Search

Orange‐fleshed sweetpotato (OFSP) [Ipomoea batatas(L.) Lam] often lacks high dry matter (DM) in elite breeding material. The objectives were to examine the potential to combine yield with high root DM and high root β‐carotene (BC) content by variance‐covariance analysis and optimization of breeding scenarios for “OFSP dry and starchy,” which was exemplarily shown for the humid tropics of Peru. The trial series across seasons suggested a good chance to combine yield and quality (high DM and high BC). The unique data were used to estimate variance components and correlations among four yield traits and four quality traits. For the optimization of selections in our breeding program, we used the selection response by Cochran's approach and developed a small R script. Responses to selection were predicted for storage root yield and a modified Elston index aggregating eight traits. The differences between two‐ and three‐stage selection scenarios were small. For storage root yield, the recommended optimized two‐stage selection increased the response relative to the standard (one‐stage selection 100%) to 141.6% for 2268 plots and to 189.1% for 540 plots; the corresponding optimized three‐stage selection increased the response to 144.3% for 2268 plots and to 190.5% for 540 plots. A two‐stage selection is recommended with two to three locations at the first stage and 6–12 locations at the second stage, according to low‐ or high‐test capacity, respectively. Optimization of breeding schemes is much more important for smaller breeding programs than those with larger test capacity, and this allows testing even larger numbers of genotypes in a small breeding program. Variance–covariance estimates in elite orange fleshed sweetpotato (OFSP) breeding material are provided.There is opportunity in OFSP to combine high storage root yields, high root dry matter and high root β‐carotene contents.Optimized yield selection scenarios can serve as appropriate selection scenarios across yield and quality traits.Optimization is much more important for smaller breeding programs than those with larger test capacity.

Most plant‐breeding programs involve multistage selection, but hitherto no flexible and computationally efficient software has been freely available for calculating the selection gain and optimizing the allocation of resources under these conditions. Here we describe the use of the newly‐developed R package selectiongainthat enables (i) calculation of the selection gain for up to 20 selection stages and (ii) a grid search of the optimum allocation of resources (number of candidates, locations, and testers) for given costs and correlations among the selection criteria in each stage. The capabilities of our software are demonstrated with an example taken from the literature on selecting doubled haploid lines in maize across three stages.

A plethora of maize (Zea maysL.) landraces is stored in gene banks worldwide. However, information about their value in breeding is scarce and strategies for identifying the most promising landraces in prebreeding are largely lacking. This study was conducted to (i) evaluate the testcross performance of 70 European flint landraces in combination with two elite dent testers and compare these results with the performance of modern hybrids for important agronomic traits, (ii) estimate the genetic variances among landraces and trait correlations for these two testcross series as well as the correlation between them, and (iii) devise a testing scheme for assessing the breeding potential of a large number of landraces for hybrid breeding. Grain yield of the landrace testcrosses was on average about 26% lower than modern hybrids. Genotypic variances among landrace testcrosses were significant for all traits, and genetic correlations were moderate to high for most trait combinations in both testcross series. Thus, it seems promising to tap this huge genetic reserve for enlarging the genetic base of the elite flint germplasm pool in Central Europe. Since the genetic correlation between the two testcross series exceeded 0.74 for all traits, we recommend assessing the breeding potential of landraces for broadening existing heterotic groups by evaluating their testcross performance in combination with one or two elite single‐cross tester(s) from the opposite heterotic pool. Subsequently, doubled haploid lines from a few of the most promising landraces could be produced to exploit the large genetic variation within landraces to the full extent.

Favorable epistatic gene complexes may be important for hybrid performance of maize (Zea maysL.). This study was conducted to assess the importance of epistasis in per se and testcross performance for grain yield and grain moisture in four crosses among four elite European flint maize lines by generation means analyses as well as genome‐wide tests for significant digenic epistatic effects between marker loci. For each cross, six generations (P1, P2, F1, F2, BC1, BC2) and testcrosses of these generations plus the F2–Syn1, F2–Syn2, and F2–Syn3 generations in combination with an unrelated dent tester were evaluated in four environments. Testcross generation means of P, BC, F1, F2, F2–Syn1, F2–Syn2, and F2–Syn3 did not significantly differ from each other for grain yield and grain moisture, indicating that epistasis between unlinked and moderately linked loci was negligible in its net effect. Depending on the cross, QTL mapping for per se and testcross performance with the dent tester was conducted with 71 to 344 lines (F3to F6) grown in four environments. In genome‐wide two‐way ANOVAs, significant epistatic interactions were found with only a few marker pairs that did not improve the fit of the model after including main‐effect QTLs previously detected by composite interval mapping. Poor correspondence of the results from per se and testcross analyses reflects dominance and epistatic interactions between parental and tester alleles. Our results suggest that epistasis is of minor importance for both traits with regard to the optimum type of population (F2vs. BC) in recycling breeding of elite maize inbreds. Estimates of digenic epistasis detected with genome‐wide tests must be treated with caution because of the problems associated with model selection in QTL mapping with the sample sizes commonly used.

The magnitude of the genotypic correlation between line per se performance (LP) and testcross performance (TP) is crucial for optimum testing schemes in hybrid breeding as well as simultaneous improvement of commercial hybrids and their inbred parents. The objectives of this study were to (i) obtain estimates of the correlation, and (ii) determine quantitative trait loci (QTL) correspondence between LP and TP within four populations (F3to F6lines) derived from intrapool crosses of European flint maize (Zea maysL.). The number of lines evaluated for both LP and TP ranged from 65 to 280 across the four populations. The LP and TP with a dent inbred tester were evaluated for grain yield, grain moisture, kernel weight, protein concentration, and plant height in four to five environments. Composite interval mapping (CIM) using a joint restriction fragment length polymorphism (RFLP) linkage map of all populations was conducted separately for LP and TP in each population, with sample sizes ranging from 71 to 344. Genotypic correlations between LP and TP, r̂g(LP, TP), were low to intermediate for grain yield (0.28–0.56) across populations and intermediate to high for the other traits (0.52–0.87). The magnitude of r̂g(LP, TP) across populations for grain yield was neither associated with the ratio between the genotypic variances for LP and TP nor with the evidence for dominance in LP or epistasis in LP or TP. Genotypic correlations between observed TP and its prediction based on QTL positions and effects for LP were smaller than corresponding values of r̂g(LP, TP) for all traits. Except for grain yield, more than half of the QTL were in common to LP and TP in the largest population A × BIThus, it seems feasible to apply marker‐assisted selection for TP based on QTL detected for LP, for traits with a large proportion of the genotypic variance accounted for by QTL.

Congruency of estimated positions and effects of QTL in different samples of the same cross or different crosses is an indicator of the reliability of these estimates and their usefulness in marker‐assisted selection (MAS). We investigated the influence of the sample and genetic background on QTL congruency among five populations of European maize (Zea maysL.). Three samples derived from the same cross comprised 344 (A × BI) and 109 (A × BII) F2:3as well as 71 F4:5(A × BIII) lines. Two other crosses comprised 109 (A × C) and 84 (C × D) F3:4lines. All lines were topcrossed to the same inbred tester and evaluated in four or five environments. A combined linkage map of RFLP marker data from all five populations was used in composite interval mapping (CIM). The total number of QTL identified for five agronomically important traits was 42 in A × BI, 18 in A × BII, 20 in A × BIII, 28 in A × C, and 23 in C × D. Averaged across traits, the proportion pof the genetic variance explained by these QTL varied between 50.4% in the largest population A × BIand 30.7% in a population of considerably smaller size (A × BII). Cross validation (CV) yielded substantially lower estimates of pBetween 10 and 24% of the 42 QTL from A × BIwere also detected within a 20‐cM interval in the other four populations. Incongruent QTL among A × B samples were due to the low power of QTL detection and the large bias in QTL estimates. The genetic correlations between predicted (based on QTL positions from one population) and observed phenotypic values in another population were highest among A × B samples with a maximum of 0.68 for plant height. Congruency of QTL was found for kernel weight, protein concentration, and plant height and was mainly attributable to one or few QTL of moderate to large size. If more cost‐effective than phenotypic selection, MAS will be promising for these traits.

Dystrophic cardiac calcinosis (DCC) occurs among certain inbred strains of mice and involves necrosis and subsequent calcification as response of myocardial tissue to injury. Using a complete linkage map approach, we investigated the genetics of DCC in an F2intercross of resistant C57BL/6J and susceptible C3H/HeJ inbred strains and identified previously a major predisposing quantitative trait locus (QTL), Dyscalc1, on proximal chromosome 7. Analysis of inheritance suggested, however, that DCC is influenced by additional modifier QTL, which have as yet not been mapped. Here, we report the identification by composite interval mapping of the DCC loci Dyscalc2, Dyscalc3, and Dyscalc4on chromosomes 4, 12 and 14, respectively. Together, the four Dyscalcloci explained 47% of the phenotypic variance of DCC, which was induced by a high-fat diet. Additive epistasis between Dyscalc1andDyscalc2enhanced DCC. Examining recombinant inbred strains, we propose a 10-cM interval containing Dyscalc1and discuss potential candidate genes.

The efficiency of breeding programs could be increased by predicting the prospects of crosses for line development before producing and testing lines derived from them. In this study, we (i) assessed the level of genetic diversity among German and Austrian winter wheat (Triticum aestivumL.) cultivars using 117 restriction fragment length polymorphism (RFLP) probes, 16 amplified fragment length polymorphism (AFLP) primer combinations, and 21 simple sequence repeat (SSR) primer pairs, (ii) investigated the correlation between coancestry (f) and genetic similarity (GS) estimated from molecular markers, and (iii) evaluated the use of fand GSfor predicting the genetic variance (σ2g) within sets of 22 F4n(n= 7 or 8) lines derived from 30 crosses between these winter wheat cultivars. The average polymorphic information content (PIC) for polymorphic bands was not significantly different between the three marker systems (0.30 ≤ PIC ≤ 0.33), whereas the marker index was low for RFLPs and SSRs but high for AFLPs. Estimates of fbetween all 55 cultivar combinations ranged from 0.01 to 0.53. GS^‐values varied between 0.52 and 0.89 for RFLPs, between 0.40 and 0.83 for AFLPs, and between 0.16 and 0.91 for SSRs. The Mantel Ztest statistic revealed no common pattern between the four dendrograms obtained by cluster analyses. Significant (P< 0.05) correlations among f̂, GS^‐RFLP, GS^‐AFLP, and GS^‐SSR were detected only for related parent combinations (f̂≥ 0.10). For all seven traits analyzed, estimates of σ2gwere not significantly associated with any measure of GSbetween parents. On the basis of these results, we recommend AFLPs for fingerprinting wheat cultivars. However, predicting the progeny variance (σ2g) remains an unsolved problem.

In the cooler temperate regions of the world, maize (Zea maysL.) is primarily cultivated for forage usage. In this study, we mapped and characterized quantitative trait loci (QTL) affecting testcross performance of important forage maize traits and investigated their consistency across environments and testers. Two elite flint inbred lines were crossed to generate 380 F2individuals, 345 of which were genotyped at 89 RFLP (restriction fragment length polymorphism) marker loci. The 380 F3lines obtained by selfing the F2individuals and the parent lines were crossed to two diverse dent inbred testers. Each series of testcrosses (TC) was evaluated in field trials with two replicates in five environments. Five traits were analyzed: dry matter yield (DMY), dry matter concentration (DMC), plant height (PHT), starch concentration (STC), and starch yield (STY). Genotypic ( σ^g2) and genotype × environment variances were highly significant for all traits in both TC series. Heritabilities ranged from 0.64 to 0.88 except for STY (0.46, 0.47). Genotypic correlations between testers were high (>0.80) for all traits. QTL analyses were performed by composite interval mapping. Between seven (STY) and 16 (PHT) QTL were detected in each testcross series, explaining between 52 and 71% of all in a simultaneous fit. Few QTL displayed significant interactions with environments. Results were highly consistent across testers for DMC and PHT, but not for DMY, STC, and STY. Although a comparison of TC generation means indicated significant epistasis for DMY and PHT, significant interactions between QTL were detected in two cases only. If a genomic region simultaneously affected two characters, the increasing alleles for the individual traits originated in most cases from the same parent for positively correlated traits and from different parents for negatively correlated traits.