:6 lines from each population, were evaluated for resistance to Sclerotinia sclerotiorum isolate 143 by a detached leaf method in the laboratory to measure lesion area on leaves inoculated with mycelium plugs. Single degree‐of‐freedom contrasts in PROC MIXED and interval analysis were used to detect putative QTLs. Twenty‐eight putative QTLs for resistance to Sclerotinia stem rot of soybeans were identified on 15 different linkage groups in five RIL populations by single degree‐of‐freedom contrasts. Alleles involved in reduction of lesion size came from both the resistant and susceptible parents, and transgressive segregates were identified in two populations. The amount of phenotypic variation explained by individual QTLs ranged from 4 to 10%. Seven QTLs on seven different linkage groups were identified in multiple populations with some QTL regions corresponding with mapped resistance genes and resistance gene analogs. The results suggest that several genes control resistance to Sclerotinia stem rot and that markers could facilitate an initial screen of segregating breeding populations.

No abstract