| Soybean is a major oil and grain crop and a leading source of protein and edible vegetable oil for human.It plays an important role in the development of agriculture and economy in China.To breed high yield and good quality varieties is the core target for soybean breeders.However,most of the target traits are controlled by multiple genes,which have small effects,and are also affected by the environment.Therefore,dissecting the genetic basis of these traits is important for soybean molecular breeding programs.A number of QTLs associated with agronomic,yield and quality traits have been reported in soybean.However,most of them are environmentally sensitive.Because of the technological limitations,the genetic maps used in most of the published studies always have low resolution and markers were not tightly linked with QTLs.These factors limit the use of QTLs in molecular-assisted selection of breeding or map-based gene cloning.On the other hand,linkage mapping based on bi-parental populations theoretically possible to identify only the polymorphic loci that are segregated between two parents,and cannot be comprehensively and effectively dissect the genetic basis of complex traits.The Chinese Jiang-Huai valley is an important soybean producing area.In this area,soybean is always planted after wheat harvest and the genotype of these varieties is different from early-maturing spring soybean.However,little is known about the genetic basis of agronomic and seed related traits in these genotypes.The objectives of this study were to map QTLs and identify stable and reliable loci that can be used for molecular assisted selection of breeding and map-based gene cloning,and then dissect the genetic basis of these traits in summer planting soybean.Four parents selected from Jiang-Huai valley were used in this study.Three soybean RIL（NJZN-RIL,NJMN-RIL and NJK3N）populations developed by crossing ZXD,M8108 and KF35（female）with NN1138-2（male）,respectively,were planted in multiple environments to investigate phenotypic data.All RIL populations were genotyped using reduced-representation genome sequencing（RRGS）technology to generate genome-wide single nucleotide polymorphism（SNP）markers for the construction of high-density genetic maps.The genetic basis,major QTLs and closely linked markers conferring flowering time,plant height,100-seed weight,seed protein and oil content of summer planting soybean were revealed through QTL mapping by composite interval mapping method.Furthermore,two major QTLs were confirmed and fine mapped using secondary segregation populations.The major research results are as follows.1.Construction of linkage maps for soybean RIL populationsAfter genotyping and SNP-based polymorphic marker development,the linkage maps for three RIL populations were constructed.The NJZN-RIL and NJMN-RIL populations were genotyped using SLAF-seq technology.In NJZN-RIL population,a total of 3255 SLAF markers were grouped into 20 Linkage groups（LG）.The genetic distance of this map was 2144.85 cM and the mean LG length was 107.24 cM.The average distance between adjacent markers was 0.66 cM.In NJMN-RIL population,a total of 2062 markers were grouped into 20 LG.The genetic distance of this map was 2054.50 cM and the mean LG length was 102.73 cM.The average distance between adjacent markers was 1.00 cM.The NJK3N-RIL population was genotyped using RAD-seq technology.A total of 1733 bin markers were grouped into 20 LGs.The genetic distance of this map was 2362.44 cM and the mean LG length was 118.12 cM.The average distance between adjacent markers was 1.36 cM.2.QTL mapping for flowering time and plant height and fine mapping of qFT-15-1The combined analysis of variance under multi-environments showed that the hereditability（h2）of flowering time（FT）and plant height（PH）in these three populations was high（more than 80.02%）,suggesting that genetic variation could explain most of the phenotypic variation.Based on the genetic maps,6,4 and 9 QTLs for FT and 8,5 and 7 QTLs for PH with LOD scores ranging from 2.60 to 32.99 and 3.61 to 57.48%of the phenotypic variation explained by individual QTLs were identified across the different environments using composite interval mapping（CIM）method in NJZN-RIL,NJMN-RIL and NJK3N-RIL population,respectively.After integration,a total of 15 QTLs controlling FT and 16 QTLs controlling PH were detected in three populations,and six QTLs for FT and four QTLs for PH were novel loci.Among them,eight QTLs for FT and seven QTLs for PH were stable in different environments or populations and can be considered as major QTLs for those traits.Meanwhile,five pairs of QTLs were located in the same marker interval,which was found to have close linkage or pleiotropy effect controlling FT and PH.The qFT-15-1 QTL was detected in NJZN-RIL and NJK3N-RIL populations and could explain 8.55%-24.27%of the phenotypic variation in NJZN-RIL population,which could be considered a novel major QTL for FT.To refine the position and identify closely linked markers of this QTL,two secondary segregation（ZNBC2F2 and ZNBC2F3）populations developed from a backcross using ZN-61（a line of NJZN-RIL population）with NN1138-2 were used.Finally,qFT-15-1 was located in a small region of approximately 250kb（from 48.79Mb to 49.04Mb）on chromosome 15,between the InDel-15-1 and BARCSOYSSR151502 markers,and this region contains 12 functional genes.3.QTL mapping for seed protein and oil content and confirmation of qOil-5-1The combined analysis of variance under multi-environments showed that the hereditability（h2）of seed protein content and seed oil content in these populations was high（more than 83.30%）,suggesting that genetic variation could explain most of the phenotypic variation.Based on the genetic maps,12,9 and 11 QTLs for seed protein content and 9,9 and 11 QTLs for seed oil content with LOD scores ranging from 2.51 to 9.07 and 3.66 to 26.16%of the phenotypic variation explained by individual QTLs were identified across the different environments using composite interval mapping（CIM）method in NJZN-RIL,NJMN-RIL and NJK3N-RIL population,respectively.After integration,a total of 22 QTLs controlling seed protein content and 26 QTLs controlling seed oil content were identified in three populations,and eight QTLs for seed protein content and two QTLs for seed oil content were novel loci.Among them,nine QTLs for seed protein content and nine QTLs for seed oil content were stable in different environments or populations and can be considered as major QTLs for those traits.Meanwhile,eight pairs of QTLs were located in the same marker interval,which was found to have close linkage or pleiotropy effect controlling seed protein content and seed oil content.The qOil-5-1 QTL was detected in all environments in NJMN-RIL population and could explain 7.5-26.2%of the phenotypic variation in certain planting environments,which could be considered a major QTL for seed oil content.To confirm and refine the position of this QTL,a secondary F2（MNBC1F2）population developed from a backcross using MN-5（a line of NJMN-RIL population）with NN1138-2 was used.Finally,the position of qOil-5-1 could be refined at a 3 cM interval between the BARCSOYSSR051226 and BARCSOYSSR051250 markers.The interval of the QTL spanned approximately 1 Mb,from position 37.85 Mb（BARCSOYSSR051217）to 38.90 Mb（BARCSOYSSR051261）on chromosome 5 of the physical map.4.QTL mapping for 100-seed weightThe combined analysis of variance under multi-environments showed that the hereditability（h2）of 100-seed weight in these populations was high（more than 88.72%）,suggesting that genetic variation could explain most of the phenotypic variation.Based on the genetic maps,14,11 and 8 QTLs for 100-seed weight with LOD scores ranging from 2.52 to 10.77 and 2.76 to 29.36%of the phenotypic variation explained by individual QTLs were identified across the different environments using composite interval mapping（CIM）method in NJZN-RIL,NJMN-RIL and NJK3N-RIL population,respectively.After integration,a total of 27 QTLs were identified in three populations,and five QTLs were novel loci.Among them,eleven QTLs were stable in different environments or populations,which can be considered as major QTLs for 100-seed weight.In summary,the genetic maps of three RIL populations were constructed in this study.Based on these genetic maps,a total of 15,16,22,26 and 27 QTLs for flowering time,plant height,seed protein content,seed oil content and 100-seed weight were mapped in all populations.Meanwhile,the stable and reliable QTLs that can be used for molecular assisted selection of breeding and map-based gene cloning were identified.Furthermore,two major QTLs were confined by using secondary segregation populations.These QTLs reflected that there was a complex genetic basis of agronomic and seed traits in summer planting soybean.The results of this study are important for a comprehensive understanding of the genetic basis of agronomic and seed traits in summer planting soybean and also provide support for the implementation of MAS for breeding soybean and map-based cloning to elucidate the mechanisms of these traits. |
