利用3个F_2群体整合高密度栽培种花生遗传连锁图

遗传图谱的构建及整合是开展花生分子育种研究的基础,利用多个作图群体整合遗传图谱是解决图谱标记密度低的有效途径。本研究采用基于锚定SSR标记的作图策略,构建3个F_2群体3张遗传连锁图,利用Join Map 3.0软件整合图谱,获得一张包含20个连锁群、792个位点、总遗传距离为2079.50 c M,标记间平均距离为2.63 c M的整合图谱,各连锁群标记数在20~66个之间,遗传距离在59.10~175.80 c M之间。将3个分离群体中检测到的与荚果及种子大小相关的QTL区段与整合连锁图的标记比较发现,各群体中检测到的位于各染色体上的QTL在整合图谱中都能出现,有些QTL标记区间在整合图谱中存在更多的标记,为今后利用这些标记进行精细定位奠定了基础。     

甘蔗SSR和AFLP分子遗传连锁图谱构建

采用甘蔗商业品种Co419与野生种割手密Y75/1/2杂交,获得269个单株,组成F1群体,用F102/356与商业品种ROC25回交获得266个单株,组成BC1群体。利用筛选的多态性条带丰富的36对SSR引物和12对AFLP引物,对两个群体进行PCR扩增和分子遗传连锁分析,构建甘蔗分子遗传连锁图谱。用F1群体获得630个分离标记,经χ2检测,298个标记为单双剂量标记,占总标记数的47%;用BC1群体获得571个分离标记,有264个标记为单双剂量标记,占总标记数的46%;4个亲本获得单双剂量标记的数量依次为Co41902/356Y75/1/2ROC25。在LOD≥5.0,相邻标记遗传距离≤40cM的条件下,F1群体有134个单双剂量标记被纳入55个连锁群,其中39个连锁群归属8个同源组,16个未列入,总遗传距离为1458.3cM,标记间平均图距为10.9cM;BC1群体有133个单双剂量标记被纳入47个连锁群,其中34个连锁群归属于8个同源组,13个连锁群未列入,总遗传距离为1059.6cM,标记间平均图距为8.0cM。从4个亲本单双剂量标记进入的连锁群数来看,Co419最多,归入34个连锁群,其次为Y75/1/2,归入20个连锁群,第3为02/356和ROC25,归入19个连锁群。研究结果表明,从单双剂量标记比例、形成连锁群数量、总遗传距离来看,F1群体构图质量要优于BC1群体。     

构建全基因组导入系中BC1F1群体大小的影响因素研究

全基因组导入系是遗传和育种研究的重要材料。导入系经受体亲本和供体亲本间连续杂交、回交构建而成, BC₁F₁群体大小是获得理想导入系群体的关键参数。然而, 各物种所需要的最小群体尚不清楚, 并且难以通过试验确定。本研究通过编写程序, 模拟减数分裂时的重组过程研究适宜的群体大小, 并通过数学运算和试验验证程序的可靠性。结果表明, 编程模拟与数学计算和试验结果一致。BC₁F₁群体大小与连锁群数目、连锁群长度和基因密度之间均为正相关。当模拟连锁群从5个增加到40个时, 群体大小需要由6.06增加到9.49; 当模拟连锁群长度从80 cM增加到200 cM时, 需要的群体大小从7.14增加到8.64; 当模拟基因密度从每基因20 cM缩小到每基因5 cM时, 群体大小从7.65增加到8.22。为测试该程序的应用范围, 对水稻、小麦、玉米、大豆等主要作物进行了BC₁F₁群体大小模拟,在保证95%的概率覆盖全基因组条件下, 水稻需要的群体最少, 为12个个体, 小麦和大豆均需13个个体, 玉米需要的个体数最多, 为14~15个。     

Quantitative trait loci for quality and agronomic traits in two advanced backcross populations in oat (Avena sativa L.)

Using the advanced backcross quantitative trait loci (AB‐QTL) strategy, we successfully transferred and mapped valuable allelic variants from the high β‐glucan (BG) accession IAH611 (PI 502955), into the genome of cultivar ‘Iltis’. By backcrossing one BC₁F₁ plant to ‘Iltis’, we developed two BC₂F_2‐6 populations A and B, comprising 98 and 72 F₂‐individuals, respectively. Genotyping of BC₂F₂ individuals with predominantly AFLP markers resulted in 12 linkage groups with a map size of 455.4 cM for Population A and 11 linkage groups with a map size of 313.5 cM for Population B. Both populations were grown at three sites in Germany over a three‐year period. Individuals were then phenotyped for 13 traits including grain yield (YD) and β‐glucan content (BG). QTL analysis via stepwise regression detected a total of 33 QTLs, most of which were clustered in three linkage groups. Two dense linkage groups A1 and B13 were found to be putatively homologous to groups KO_6 and KO_11 of the ‘Kanota’/‘Ogle’ map, respectively.     

Genetic mapping of AFLP markers in Japanese bunching onion (<Emphasis Type="Italic">Allium fistulosum</Emphasis>)

Summary The first genetic linkage map of Japanese bunching onion (Allium fistulosum) based primarily on AFLP markers was constructed using reciprocally backcrossed progenies. They were 120 plants each of (P₁)BC₁ and (P₂)BC₁ populations derived from a cross between single plants of two inbred lines: D1s-15s-22 (P₁) and J1s-14s-20 (P₂). Based on the (P₂)BC₁ population, a linkage map of P₁ was constructed. It comprises 164 markers – 149 amplified fragment length polymorphisms (AFLPs), 2 cleaved amplified polymorphic sequences (CAPSs), and 12 simple sequence repeats (SSRs) from Japanese bunching onion, and 1 SSR from bulb onion (A. cepa) – on 15 linkage groups covering 947 centiMorgans (cM). The linkage map of P₂ was constructed with the (P₁)BC₁ population and composed of 120 loci – 105 AFLPs, 1 CAPS, and 13 SSRs developed from Japanese bunching onion and 1 SSR from bulb onion – on 14 linkage groups covering 775 cM. Both maps were not saturated but were considered to cover the majority of the genome. Nine linkage groups in P₂ map were connected with their counterparts in P₁ map using co-dominant anchor markers, 13 SSRs and 1 CAPS.     

Quantitative trait loci analysis in Theobroma cacao using molecular markers. Inheritance of polygenic resistance to Phytophthora palmivora in two related cacao populations

Two related segregating populations of Theobroma cacao L. were analysed for their resistance to Phytophthora palmivora. The first F₁ population was obtained by crossing two susceptible cacao clones of Catongo (a highly homozygous genotype) and Pound 12(a highly heterozygous genotype) and the second population was obtained by backcrossing a single F₁ tree with Catongo. The genetic maps obtained for each population were compared. The F₁ map includes 162 loci and the backcross has 140 loci. The two maps, F₁ and BC₁, exhibit high co-linear loci organization covering respectively, 772 and 944 cM.Phytophthora resistance was assessed by measuring the size increase of a lesion at five (DL5)and ten days (DL10) after pod inoculation. Six different QTL were detected in the F₁ and BC₁ populations. One QTL was found in both populations, and appeared to be a major component of disease resistance, and explaining nearly 48% of the phenotypic variance in the F₁ population. The absence of some yield QTL detection in the BC₁ in comparison with the F₁ population is due to the lack of transmission of the favouring alleles for these QTL from the single F₁ tree used for the backcross. The phenotypic variance explained by the action of the quantitative trait alleles indicated that genetic factors of both major and minor effects were involved in the control of the character studied. QTL conferring increased resistance to Phytophthorawere identified in both susceptible parents, suggesting the presence of transgressive traits and the possibility of selection in cacao. Pleiotropic and epistatic effects for the QTL were also detected. Finally, the use of marker assisted selection (MAS) in cacao breeding programs is discussed. This revised version was published online in July 2006 with corrections to the Cover Date.     

Different recombination frequencies in wheat doubled haploid populations obtained through maize pollination and anther culture

This study compared the meiotic recombination frequency between wheat doubled haploid (DH) populations obtained through two different methods, maize pollination (MP♀) and anther culture (AC♂). The comparison was based on a genetic linkage analysis, performed with DNA markers. Thirty-five polymorphic markers (15 SSR, 15 AFLP, 5 RAPD) were screened in MP♀ and AC♂ doubled haploids populations, derived from the same hybrid genotype (F₁ of ‘Eta’ × ‘Darkhan 15’). Nine linkage groups, comprising 35 loci (the MP♀ lines) and 31 loci (the AC♂ lines), were constructed. The linkage groups in both DH populations showed identical orders of markers, except for one group mapping to chromosome 6B. The MP♀ and AC♂ linkage maps differed significantly in recombination frequencies for corresponding intervals. In total, the AC♂ linkage map (495.5 cM) was 40.5% longer than the MP♀ map (352.8 cM), indicating a significantly higher meiotic recombination rate in pollen mother cells. The enhancement in recombination was visible in five of nine linkage groups, and in 7 intervals between individual loci out of 19 compared. Moreover, for 6 other intervals a lack of linkage was observed in the AC♂ population, as compared to the MP♀ map.     

The first genetic maps of cashew (Anacardium occidentale L.)

Cashew (Anacardium occidentale) is a widespread tropical tree crop that is grown primarily for its nuts and has a global production of over 2 million Mt. In spite of its economic importance to many countries, however, no linkage map containing STS anchor sites has yet been produced for this species. This is largely attributable to a prolonged juvenile phase of the tree (limiting mapping to F₁ progenies) and difficulty in effecting sufficient hand-pollinations to create mapping populations of effective size. Here, we produce an F₁ mapping population of 85 individuals from a cross between CP 1001 (dwarf commercial clone) and CP 96 (giant genotype), and use it to generate two linkage genetic maps comprising of 205 genetic markers (194 AFLP and 11 SSR markers). The female map (CP 1001) contains 122 markers over 19 linkage groups and the male map (CP 96) comprises 120 markers assembled over 23 linkage groups. The total map distance of the female map is 1050.7 cM representing around 68% genome coverage, whereas the male map spans 944.7 cM (64% coverage). The average map distance between markers is 8.6 cM in the female map and 7.9 cM in the male map. Homology between the two maps was established between 13 linkage groups of the female map and 14 of the male map using 46 bridging markers that include 11 SSR markers. These maps represent a platform from which to identify loci controlling economically important traits in this crop.     

Construction of a high-density reference linkage map of tea (Camellia sinensis)

A few linkage maps of tea have been constructed using pseudo-testcross theory based on dominant marker systems. However, dominant markers are not suitable as landmark markers across a wide range of materials. Therefore, we developed co-dominant SSR markers from genomic DNA and ESTs and constructed a reference map using these co-dominant markers as landmarks. A population of 54 F₁ clones derived from reciprocal crosses between ‘Sayamakaori’ and ‘Kana-Ck17’ was used for the linkage analysis. Maps of both parents were constructed from the F₁ population that was taken for BC₁ population. The order of most of the dominant markers in the parental maps was consistent. We constructed a core map by merging the linkage data for markers that detected polymorphisms in both parents. The core map contains 15 linkage groups, which corresponds to the basic chromosome number of tea. The total length of the core map is 1218 cM. Here, we present the reference map as a central core map sandwiched between the parental maps for each linkage group; the combined maps contain 441 SSRs, 7 CAPS, 2 STS and 674 RAPDs. This newly constructed linkage map can be used as a basic reference linkage map of tea.     

Combining different linkage maps in sugar beet (Beta vulgaris L.) to make one map

Several genetic maps for sugar beet (Beta vulgaris L.), from different German research groups, have been published and it is now possible to consider combining them with the aid of the common markers. The computer program JOINMAP (versions 1.3 and 2.0) was used for pair-wise combination of three populations. Several problems arose: the genetic background of the populations, different population structures (F₂ versus F₁× F₁), different number of polymorphic loci for common probes in the populations to be combined, different estimates of the recombination rates between the same markers and differences between the Join Map versions. The maps from two F₂ populations could be integrated into a single map, but it was more appropriate to construct separate maps for the F₂ populations and the F₁× F₁ population using common markers as reference points only.     

Construction of a genetic linkage map and QTL analysis of fiber-related traits in upland cotton (Gossypium hirsutum L.)

A genetic linkage map with 70 loci (55 SSR, 12 AFLP and 3 morphological loci) was constructed using 117 F₂ plants obtained from a cross between two upland cotton cultivars Yumian 1 and T586, which have relatively high levels of DNA marker polymorphism and differ remarkably in fiber-related traits. The linkage map comprised of 20 linkage groups, covering 525 cM with an average distance of 7.5 cM between two markers, or approximately 11.8% of the recombination length of the cotton genome. The present genetic linkage map was used to identify and map the quantitative trait loci (QTLs) affecting lint percentage and fiber quality traits in 117 F_2:3 family lines. Sixteen QTLs for lint percentage and fiber quality traits were identified in six linkage groups by multiple interval mapping: four QTLs for lint percentage, two QTLs for fiber 2.5% span length, three QTLs for fiber length uniformity, three QTLs for fiber strength, two QTLs for fiber elongation and two QTLs for micronaire reading. The QTL controlling fiber-related traits were mainly additive, and meanwhile including dominant and overdominant. Several QTLs affecting different fiber-related traits were detected within the same chromosome region, suggesting that genes controlling fiber traits may be linked or the result of pleiotropy.     

Cytogenetic study of Helianthus laevigatus and its F1 and BC1F1 hybrids with cultivated sunflower, Helianthus annuus

Although the wild sunflower species Helianthus laevigatus has not been extensively studied it may be considered for sunflower breeding as a potential source of desirable genes for Sclerotinia stalk rot resistance and high contents of proteins and linoleic acid in the seed. A set of six H. laevigatus populations was crossed to cultivated sun~ower lines and produced nine F₁ (2-14 plants) and 66 BC₁F₁ hybrid combinations (1-13 plants). Male sterility occurred in F₁ and BC₁F₁ hybrid combinations and pollen viability was lower in the progenies than in the parents (51.6-77.2%in F₁ and in F₁ and 4.8-34.0% in BC₁F₁). Meiosis was normal in the H. laevigatus populations It was found that this tetraploid species also occurred in a hexaploid form Numerous irregularities were observed in the meiosis of the F₁ interspecific hybrids During diakinesis, quadrivalents and hexavalents were recorded in addition to bivalents Dislocated chromosomes and chromosome bridges were present in the other phases The chromosome number in F₁ was 68 (tetraploid). Irregularities in chromosome pairing were observed in the interspecific hybrids at BC₁F₁. There were many univalents, and trivalents quadrivalents and hexavalents were also present The chromosome number in the BC₁F₁ generation ranged from 34 to 60. The occurrence of meiotic irregularities in the F₁ and BC₁F₁ interspecific hybrids indicates that H. laevigatus and the cultivated sunflower differ in genome constitution.     

A genetic linkage map of rhodesgrass based on an F1 pseudo-testcross population

The linkage relationships between 164 polymorphic amplified fragment length polymorphism (AFLP) and 25 restriction fragment length polymorphism (RFLP) fragments assayed in a pseudo‐testcross population generated from the mating of single genotypes from two divergent cultivars were used to construct female, ‘Katambora’ (‘Kat’) and male, ‘Tochirakukei’ (‘Toch’) parental genetic maps for rhodesgrass. The ‘Kat’ genetic map consists of 84 marker loci (72 AFLP and 12 RFLP markers) distributed on 14 linkage groups and spans a total length of 488.3 cM, with an average distance of 7.8 cM between adjacent markers. The ‘Toch’ genetic map consists of 61 marker loci (52 AFLP and nine RFLP) mapped on 12 linkage groups spanning a total length of 443.3 cM, with an average spacing of 9.0 cM between adjacent markers. About 23% of the markers remained unassigned. The level of segregation distortion observed in this cross was 11.1%. In both maps, linked duplicated RFLP loci were found. These linkage maps will serve as a starting point for linkage studies in rhodesgrass with potential application for marker‐assisted selection in breeding programmes.     

Mapping of a gene for leaf scald resistance in barley line 'B87/14' and validation of microsatellite and RFLP markers for marker-assisted selection

Seedlings of the barley line ‘B87/14’ were resistant to 22 out of 23 Australian isolates of Rhynchosporium secalis, the causal agent of leaf scald.‘B87/14’‐based populations were developed to determine the location of the resistance locus. Scald resistance segregated as a single dominant trait in BC₁F₂ and BC₁F₃ populations. Bulked segregant analysis identified amplified fragment length polymorphisms (AFLPs) with close linkage to the resistance locus. Fully mapped populations not segregating for scald resistance located these AFLP markers on chromosome 3H, possibly within the complex Rrs1 scald locus. Microsatellite and restriction fragment length polymorphism markers adjacent to the AFLP markers were identified and validated for their linkage to scald resistance in a second segregating population, with the closest marker 2.2 cM from the resistance locus. These markers can be used for selection of the Rrs.B87 scald‐resistance locus, and other genes at the chromosome 3H Rrs1 locus.     

A high-density integrative linkage map for <Emphasis Type="Italic">Gossypium hirsutum</Emphasis>

Sequence-related amplified polymorphism (SRAP) combined with SSRs, RAPDs, and RGAPs was used to construct a high density genetic map for a F₂ population derived from the cross DH962 (G. hirsutum accession) × Jimian5 (G. hirsutum cultivar). A total of 4,096 SRAP primer combinations, 6310 SSRs, 600 RAPDs, and 10 RGAPs produced 331, 156, 17 and 2 polymorphic loci, respectively. Among the 506 loci obtained, 471 loci (309 SRAPs, 144 SSRs, 16 RAPDs and 2 RGAPs) were assigned to 51 linkage groups. Of these, 29 linkage groups were assigned to corresponding chromosomes by SSR markers with known chromosome locations. The map covered 3070.2 cM with a mean density of 6.5 cM per locus. The segregation distortion in this population was 9.49%, and these distorted loci tend to cluster at the end of linkage groups or in minor clusters on linkage groups. The majority of SRAPs in this map provided an effective tool for map construction in G. hirsutum despite of its low polymorphism. This high-density linkage map will be useful for further genetic studies in Upland cotton, including mapping of loci controlling quantitative traits, and comparative and integrative analysis with other interspecific and intraspecific linkage maps in cotton.     

Molecular mapping and identification of QTL's associated to oat crown rust partial resistance

Summary Molecular mapping is a promising strategy for studying and understanding traits with complex genetic control, such as partial resistance to oat crown rust. The objectives of this research were to develop molecular maps from the progenies of the cross UFRGS7 (susceptible) × UFRGS910906 (partially resistant) and to identify QTLs (quantitative trait loci) associated to partial resistance to oat crown rust in two generations of that population.DNA of 86 genotypes of the F₂ and 90 genotypes of the F₆ UFRGS7 × UFRGS910906 population were used to generate AFLP markers. Molecular maps were constructed using Mapmaker Exp. 3.0 and QTLs for partial resistance to oat crown rust were identified with Mapmaker/QTL software. Five hundred and fifty seven markers in the F₂ and 243 markers in the F₆ generations were identified. The F₂ map integrated 250 markers in 37 linkage groups. The F₆ map integrated 86 markers in 17 linkage groups.Five QTLs were identified for partial resistance to oat crown rust in the F₂ generation and three QTLs in the F₆. The QTL identified on F₆ through the PaaaMctt340 AFLP marker showed consistency across two environments and two generations (F₄ and F₆), and appear to have potential for marker-assisted selection in oat.     

An SSR-based molecular genetic map of cassava

Summary Microsatellites or simple sequence repeats (SSR) are the markers of choice for molecular genetic mapping and marker-assisted selection in many crop species. A microsatellite-based linkage map of cassava was drawn using SSR markers and a F₂ population consisting of 268 individuals. The F₂ population was derived from selfing the genotype K150, an early yielding genotype from an F₁ progeny from a cross between two non-inbred elite cassava varieties, TMS 30572 and CM 2177-2 from IITA and CIAT respectively. A set of 472 SSR markers, previously developed from cassava genomic and cDNA libraries, were screened for polymorphism in K150 and its parents TMS 30572 and CM 2177-2. One hundred and twenty two polymorphic SSR markers were identified and utilized for linkage analysis. The map has 100 markers spanning 1236.7 cM, distributed on 22 linkage groups with an average marker distance of 17.92 cM. Marker density across the genome was uniform. This is the first SSR based linkage map of cassava and represents an important step towards quantitative trait loci mapping and genetic analysis of complex traits in M. esculenta species in national research program and other institutes with minimal laboratory facilities. SSR markers reduce the time and cost of mapping quantitative trait loci (QTL) controlling traits of agronomic interest, and are of potential use for marker-assisted selection (MAS).     

Construction of high-density linkage map and identification of QTLs associated with resistance to black rot in radish (Raphanus sativus) by RAD sequencing

High-density marker-based QTL mapping can serve as an effective strategy to identify novel genomic information to facilitate crop improvement. In this study, we genotyped an F₂ population (KB12-1 × PP12-1) using a RAD-seq approach and constructed a high-density linkage map for radish. After a series of filtering procedures were performed, 17,124 SNPs and 3,336 indels with aa × bb genotyping were retained to obtain bin markers. Then, a linkage map comprising a total of 1,221 bin markers in nine linkage groups spanning 1,467.3 cM with an average marker interval of 1.2 cM was constructed. We evaluated the resistance of the F₂ mapping population to black rot using F₃ progeny, and two major QTLs related to black rot resistance were identified based on this map. Among these QTLs, qBRR2 on Chr.2 explained 26.97% of the phenotypic variation with a LOD score of 11.93, and qBRR7 on Chr.7 accounted for 27.06% of the phenotypic variation with a LOD score of 11.83. The additive effect of qBRR2 was positive (14.97); however, qBRR7 had the opposite effect (−11.99). The high-density linkage map and the major QTLs qBRR2 and qBRR7 provide new important information for disease resistance gene discovery and utilization in genetic improvement.     

Molecular mapping of black rot resistance locus Xca1bo on chromosome 3 in Indian cauliflower (Brassica oleracea var. botrytis L.)

Black rot is the most devastating disease of cauliflower worldwide causing severe damage to crop. The identification of markers linked to loci that control resistance can facilitate selection of plants for breeding programmes. In the present investigation, F₂ population derived from a cross between ‘Pusa Himjyoti’, a susceptible genotype, and ‘BR‐161’, a resistant genotype, was phenotyped by artificial inoculation using Xcc race 1. Segregation analysis of F₂ progeny indicated that a single dominant locus governed resistance to Xcc race 1 in ‘BR‐161’. Bulk segregant analysis in resistant and susceptible bulks of F₂ progeny revealed seven differentiating polymorphic markers (three RAPD, two ISSR and two SSR) of 102 markers screened. Subsequently, these markers were used to genotype the entire F₂ population, and a genetic linkage map covering 74.7 cM distance was developed. The major locus Xca1bo was mapped in 1.6‐cM interval flanked by the markers RAPD 04₈₃₃ and ISSR 11₆₃₅. The Xca1bo locus was located on chromosome 3. The linked markers will be useful for marker‐assisted resistance breeding in cauliflower.     

Development of an interspecific Vigna linkage map between Vigna umbellata (Thunb.) Ohwi & Ohashi and V. nakashimae (Ohwi) Ohwi & Ohashi and its use in analysis of bruchid resistance and comparative genomics

To facilitate transfer of bruchid resistance to azuki bean (Vigna angularis) from its relatives an interspecific mapping population was made between rice bean, V. umbellata, and the related wild species V. nakashimae. The V. umbellata parent is completely resistant and V. nakashimae is completely susceptible to the bruchid beetle pests, azuki bean weevil (Callosobruchus chinensis) and cowpea weevil (C. maculatus). There is very low cross compatibility between V. umbellata and azuki bean. Therefore, V. nakashimae, that crosses with both V. umbellata and V. angularis without the need for embryo rescue, is used as a bridging species. A genetic linkage map was constructed based on an interspecific F₂ mapping population between V. umbellata and V. nakashimae consisting of 74 plants. A total of 175 DNA marker loci (74 RFLPs and 101 SSRs) were mapped on to 11 linkage groups spanning a total length of 652 cM. Segregation distortion was observed but only three markers were not linked to any linkage group due to severe segregation distortion. This interspecific genome map was compared with the genome map of azuki bean. Of 121 common markers on the two maps, 114 (94.2%) were located on the same linkage groups in both maps. The marker order was highly conserved between the two genome maps. Fifty F₂ plants that produced sufficient seeds were used for quantitative trait locus (QTL) analysis and locating gene(s) for C. chinensis and C. maculatus resistance in V. umbellata. The resistance reaction of these F₂ plants differed between C. chinensis and C. maculatus. Both resistances were quantitatively inherited with no F₂ plants completely susceptible to C. chinensis or C. maculatus. One putative QTL for resistance to each of these bruchid species was located on different linkage groups. Other putative QTLs associated with resistance to both C. chinensis and C. maculatus were localized on the same linkage group 1. Linked markers associated with the bruchid‐resistant QTL will facilitate their transfer to azuki bean breeding lines.