YANG Meng-jiao,WANG Cai-rong,,Muhammad Adeel HAssAN,WU Yu-ying,XIA Xian-chun,sHI shu-bingXIAO Yong-gui,HE Zhong-hu,

1 College of Agriculture,Xinjiang Agricultural University,Urumqi 830052,P.R.China

2 National Wheat Improvement Centre,Institute of Crop Sciences,Chinese Academy of Agricultural Sciences (CAAS),Beijing 100081,P.R.China

3 Institute of Agricultural Science of Yili Prefecture,Yining 835000,P.R.China

4 International Maize and Wheat Improvement Centre (CIMMYT) China Office,c/o CAAS,Beijing 100081,P.R.China

Abstract Plant nitrogen assimilation and use efficiency in the seedling’s root system are beneficial for adult plants in field condition for yield enhancement.Identification of the genetic basis between root traits and N uptake plays a crucial role in wheat breeding.In the present study,198 doubled haploid lines from the cross of Yangmai 16/Zhongmai 895 were used to identify quantitative trait loci (QTLs) underpinning four seedling biomass traits and five root system architecture (RSA) related traits.The plants were grown under hydroponic conditions with control,low and high N treatments (Ca(NO3)2·4H2O at 0,0.05 and 2.0 mmol L?1,respectively).Significant variations among the treatments and genotypes,and positive correlations between seedling biomass and RSA traits (r=0.20 to 0.98) were observed.Inclusive composite interval mapping based on a high-density map from the Wheat 660K single nucleotide polymorphisms (SNP) array identified 51 QTLs from the three N treatments.Twelve new QTLs detected on chromosomes 1AL (1) in the control,1DS (2) in high N treatment,4BL (5) in low and high N treatments,and 7DS (3) and 7DL (1) in low N treatments,are first reported in influencing the root and biomass related traits for N uptake.The most stable QTLs (RRS.caas-4DS) on chromosome 4DS,which were related to ratio of root to shoot dry weight trait,was in close proximity of the Rht-D1 gene,and it showed high phenotypic effects,explaining 13.1% of the phenotypic variance.Twenty-eight QTLs were clustered in 12 genetic regions.SNP markers tightly linked to two important QTLs clusters C10 and C11 on chromosomes 6BL and 7BL were converted to kompetitive allele-specific PCR (KASP) assays that underpin important traits in root development,including root dry weight,root surface area and shoot dry weight.These QTLs,clusters and KASP assays can greatly improve the efficiency of selection for root traits in wheat breeding programmes.

Keywords:KASP marker,QTL analysis,root traits,SNP array,Triticum aestivum

1.Introduction

Nitrogen (N) is the principal nutrient element in root development and photosynthate accumulation during the plant growth cycle (Hermanset al.2006;Fordeet al.2014;Cormieret al.2016).N fertilizers are essential for the maximum ecosystem productivity and meeting food demand(Moore and Lobell 2015).In recent years,N fertilizer use in terms of pure N increased from 92 to 108 Mt in the world(FAOSTAT 2020).Over-application and low nitrogen use efficiency (NUE) have caused major resource concerns and substantial greenhouse gas emissions (Diaz and Rosenberg 2008;Guoet al.2010;Zhang and Wang 2015;Zhang Wet al.2016).In China,the agricultural system generally uses high to excessive N fertilizer,and the total average application of N for winter wheat (Triticum aestivumL.) has increased to more than 500 kg N ha?1.Whereas the NUE in the wheat production system was lower than in maize (Zea maysL.) and rice (Oryza sativaL.),by approximately 25%(Zhang Wet al.2016;Cuiet al.2018),substantial regulation and breeding new varieties for high N acquisition would be an effective approach for improving NUE and yield potential(Rengel and Marschner 2005).

In wheat,root system architecture (RSA) defines the spatial configuration of root structure that includes the root’s number,length,tip number,emergence angles,width,depth,convex hull area,and root mass center.An increased root biomass could help plants to maintain a balance between the shoot and root under N deficiency,and also has a significant impact on yield enhancement (Barracloughet al.1989;Robinsonet al.2001;Reynoldset al.2007;Cormieret al.2016;Bettembourget al.2017).Development of RSA-related traits depend on genetic features of individual plants and the growing environment.It is difficult for genetic studies to investigate underground root traits for large sets of samples.Thus,there is a need for high-throughput methods for root phenotyping to determine the best-performing wheat genotypes for precise selection.

Identifying multiple genes for root growth in response to N uptake provides a promising way to optimize RSA,and reduces N fertilizer requirements while maintaining yield stability and to overcome the deceasing resource problems in the future (Wassonet al.2012).TheRhtgenes controlling shoot height have been reported to reduce root proliferation and promote increased N uptake from underground resources by manipulating root systems (Lynchet al.2007;Hundet al.2009;Baiet al.2013;Cuiet al.2014;Narayanan and Prasad 2014;Ryanet al.2015;Azizet al.2017).Quantitative trait loci (QTLs) mapping based on high-density maps has increased the understanding of the genetic basis for root traits,and marker-assisted selection(MAS) in multiple traits has contributed to genetically improved wheat varieties (Guoet al.2012;Atkinsonet al.2015;Subiraet al.2016).An increased grain yield was achieved through introgression of QTLs into backcrossderived lines,and some studies have provided evidence for the feasibility of improving grain yield by manipulating root systems (Li P Cet al.2015).Moreover,a major locus on chromosome 2BS,qTaLRO-B1,was determined to affect root length and biomass accumulation of wheat seedlings,and the linked markers were developed for further breeding purposes (Caoet al.2014).

The objectives of this study were to screen the doubled haploid (DH) lines derived from the Yangmai 16/Zhongmai 895 cross for seedling biomass and root traits using hydroponic culture,to identify QTLs at the seedling stage under three N conditions using 660K single nucleotide polymorphisms(SNP) array,and to develop kompetitive allele-specific PCR(KASP) markers associated with important loci to increase N uptake capacity in further wheat breeding programs.

2.Materials and methods

2.1.Plant materials

The mapping population used in the present study included 198 DH lines derived from a cross between two Chinese wheat varieties,Yangmai 16 and Zhongmai 895.Zhongmai 895 is a facultative variety from the southern part of the Yellow and Huai River Valley Winter Wheat Zone,China with high yield potential and strong root vigor,whereas Yangmai 16 is a spring wheat variety widely grown in the Middle and Lower Yangtze River Valley Wheat Zone,China.

2.2.Hydroponic culture and experimental design

Root screening in hydroponics at the seedling stage was performed using Hoagland’s nutrient solution (Hoagland and Arono 1950),with three N treatments,viz.,control,low and high N based on Ca(NO3)2·4H2O concentrations of 0,0.05 and 2.0 mmol L?1,respectively.To maintain the Ca nutrient in a common concentration,CaCl2·2H2O solution (2.0,1.95 and 0 mmol L?1,respectively) was applied corresponding to the Ca in the three N levels (Li Fet al.2015) (Table 1).A randomized complete block design was used to minimize the experimental errors,with three replications in a temperaturecontrolled greenhouse from March 15 to April 28,2016.

Thirty seeds of each line were surface-sterilized in a 10% H2O2solution for 15–20 min,rinsed in sterilized water 5–6 times,and germinated on moist germination paper in Petri dishes.The germinated seeds were transferred to plastic trays filled with quartz sand (2 mm diameter) and held in darkness at 24°C for 72 h.Nine seedlings for each N treatment,i.e.,three seedlings for each replication,were chosen and placed into holes fastened by spongy material and then transferred into a plastic tank (660 mm×480 mm×280 mm) containing 20 L of nutrient solution in an environmentally controlled room(16°C in day time with a light intensity of 400 μmol m–2s–1photosynthetically active radiation and 13°C at night,and a relative humidity of 70%).The solution was changed every 3 days.Plants were harvested after 10 days and stored in 30% ethanol prior to imaging.

Table 1 Nutrient solution ingredients for wheat seedling growth

2.3.Trait measurements

The seedling biomass traits,viz.,shoot dry weight (SDW)and root dry weight (RDW),were determined after oven drying at 70°C for 72 h using 1/1 000 balances.The total dry weight (TDW) was calculated as the sum of SDW and RDW,and the ratio of root to shoot dry weight (RRS) was defined as the ratio of RDW to SDW.RSA related traits,viz.,main root length (RL),root diameter (RD),root surface area(ROSA),root tip number (RTN),and root volume (RV),were measured using a recording scanner (Perfection V700/V750 2.80A;Epson,China),and images were analyzed by semiautomated Software RootNav V1.7.5 (Poundet al.2013).

2.4.sNP genotyping and QTL detection

The DH lines and parents were genotyped using the Wheat 660K iSelect SNP array from Affymetrix at Capital Bio Corporation (Beijing,China;http://www.capitalbio.com).The genetic map was constructed by Wanget al.(2017).The linkage map of Yangmai 16/Zhongmai 895 population contained 10 242 SNP markers in 25 linkage groups,covering all 21 wheat chromosomes.Inclusive composite interval mapping (ICIM) was used to QTL,based on the best linear unbiased estimation (BLUE) values from three replicates by the Software IciMapping V4.1 (Menget al.2015).The genotypes of Yangmai 16 and Zhoumai 985 were defined as 0 and 2,respectively.Hence,alleles from Zhongmai 895 increased trait values when the additive effects were positive.Recombination frequencies were converted into map distances using the Kosambi mapping function (Kosambi 1943).In order to ensure the existence of accurate QTLs,the LOD value of 2.5 was used as a threshold based on the 1 000 permutation test.The QTL was computed with a walking step of 1.0 cM,a PIN of 0.001 and a Type I error of 0.05 (Churchill and Doerge 1994).

2.5.Conversion of sNP markers to KAsP assay

The physical positions of markers were based on wheat genome sequences from the International Wheat Genome Sequencing Consortium (IWGSC RefSeq 1.0,http://www.wheatgenome.org/).Allele specific primers for KASP assays were designed using the PolyMarker (http://polymarker.tgac.ac.uk/) following Rasheedet al.(2016).The detailed information for KASP markers corresponding to two SNPsAX-109558906on chromosome 6B andAX-95025477on chromosome 7B,respectively,is provided in Appendix A.The primer mixture included 46 μL ddH2O,30 μL common primer(100 μmol L–1) and 12 μL of each tailed primer (100 μmol L–1).Assays were tested in 384-well formats and set up as～3 μL reactions (10–20 ng μL–1DNA,3 μL of 1× KASP master mixture and 0.056 μL of primer mixture).PCR cycling was performed by the following procedure:hot start for 15 min at 95°C,10 touchdown cycles (95°C for 20 s;touchdown at 65°C initially and decreasing by ?1°C per cycle for 25 s),then 30 additional cycles of denaturation and annealing/extension (95°C for 10 s;57°C for 60 s).PCR was performed in a Bio-Rad CFX Real-Time PCR System,and fluorescence was detected using the Bio-Rad CFX Manage 3.1 Software.

2.6.statistical analysis

Data analysis was performed using the R package (R Core Team 2013).Correlations were determined by Pearson’s coefficient,and significance tests among DH lines and parents were made using a mixed linear model atP<0.05.Broad-sense heritability for each trait was estimated by:

h2=σg2/(σg2+σgt2/r+σε2/rt)

whereσg2is the genetic variance,σgt2is the genotype(line)×N treatment interaction,σε2is the error variance,andtandrrepresent the numbers of N treatments and replications,respectively.

3.Results

3.1.Phenotypic performance of DH lines for seedling biomass and RsA related traits under three N levels

The phenotypic performance of DH lines was normally distributed,and transgressive segregations were observed for seedling biomass and RSA related traits under all three N levels (Appendices B and C).Significant variations(P<0.001) were observed among the DH lines with high broad-sense heritabilities (0.83 to 0.98).Variations for all traits,treatments and the genotype×treatment interaction were also significant (Table 2).In particular,DH lines exhibited higher values for RRS,RL and RTN under the control and low N conditions,and higher values for SDW,TDW,RD,and RV at the high N level.Yangmai 16 had higher values of SDW,RDW,TDW,RL,and RTN under the control and low N conditions,whereas Zhongmai 895 showed higher values at the high N level for all traits except RTN(Fig.1).

High and positive correlations between most seedling biomass and RSA traits were observed across all three N levels (r=0.20 to 0.98;Fig.2).RD showed significantly negative correlations with RTN (r=?0.45 to ?0.62),and with RL at the low N (r=?0.29) and high N levels (r=?0.40).Negative correlations were also observed between SDW and RRS across the three N treatments.

3.2.QTL for seedling biomass traits

Twenty-five QTLs were mapped on 12 chromosomes for seedling biomass traits (Table 3).Ten major QTLs,accounting for >10% of the phenotypic variances,were located in the three N treatments.Twelve QTLs were detected in the high N treatment,seven in the control and six in the low N treatment.One stable QTL on 4DS was detected across the three N treatments.This corresponded toRht-D1region with favorable alleles from the male parent Zhongmai 895,explaining 13.1 to 19.0% of the phenotypic variances.A QTL for RRS on 4BS was detected under both low and high N treatments,corresponding to theRht-B1region.The presence ofRht-B1bin Yangmai 16 andRht-D1bin Zhongmai 895 and their segregation in DH population was confirmed through functional markers in a previous study (Hassanet al.2019;Appendix D).At the high N level,an important pleiotropic QTL on 6BL simultaneouslycontrolling SDW,RDW and TDWexplained 10.4,9.9 and 10.3% of the phenotypic variances,respectively.

Table 2 Significance test and heritability (h2) values for the measured traits of doubled haploid (DH) lines in three nitrogen (N)treatments

Fig.1 Phenotypic differences between parents Yangmai 16 and Zhongmai 895 under different nitrogen (N) treatments.C,control;L,low N treatment;H,high N treatment.SDW,shoot dry weight;RDW,root dry weight;TDW,total dry weight;RRS,ratio of root to shoot dry weight;RL,root length;RD,root diameter;RV,root volume;RTN,root tip number;ROSA,root surface area.Error bars represent standard deviations for each proportion (n=3);different letters indicate significant differences between the genotypes determined by Duncan’s multiple range tests.*,significant at P<0.05.

3.3.QTL for RsA-related traits

Twenty-six QTLs were identified for all five RSA traits(Table 3).Among them,five QTLs for RL were identified on chromosomes 1AL (2),2BS,3BS,and 7DS,nine for RD on chromosomes 1AL,1DS,2AL,3AS,4AS,4BS,4BL,5AL and 5BL,four for RV on chromosomes 4BL (2),5AS and 7DS,five for RTN on chromosomes 1DS,2BS,4DS,5AL and 7DS,and three for ROSA on chromosomes 2AS,4BL and 7BL.The phenotypic variances explained (PVE) of these QTLs ranged from 4.7 to 19.1%.Five major QTLsQRD.caas-1AL,QRD.caas-4BL,QRV.caas-4BL,QRTN.caas-5AL,andQROSA.caas-4BLwith larger effects,explained 19.1,18.3,13.3,14.5,and 12.5% of phenotypic variances,respectively.Under the low N treatment,one pleiotropic QTL for RL and RTN on 2BS,and one for RL and RV on chromosome 7DS were identified in the marker intervals ofAX-108920782–AX-110463005(2BS)andAX-1089522259–AX-111881572(7DS) under low N treatment,respectively.Two pleiotropic QTLs conditioning RD and RTN in the high N treatment were identified on 1DS and 5AL in the marker intervals ofAX-109849862–AX-108727857(1DS) andAX-109958693–AX-94700681(5AL),respectively.These QTLs explained 5.9 to 14.5%of the phenotypic variances.

3.4.QTL clusters

Twelve clusters (C1–C12) including 28 QTLs for different traits were identified on 10 chromosomes (i.e.,1AL,1DS (2),2BS,4BL (2),4DS,5AL,6AL,6BL,7BL,and 7DS) across the three N treatments (Figs.3 and 4).Among these,more QTL clusters under the high N treatment (5) were identified than those under the control (1) and the low N treatment(3),and the other three clusters were detected in two or three N treatments.

Fig.2 Correlations between seedling biomass and root system architecture (RSA)-related traits under three nitrogen (N) treatments.Color intensity indicates the levels of positive and negative significance at P<0.05.Red and yellow colors indicate significantly positive and negative correlations,respectively,whereas white color indicates no significant correlation.C,control;L,low N treatment;H,high N treatment.SDW,shoot dry weight;RDW,root dry weight;TDW,total dry weight;RRS,ratio of root to shoot dry weight;RL,root length;RD,root diameter;RV,root volume;RTN,root tip number;ROSA,root surface area.

In the control,C1 for RL and RRS was detected on chromosome 1AL in the marker intervalAX-89541634–AX-109280493.In the low N treatment,three QTL clusters were identified including C3 for SDW and TDW on chromosome 1DS,C4 for RDW,RL and RTN on 2BS,and C12 for RL and RV on 7DS.In the high N condition,C8,C9,C10,and C11 were mapped on chromosomes 5AL,6AL,6BL,and 7BL,respectively,comprising 14 QTLs and explaining 5.1 to 14.5% of the phenotypic variances.In addition,C5 for RDW and RV and C6 for RDW,RV and ROSA were detected on chromosome 4BL.In addition,C7 for SDW and RRS wasmapped on 4DS (AX-109861583–AX-109478820),with favorable alleles contributed by Zhongmai 895.

Table 3 Quantitative trait loci (QTLs) for seedling biomass and root system architecture (RSA) traits in the Yangmai 16/Zhongmai 895 doubled haploid (DH) population1)

Fig.3 Twelve QTL clusters for seedling biomass and root system architecture (RSA) traits.Different colors indicate traits and shapes indicate the three N treatments.C,control;C1–12,QTL clusters 1–12;L,low N treatment;H,high N treatment.SDW,shoot dry weight;RDW,root dry weight;TDW,total dry weight;RRS,ratio of root to shoot dry weight;RL,root length;RD,root diameter;RV,root volume;RTN,root tip number;ROSA,root surface area.

3.5.KAsP marker development for important loci

Two KASP assays were developed for new QTL clusters C10 and C11.The first assay forAX-109558906-6Blocus in C10 was related to SDW,RDW and TDW,and the second one forAX-95025477-7Blocus in C11 controlled SDW,RDW and ROSA (Appendix A).Fourteen varieties and advanced breeding lines and two parents were chosen for dissecting the characteristics of SDW,RDW,TDW,and ROSA to verify the markers using Gel-free KASP assays.AtAX-109558906-6Bloci,the genotypes showing the blue color for“AA”allele were the same as Zhongmai 895,while the red for“GG”allele were the same as Yangmai 16.Higher SDW,RDW and TDW were observed in six varieties with the“AA”allele than those eight with the“GG”allele.AtAX-95025477-7Blocus,the genotypes showing the blue color for“GG”allele were the same as Yangmai 16,whereas those with the red color for“CC”allele were the same as Zhongmai 895.The ROSA among six varieties with the“CC”allele were greater than those eight varieties with the“GG”allele (Fig.5;Appendix E).The genotyping results of the entire population with two KASP markers were the same as those of the corresponding SNPs from the Wheat 660K SNP assay.

4.Discussion

4.1.Phenotypic trait variation and correlations

Considering the importance of root traits for nutrient uptake,root characteristics at seedling stage could help to predict the mature root system and important yield related traits as previously established by correlation analysis (Atkinsonet al.2015).Complexity in observing these“hidden half”has remained a bottleneck in understanding their genetic control and effective optimization for efficient nutrient uptake(Meisteret al.2014).Advanced phenotyping approaches have opened a new avenue for detecting precise and novel phenotypic information which allows for effective improvement of relevant traits.

In this study,rapid screening of a large DH population for seedling biomass and RSA related traits was accomplished to dissect their genetic basis under three N conditions through a hydroponic culture-based pipeline.Significant phenotypic variations (P<0.001) and high repeatability indicated that DH lines presented high genetic diversity for root traits at seedling stage.It had great value for dissecting the genetic basis of plant growth and N uptake(Anet al.2006;Melinoet al.2015).The significant genotype×treatment interaction detected under different N levels demonstrated that the deep information of root traits was important,and could be significantly affected by nitrogen supply conditions.Similar results were observed in root traits under nutrient supply manipulations by Guoet al.(2012) and Hornet al.(2016).

Fig.4 Root phenotyping at the high nitrogen (N) level and effects contributed by 12 quantitative trait loci (QTLs) clusters from Yangmai 16 and Zhongmai 895.YM16,Yangmai 16;ZM895,Zhongmai 895.C,control;C1–12,QTL clusters 1–12;L,low N treatment;H,high N treatment.SDW,shoot dry weight;RDW,root dry weight;TDW,total dry weight;RRS,ratio of root to shoot dry weight;RL,root length;RD,root diameter;RV,root volume;RTN,root tip number;ROSA,root surface area.

Fig.5 Scatter plots for two kompetitive allele-specifc PCR (KASP) marker assays in 16 wheat varieties.A,KASP assay for AX-109558906-6B,the blue color for“AA”allele corresponded to the Zhongmai 895 genotype,whereas the red color for“GG”allele corresponds to the Yangmai 16 genotype.B,KASP assay for AX-95025477-7B,the blue color for“GG”allele corresponds to the Yangmai 16 genotype,and those with the red color for“CC”allele correspond to the Zhongmai 895 genotype.Scatter plot for KASP assays showing clustering of genotypes on the X-(FAM) and Y-(HEX) axes.Blue genotypes have the FAM-type allele;red genotypes have the HEX-type allele;black dots indicate the NTC (non-template control).

DH lines performed better for all traits investigated,such as higher RTN,RRS and RDW in low N condition and RV,TDW and SDW in high N condition (Appendix C).High levels of these traits in correlation with root early vigor and efficient accumulation of nutrients,and transgressive segregation indicated that DH lines could be a potential source for the identification of novel alleles (Gaoet al.2017).Flexibility and alteration in root traits could help adult plants by modifying the cellular mechanisms according to the influence of varied growing environments (López-Bucioet al.2002;Hornet al.2016).In the present study,we observed longer RL and RRS,and more RTN under the control and low N conditions.Shorter RL,less RTN and higher SDW and TDW were found in the high N condition.These types of changes could improve plant stability under diverse nutrient conditions.Strong correlations among measured traits were indicative of closely linked or pleiotropic loci,and RL,RRS and RTN were crucial in N uptake in the three N levels.Zhongmai 895 showed higher values of traits at the high N level compared to Yangmai 16,because Zhongmai 895 was developed through selection by screening for high yield potential,NUE and drought resistance (Heet al.2014).Therefore,these traits could be selection targets in breeding,and significant variations could help to identify important QTLs for adaptive traits (Liet al.2011;Yuanet al.2017).

4.2.QTL identified for nitrogen supply

Efficient N accumulation and utilization are crucial for plant growth in fulfilling N requirements (Ryanet al.2015;Hornet al.2016).In present study,51 QTLs were identified under three N treatments.In low N treatment,two new QTLs on 4BL for RDW and ROSA,three new QTLs for RL,RV and RTN on 7DS and one for TDW on 7DL showed significant phenotypic variances,ranging from 6.1 to 12.5%.Five new QTLs were identified under high N conditions,including two for RD and RTN on 1DS,and three for RDW,RD and RV on chromosome 4BL.These QTLs were linked with the genes that might affect root early vigor and would be important for the development of N-efficient wheat varieties (Table 3).

Fourteen QTLs were co-localized with previously reported loci for seedling and yield-related traits (Anet al.2006;Liet al.2011;Yuanet al.2017;Guanet al.2018).Among them,the most stable QTL,QRRS.caas-4DSforRht-D1b,was confirmed through its functional marker by Hassanet al.(2019),and Zhongmai 895 contributed the favorable allele.This could provide a useful selection criterion in N-deficient and sufficient conditions,by using a gene-specific marker forRht-D1b(Rasheedet al.2016),which could help in evaluating root phenotypes and might be used for markerassisted selection.Moreover,a major QTL with significant phenotypic impact was detected on chromosome 4BS (for RRS and RD) corresponding toRht-B1on 4BS at three different confidence intervals,with a favorable allele from Yangmai 16.Previously,it had been discussed that the range of SNP confidence intervals linked withRht-B1andRht-D1on chromosome 4D and 4B were wide when cross checked through genomic prediction analysis (Hassanet al.2019).These QTLs have been shown to play a key role in early vigor at the seedling stage by affecting root length,and root and shoot weight (Hornet al.2016;Gaoet al.2017).Notably,Lapercheet al.(2007) demonstrated the role of dwarfing geneRhtin increasing N content (g N m?2)as well as grain protein content in plants.The QTL on 2BS showed significant effects on RDW,RL and RTN,which were likely to be co-localized with thePpd-B1gene and an important locus,qTaLRO-B1,which is involved in shoot height fluctuations,seedling biomass accumulation and mineralization of N at later growth stages (Caoet al.2014).In our results,the presence of dwarfing alleles of plant height on 4BS,4DS and 5AL and the photoperiodism gene on 2BS had significant influences in biomass accumulation,root development and N accumulation (Lapercheet al.2007;Quraishiet al.2011).

At the high N level,10 QTLs were located on chromosomes 5AL,6AL,6BL,and 7BL.Among them,5AL had significant influences on RD and RTN.It also had roles in increasing the duration of organ development for nitrogen accumulation,and grain yield under rain-fed conditions (Quraishiet al.2011;Gahlautet al.2017;Guanet al.2018).Three QTLs on 6BL for SDW,RDW and TDW were likely in the same region as a previously reported gene for NUE (Guoet al.2012),and thousand-grain weight and grain yield (Yuanet al.2017).In addition,an important QTL on chromosome 7B was reported to have a strong relationship with kernel number per spike (Guanet al.2018).Three QTLs on 7BL for SDW,RDW and ROSA were detected,with favorable alleles from Zhongmai 895.These important QTL for seedling biomass and RSA traits could be useful for root morphogenesis and early vigor,and for high degrees of development of mature plant organs and yield stability under varied N conditions.

4.3.QTL clusters

The presence of favorable genetic relationships between the traits and stable loci in QTL clusters are more reliable under varied environments (Gonget al.2016).In the present study,12 QTL clusters were identified for closely associated traits (Fig.3).C1 on 1AL showed 9.9 to 10.8%of PVE for RL and RRS under control treatment,and C2 and 3 on 1DS exhibited 8.5 to 8.8% of PVE for SDW,RD and RTN under low and high N levels.These are likely to be new QTL for the alteration of root and biomass-related seedling traits.C4 on chromosome 2BS was likely to be linked withPpd-B1which has been cloned as an important gene for harvest index and NUE.C7 was the most stable cluster for RRS and SDW detected in all three N treatments(AX-109816583–AX-109478820) on chromosome 4DS,with a significant phenotypic effect ranging from 5.3 to 19%,and all of the favorable alleles from C7 were provided by Zhongmai 895;also,it is likely to be the reduced height alleleRht-D1b,affecting root and shoot biomass (Ryanet al.2015 ;Hornet al.2016).Rhtloci have effects on increasing N content,thousand-grain weight and grain yield.Therefore,the presence ofRht1andRht2genes in DH lines seems to have a great influence in N uptake.Besides the clusters containing dwarfing and photoperiod related genes,the important QTLs cluster of C10 on chromosome 6BL(AX-109558906–AX-110028322) significantly increased SDW,RDW and TDW with 9.9 to 10.4% PVE,and C11 on chromosome 7BL (AX-95025477–AX-95121748) enhanced SDW,RDW and ROSA with 5.1 to 7.8% of PVE at the high N level.KASP markers for C10 and C11 on chromosomes 6BL and 7BL were developed and validated in the study.Previously,the QTL in cluster of C10 was reported for N accumulation and thousand-grain weight in RIL populations under field conditions (Gahlautet al.2017;Yuanet al.2017).While the QTL identified in C11 were likely to be new for fluctuations of seedling traits in response to N,these QTLs have been reported for kernel number per spike in a previous study (Zhang Het al.2016).Thus,the above QTL clusters,particularly C10 and C11,could provide important information of genetic background for early vigor selection of N uptake efficiency genotypes.

5.Conclusion

Fifty-one QTLs for seedling biomass and RSA-related traits were detected across three N treatments using a hydroponic experiment for high-throughput genotyping in wheat.Twelve new QTLs on chromosomes 1AL,1DS (2),4BL (5),7DS(3) and 7DL influenced the root biomass traits for N uptake.The QTL on chromosome 4DS for RRS with high phenotypic effects was most stable across the N treatments.Among 12 QTL clusters,C10 on 6BL for SDW,RDW and TDW,and C11 on 7BL for SDW,RDW and ROSA were new loci under high N levels,and these were also reported to have links with yield-related traits.These QTLs and clusters associated with seedling biomass and RSA-related traits,together with KASP markers developed for C10 and C11,could help to genetically improve the N uptake efficiency in wheat.

Acknowledgements

We thank Prof.R.A.McIntosh from Plant Breeding Institute,University of Sydney,for review of this manuscript.This work was funded by the National Key R&D Program of China (2016YFD0101804-6),the National Natural Science Foundation of China (31671691) and the International Science &Technology Cooperation Program of China(2016YFE0108600).

Declaration of competing interest

The authors declare that they have no conflict of interest.

Appendicesassociated with this paper are available on http://www.ChinaAgriSci.com/V2/En/appendix.htm