QTL Mapping of Bread Wheat (Triticum aestivum L ...

ISSN 1021-4437, Russian Journal of Plant Physiology, 2017, Vol. 64, No. 1, pp. 48–58. © Pleiades Publishing, Ltd., 2017. Original Russian Text © Yu.V. Chesnokov, G.V. Mirskaya, E.V. Kanash, N.V. Kocherina, U. Lohwasser, A. Börner, 2017, published in Fiziologiya Rastenii, 2017, Vol. 64, No. 1, pp. 55–68.

RESEARCH PAPERS

QTL Mapping of Bread Wheat (Triticum aestivum L.) Grown under Controlled Conditions of an Agroecobiological Testing Ground Yu. V. Chesnokova, b, *, G. V. Mirskayaa, E. V. Kanasha, N. V. Kocherinab, U. Lohwasserc, and A. Börner c a

Agrophysical Research Institute, St. Petersburg, 195220 Russia Vavilov Institute of Plant Genetic Resources, St. Petersburg, 190000 Russia c Leibniz Institut fur Pflanzengenetik und Kulturpflanzenforschung (IPK), Gatersleben D-06466, Germany *e-mail: [email protected]

b

Received February 26, 2016

Abstract—To determine the effects of physiological and genetic interaction between the genotype and environment, QTL (quantitative trait loci) mapping of valuable traits of bread wheat (Triticum aestivum L.) manifesting under controlled conditions of an agroecobiological testing ground has been first carried out. In the course of two experiments, differing from each other only by temperature and illumination regimes and providing the strict control and invariability of other growing parameters, 99 QTLs determining 30 different agronomically important traits have been identified. According to the results of the QTL mapping and a single-factor ANOVA, changes in the temperature and illumination regimes did not influence 21 of 30 studied traits, which remained stable in their manifestation; only nine traits varied under these conditions, which indicates that their manifestation is dependent on changes in these environmental factors. Both statistical approaches used in this study demonstrated complementary results; for each of them, the maximum likelihood criterion was used, statistical significance was determined, and significance of results was evaluated. The significance of a correlation between the identified QTLs and the polymorphism of individual traits studied was assessed using the threshold value of LOD (logarithm of odds) score. In addition, QTL analysis allowed a block structure of the T. aestivum genome to be revealed, the percentage of a phenotypic variability determined by each of the identified QTLs to be calculated, and the determination of which of the parents donated individual QTL alleles. The obtained results can be used for the further study of the physiological and genetic mechanisms of realization of traits evaluated within the framework of the “genotype–environment” interaction and also for the marker-assisted breeding of wheat. Keywords: Triticum aestivum, agronomically important traits, QTL mapping, controlled conditions of agroecobiological testing ground, “genotype-environment” interaction, genome structure DOI: 10.1134/S1021443716060029

eighties using an approach called later as the QTL analysis [2]. In the beginning of the nineties, an international project on the molecular-genetic study of hexaploid wheat titled “International Triticeae Mapping Initiative” (ITMI) was started. The resulting mapping population of recombinant inbred lines (RIL) was saturated with RFLP, SSR, and SNP markers [3]. Using this population, scientists obtained an independent confirmation of the genome location of several genes with the known chromosomal localization; these genes included those associated with the red color of kernel (R1 and R3) and coleoptile (Rc1 and Rc3), inhibition of a wax coating formation in plants (W21), grain hardness (Ha), response to vernalization (Vrn1 and Vrn3), and leaf rust resistance (Lr34) [4, 5]. At the

INTRODUCTION The majority of morphological and economically important traits is quantitative and, as a rule, is determined by the allele structure of a certain number of genetic loci. According to the existing data, the set of genes determining the average value and genetic variance of a quantitative trait is usually specified by a limiting environmental factor. Any changes of the limiting factor entails the corresponding changes in the set of gene loci controlling the variability of the trait [1]. In addition to this fact, there are individual key genes that contribute to the formation of the given quantitative trait under any conditions, though the degree of this contribution is regulated by the environment. Such genetic loci were called quantitative trait loci (QTL) and have been actively studied since the beginning of 48

QTL MAPPING OF BREAD WHEAT (Triticum aestivum L.)

same period, some new QTLs were described, which regulated the resistance to stem and leaf rust [4], resistance to Pyrenophora tritici-repentis [6], and stripe rust resistance [7]. In Russia, the use of this mapping population was initially connected with the laboratory study and localization of QTLs determining grain quality parameters [8, 9]. Later the studies on the identification and molecular-genetic mapping of QTLs determining morphological and agronomically important traits of bread wheat (Triticum aestivum L.) were carried out using the ITMI mapping population in different ecogeographical regions of Russia [10, 11]; in addition, the ecological and genetic trials and the mapping of QTLs, controlling the manifestation of agronomically important traits in a gradient of nitrogen nutrition doses [12] and water regimes [13], were carried out. In general, the existing data demonstrate a general strategy for the study of the fundamentals of a “genotype–environment” interaction. At the current stage of development of the plant genetics and physiology, the basics of such interaction represent a key element in the determination of the “fine” structural organization and functioning principles of the higher plant genome. The first attempts to localize QTLs of the main agronomically important traits in different ecogeographical regions showed that QTL positions varied in different experiments that evidenced a significant environmental impact on the gene determination of quantitative traits. Thus, since 1991, a series of experiments was arranged to study the physiological and genetic interaction between the genotype and environment that made it possible to analyze the stability of the QTL manifestation under various environmental conditions. In the case of tomato, researchers from United States showed that, depending on environmental conditions, different QTLs can be manifested for the same trait [14]. Similar results were obtained for maize [15], arabidopsis [16], and wheat [5, 17, 18]. In Russia, the study of the QTL mapping in wheat under different ecogeographical conditions is carried out only by the research group of Chesnokov [10–12]. There are also some studies on the QTL mapping of wheat under controlled conditions of greenhouses [19], specially designed polytunnels [20], growth chambers [5, 17], and even in so-called “Phenomics greenhouses” [20]. However, according to Parent et al. [20], all these cases provided rather semicontrolled growth conditions, since the complete control of the plant growth and development in greenhouses, polytunnels, or Phenomics greenhouses is impossible, whereas growth chambers usually have too small area that does not allow researchers to place in them a large number of vegetation pots and, therefore, to grow a large number of plants under identical conditions of the vegetation experiment. The only exception is a specialized agroecobiological testing ground or in other words agroecobiopolygon, providing regulated and controlled conditions and having vegetation facilities isolated from the sunlight and any other external influRUSSIAN JOURNAL OF PLANT PHYSIOLOGY

Vol. 64

49

ence. This agroecobiopolygon is equipped with the climate control systems and the corresponding light sources of different types providing the year-round intensive cultivation of plants of different height (that is especially important for the work with agricultural crops). In addition, agroecobiopolygon also has equipment for the distant and contact diagnostics of the physiological and morphobiological state of vegetating plants. The strictly controlled conditions of such agroecobiological testing ground make it possible to arrange physiological and genetic experiments with a large number of plants using either separate vegetation pots or special stands filled with a peat ground [21, 22]. Under these conditions, the study of the effect of low positive temperatures and day length on the duration of the periods of vegetative growth of wheat was carried out using contrast regimes for the studied factors and, at the same time, maintaining the stability of other growing parameters. The series of vegetation cycles of the year-round plant cultivation under completely reproducible conditions allowed the research group of E.I. Ermakov and G.G. Panova to reveal the patterns of inheritance for the duration of the “sprouting–booting” and “booting–earing” ontogenetic periods in wheat [21, 22]. These authors also specified some theoretical statements concerning the selection of genotypes by transgressive traits and developed an approach for the practical development of wheat lines, in which the parameters of a selected trait are given in correspondence with the soil and climatic conditions of a zone in which these lines are planned to be cultivated [21, 22]. However, to date, there are no studies in which QTL mapping would be carried out to reveal chromosomal loci and/or genes determining the manifestation of agronomically important traits under controlled conditions of the agroecobiological testing ground, and to establish the character of interaction between certain (and/or determined) components of the “genotype– environment” interaction independently on the uncontrolled environmental impacts stochastically influencing on the manifestation of the studied and/or other traits, especially quantitative ones. Such studies would represent a practical basis for the “fine” identification and localization of QTLs in the “genotype– controlled environment” interaction that would be of both fundamental and practical importance for the plant physiology and genetics [2, 5, 12]. The purpose of this study was the revealing of the number and chromosomal localization of QTLs involved in the physiological and genetic process of realization of complex agronomically important traits in bread wheat (T. aestivum L.) and manifested under controlled conditions of the regulated agroecosystem of a biotesting ground. MATERIALS AND METHODS Plant material. The objects of our studies were recombinant inbred lines (RIL) of the ITMI mapping population of bread wheat (Triticum aestivum L.). The No. 1

2017

50

CHESNOKOV et al.

ITMI population was developed by the pollination of a bread wheat variety Opata-85 with synthetic hexaploid wheat W7984, an amphidiploid obtained by the crossing of Aegilops tauschii Coss. (accession CIGM86.940 (DD); male parent) and tetraploid hard wheat T. turgidum var. durum (variety Altar 84 (ААВВ); female parent) as described earlier [11–13]. The details of the genotype selection for the creation of the RIL collection and their characteristics are also described in [11–13]. The RIL growing and evaluation was carried out under controlled conditions of the agroecobiological testing ground (agroecobiopolygon) at the Agrophysical Research Institute (AFI, St. Petersburg). The study of 114 ITMI lines and their parental forms Opata 85 and W7984 was performed in vegetation irradiation chambers (VICs) equipped with DNaT-400 lamps providing a preset irradiance level. The chambers were located in a stationary facility isolated from the sunlight and equipped with a heating system and combined extract/input ventilation. The microclimatic parameters of plant cultivation in two arranged experiments differed by the irradiance level of top leaves (40 ± 0.5 W/m2 PAR and 50 ± 0.5 W/m2 PAR in the first and second experiments, respectively) and temperature regime (24−25°C day/19–20°C night and 28−29°C day/23−24°C night in the first and second experiments, respectively); other strictly controlled cultivation parameters were maintained invariable. As plants grew, the PAR value was corrected by the change of the distance between lamps and plants; thus, it was maintained constant for the whole vegetation period. Both experiments were carried out at a 16-h photoperiod. Plants were grown in 2-L vegetation pots. Each pot was used for the growing of one plant; the total plant density was 50 plants per 1 m2. For each line and parental form, the experiment was carried out in three replications. An “Agrophyt” substrate developed at the AFI on the basis of a high-moor peat of a low decomposition degree was used as a rooting medium, since it provides the most favorable hydrophysical and biological properties as compared with other organic and mineralrooting media used under sheltered ground conditions. The “Agrophyt” substrate has the following proportion of components: high-moor peat (1 L), Cambrian clay (60 g), chalk (5 g), and superphosphate (1.5 g) [21]. During the whole vegetation period, the humidity of the root medium was maintained at the level of 70–80% of the total moisture, since this value is optimal for wheat. Watering was carried out every day and in equal doses; water was alternated with Knop’s solution. Phenological observations were performed every 2 days, excepting basic ontogenetic stages, when the observations were performed on a daily basis. After the completion of vegetation experiments, the basic productivity parameters were determined for each plant.

Analysis of traits. The analysis of quantitative traits was carried out according to the methods accepted at the Vavilov Institute of Plant Genetic Resources (VIR) as described earlier [11–13]. Only traits manifesting a sufficient expressivity were considered. The total number of traits analyzed in the course of the whole vegetation period was 30 (Table 1). Statistical analysis. QTL analysis was performed using the Mapmaker/QTL program [5, 13]. Since this program uses a mathematical formula derived by Haldane [23], the mapping data published in the GrainGenes database (gopher: http://www.greengenes.cit. cornell.edu) were used to recalculate genetic distances on the map using the Mapmaker/EXP 3.0 program [24]. The data obtained concerning the phenotypic analysis were integrated into the existing basic map developed for the ITMI population [3]. The QTL identification and localization was carried out using the QGENE program as described earlier [5, 13]; only markers corresponding to the Kosambi mapping function, which takes account of the interference, were used. The reliability of the interrelation between the revealed loci and the polymorphism of any trait was evaluated based on the threshold value of LOD (logarithm of odds) score [10–12]. For each trait, a separate QTL analysis was carried out in each experiment, and the trait variation degrees (R2) explained by the given QTL were calculated. The significance of each LOD value was determined by a permutation test (1000 repetitions). Only loci with LOD ≥ 3.0 (p < 0.001), 2 < LOD < 3 (p < 0.01), and 1.5 < LOD < 2 (p < 0.1) were taken into account [11, 12]. A complex evaluation of compared mean values of traits determined under different growing conditions was carried out by one-factor ANOVA with the calculation of variation indices, such as the error mean squares and their variance ratio F; the significance of results was also evaluated [25]. The value p < 0.05 was considered as an acceptable statistical significance limit, since this level includes a 5% probability of error. Results significant at the level of p < 0.01 were considered as statistically significant, whereas results with the level of p < 0.001 were considered to be highly significant. All calculations were carried out using the Statistica 6.0 program. RESULTS The results of the study shown that the variation in the manifestation of agronomically important traits— studied under strictly controlled conditions of the agroecobiological testing ground—between the same RIL genotypes was significant for some QTL parameters and depended on the differing experimental conditions (Table 1). For example, the identification of chromosomal loci determining the manifestation of the trait controlling the duration of a “sprouting−booting” period resulted in the identification and mapping of

RUSSIAN JOURNAL OF PLANT PHYSIOLOGY

Vol. 64

No. 1

2017


QTLs localized at chromosomes 5A and 7D. Positions of one of QTLs localized closer to the central part of chromosome 5A (genetic distances were 191.8 and 89.0 cM, respectively) slightly differed in both experiments. In experiment 1, the maximum LOD value for this QTL was observed in the 5A position at a distance of 47.3 cM, whereas it was observed in the same position but at a distance of 89.0 cM under conditions of experiment 2. The QTL detected at the 7D position (131.5 cM) did not change its localization and remained stable determining a significant level of phenotypic variability of the studied trait in both experiments. The trait controlling the duration of a “sprouting−earing” period revealed more signiﬁcant differences concerning the involvement of chromosomal loci in the manifestation of this trait. Under conditions of experiment 1, two QTLs were detected on chromosomes 5A and 6A. At the same time, under conditions of experiment 2, both detected QTLs were localized on chromosome 5A. Only one QTL from the 5A linkage group had a coincided localization (89.0 cM); in both cases, the LOD value was high (3.87 and 3.78, respectively). The localization of the second QTL of this trait differed in two experiments. Note that, in the case of experiment 1, the higher level of manifestation (21.07%) of the trait controlling the duration of the “sprouting−earing” period was contributed by the QTL localized on chromosome 6A (it was probably determined by its participation in the realization of the trait, controlling the duration of the “sprouting−ripening” period) with the similar contribution into the phenotypic variability (21.48%). The second QTL controlling the duration of the “sprouting−ripening” period was also localized on chromosome 6A but at a different distance (108.2 cM). Quite a different picture was observed for the same trait under conditions of experiment 2: both QTLs controlling the duration of the “sprouting−ripening” period were localized on chromosome 5A at exactly the same position as the QTLs controlling the duration of the “sprouting−earing” period. The maximum level of phenotypic variability revealed for both traits was contributed by QTL localized on chromosome 5A (89.0 cM). The trait controlling plant height remained stable under both experimental conditions. Two QTLs were detected on chromosomes 6D and 3A. Though these QTLs did not show high LOD values, they contributed a significant level of phenotypic variability determining the manifestation of this trait in both experiments. The same situation was observed for the trait controlling the flag leaf position in the beginning of the earing phase. Both experiments revealed two QTLs localized on chromosome 5A at the distance of 230.9 cM; in addition, the QTLs localized at the distances of 170.2 and 170.7 cM were detected in experiments 1 and 2, respectively. The corresponding levels of phenotypic variability were 19.44% and 16.92% under conditions of experRUSSIAN JOURNAL OF PLANT PHYSIOLOGY

Vol. 64

51

iment 1 or 15.48% and 15.46% under conditions of experiment 2. Traits controlling the length of the upper internode and the size of a stem node also showed a stable localization in both experiments. For each of them, two QTLs were detected on chromosomes 1A and 3A (upper internode length) or 1B and 5D (stem node size). In the case of the trait controlling the stem node size, the localization of QTL on chromosome 5D was identical in both experiments (286.6 cM), whereas the distance made 145.4 and 169.2 cM for experiments 1 and 2, respectively, for chromosome 1B. In the case of the QTL controlling the upper internode length, the distance on chromosome 3A in experiments 1 and 2 was 51.3 and 56.3 cM, respectively; the distance on chromosome 1A in experiments 1 and 2 was 142.6 and 205.7 cM, respectively. In all the above-mentioned cases, the allele determining the manifestation of the traits was donated by the maternal form (Opata 85) excepting chromosome 7D, whose allele was contributed by W7984 (for the trait controlling the duration of the “sprouting−booting” stage). The manifestation of two other quantitative traits controlling the length and width of a f lag leaf was similar to the upper internode length and stem node size traits in both experiments. In the case of the f lag leaf length, one QTL was detected on chromosome 6A; for experiments 1 and 2, the distance was 151.6 and 118.3 cM, respectively. Both alleles were contributed by Opata 85, but the level of phenotypic variation in experiment 2 (17.83%) was two times higher than in experiment 1. For the trait controlling the flag leaf width, two QTLs were detected in both experiments on chromosomes 2D and 7A. All alleles were donated by Opata 85. For chromosome 7A, the difference in the revealed distance was not so significant (271.8 and 247.7 cM in experiments 1 and 2, respectively) as in the case of chromosome 2D (217.2 and 80.4 cM, respectively). At the same time, the LOD values in experiment 2 were significantly higher (3.95 and 2.43) than in experiment 1 (1.94 and 1.81), though the level of phenotypic variability determined by the detected QTLs was similar (13.84 and 14.81% for chromosome 2D and 16.94 and 10.89% for chromosome 7А). Traits controlling the presence of a wax coating on the axial and abaxial leaf sides and also on the stem and ear were stable concerning both chromosomal localization and genetic distance. For example, in the case of traits controlling the wax coating on the ear and stem, we detected two QTLs in each experiment, both localized on chromosome 2D. For both experiments and traits, these QTLs were localized at a distance of 295.8 and 300.0 сМ, respectively. The maximum levels of phenotypic variability for the traits controlling the wax coating on the ear and stem were 35.27 and 57.15%, respectively. In the case of the trait controlling the wax coating on the axial leaf side, one QTL on chromoNo. 1

2017

52

CHESNOKOV et al.

Table 1. Traits and QTLs revealed in the ITMI mapping population grown under controlled conditions of the agroecobiological testing ground Experiment** 1 Trait

Designation

Duration of the “sprouting— booting” period

VSB

Duration of the “sprouting— earing” period

VSH

Duration of the “sprouting— ripening” period

VSM

Plant height

PH

Upper internode length

StLul

Stem node size

StNS

Flag leaf position at the beginning of the earing phase

LFP

Flag leaf length

LFL

Flag leaf width

LFW

Wax coating on the axial leaf side

LWBi

Wax coating on the abaxial leaf side

LWBo

Wax coating on the stem

StWB

Wax coating on the ear

SpWB

Leaf: ligule color

LLC

Leaf: color of auricles

AuC

Flowering type

ChF

Ear texture Ear beardedness Ear fragility

SpT SpAwn SpBR

Form of ear/flower glume

GS

Kernel color

KC

Hard-threshing

DifThC

Ear length

SpL

Number of spikelets per ear

NSpt

Experiment** 2

localization

LOD

R2

localization

LOD

R2

5A (47.3) 7D (131.5) 5A (191.8) 5A (89.0) 6A (88.8) 6A (108.2) 6A (88.8) 6D (150.7) 3A (56.3) 3A (51.3) 1A (142.6) 5D (286.6) 1B (145.4) 5A (170.2) 5A (230.9) 6A (151.6) 2D (217.2) 7A (271.8) 2D (300.0) 3D (3.8)

3.67 3.29 3.24 3.87 2.67 3.36 2.73 2.41 1.79 3.06 2.76 2.32 2.16 2.63 2.54 2.02 1.94 1.81 2.87 1.99

14.26 19.45 21.36 16.02 21.07 22.42 21.48 19.21 15.22 21.89 18.29 15.86 9.95 19.44 16.92 8.65 13.84 14.81 11.95 13.74

2D (295.8) 2D (300.0) 2D (295.8) 2D (300.0) 1B (206.5) 5D (35.7) 1D (221.8) 4A (47.3) 3A (39.5) 7A (172.9) 5D (308.9) 2A (202.4) 5A (138.2) 3D (0.0) 3D (2.1) 3D (97.1) 3D (88.2) 5A (47.3) 2D (255.1) 4A (124.4) 2D (190.3) 2D (171.6)

18.48 17.00 3.69 3.35 2.24 1.65 2.91 1.98 1.82 1.68 1.69 1.91 2.14 2.61 2.05 6.39 6.03 3.48 1.98 2.46 3.55 2.91

55.19 53.59 15.20 14.42 10.17 6.87 20.97 12.89 16.04 11.56 13.90 14.54 12.29 12.25 14.14 42.00 37.54 15.81 14.11 11.96 17.70 19.99

5A (89.0) 5A (47.3) 7D (131.5) 5A (89.0) 5A (74.4) 5A (89.0) 5A (74.4) 6D (150.7) 3A (52.5) 3A (56.3) 1A (205.7) 5D (286.6) 1B (169.2) 5A (170.7) 5A (230.9) 6A (118.3) 2D (80.4) 7A (247.7) 2D (300.0) 2D (300.0) 3D (9.7) 2D (295.8) 2D (300.0) 2D (295.8) 2D (300.0) 1B (17.6) 5D (35.7) 4A (248.6) 2A (224.4) 6B (67.9) 7A (172.9) 5D (308.9) 2A (202.4) 5A (138.2) 3D (0.0) 3D (2.1) 3D (88.2) 3D (97.1) 5A (47.3) 2D (255.1) 4A (240.7) 4D (53.4) 2D (295.8)

4.03 3.47 2.93 3.78 3.10 3.80 2.28 2.12 2.00 2.14 2.14 2.19 3.09 2.26 2.22 2.64 3.95 2.43 4.45 5.02 2.00 18.04 17.39 9.26 8.07 2.08 1.68 3.00 2.36 2.32 2.08 2.57 1.75 1.60 2.83 2.69 10.18 8.02 2.72 2.02 2.76 2.65 2.54

15.38 13.51 17.53 16.74 13.81 18.24 11.26 18.76 17.46 20.05 20.05 17.03 26.16 15.48 15.46 17.83 16.94 10.89 19.22 21.41 14.23 57.15 56.58 35.27 32.09 14.53 7.67 21.85 16.92 16.04 14.33 21.47 14.34 9.13 12.82 18.37 38.01 33.65 13.55 16.08 13.33 12.94 12.33

Vol. 64

No. 1


Total*

5+1+0

3+1+0 2+2+0 0+3+1 1+3+0 1+3+0 0+4+0 0+2+0 1+1+2 1+1+0 1+1+1 4+0+0 4+0+0 0+2+2 1+2+1 0+2+2 0+1+1 0+0+2 0+1+1 0+4+0 4+0+0 1+2+1 0+2+0 1+3+0

2017


53

Table 1. (Contd.) Experiment** 1 Trait

Designation

Number of grains per spikelet Number of grains per ear

NseSpt NseSp

Grain weight per ear

GMSp

Weight of 1000 grains

TGW

Grain glassiness Number of ears

GrT NStS

Total*

Experiment** 2 R2

Total*

localization

LOD

R2

localization

LOD

6D (196.2) 6D (196.2) 2B (216.9)

1.83 3.87 3.91

22.54 41.74 46.29

3D (145.6) 7D (60.6) 6A (35.2) 5A (74.4) 16 + 20 + 13

2.80 1.99 2.84 2.59

20.26 22.52 31.90 11.92

1D (161.5) 1D (161.5) 4A (80.2) 7D (60.6) 7D (60.6)

3.34 2.79 3.42 2.80 6.01

24.03 1 + 0 + 1 20.47 1 + 1 + 0 28.50 2+1+0 24.90 45.93 1+1+1

6A (35.2) 5A (63.8) 20 + 27 + 3

2.29 4.06

11.40 0 + 2 + 0 28.36 1 + 1 + 0 36 + 47 + 16 = 99

* Main QTLs (LOD ≥ 3) are indicated by bold type, strong QTLs (3 > LOD ≥ 2) are indicated by underlined type; minor QTLs (2 > LOD ≥ 1.5) are indicated by regular type. ** Chromosomal localization of QTL is indicated in round brackets under the number of the chromosome with the corresponding lettering (QTLs contributed by the Opata 85 variety are indicated in italics, whereas QTLs contributed by W7984 are indicated by regular type). R2 indicates the level of phenotypic variability determined by the given QTL. Level of irradiation at the top leaf layer: 40 ± 0.5 W/m2 PAR and 50 ± 0.5 W/m2 PAR in experiments 1 and 2, respectively. Temperature mode (day/night): 24–25°C/19–20°C and 28–29°C/23–24°C in experiments 1 and 2, respectively. All other growing parameters remained strictly invariable.

some 2D (300 cM) was detected in both experiments. At the same time, for the trait controlling the wax coating on the external leaf side, we detected one QTL on chromosome 2D (300.0 cM for both experiments) and another QTL on chromosome 3D (3.8 and 9.7 cM for experiments 1 and 2, respectively). However, relatively low LOD values for the QTL localized on chromosome 3D (1.99 and 2.00) indicate insufficient expressivity of this trait. The maximum level of phenotypic variability for the trait controlling the wax coating on the stem in experiments 1 and 2 was 55.19 and 57.15%, respectively. For traits controlling the color of ligule and auricles, we observed a different level of identification of the corresponding QTLs. In both experiments, the ligule color was determined by two QTLs localized on chromosomes 1B and 5D. While the localization on chromosome 5D remained unchanged (35.7 cM), the distance for chromosome 1B was 206.5 and 17.6 cM for experiments 1 and 2, respectively. The trait controlling the color of auricles did not demonstrate such stability excepting the number of detected QTLs. For this trait, we also detected two associated QTLs localized on chromosome 1D (221.8 сМ) and 4А (47.3 сМ) in experiment 1 and on chromosomes 2А (224.4 сМ) and 4А (248.6 сМ) in experiment 2. Note that, under the conditions of experiment 1, the QTL allele localized on chromosome 4A was donated by W7984, whereas it was contributed by Opata 85 under conditions of experiment 2. The maximum level of phenotypic variability in experiments 1 and 2 was 20.97% (LOD = 2.91) and 21.85% (LOD = 3.00), respectively. RUSSIAN JOURNAL OF PLANT PHYSIOLOGY

Vol. 64

Two QTLs associated with the type of flowering were detected, these QTLs were localized on chromosomes 3А and 7А (experiment 1) or 6В and 7А (experiment 2). The QTL localization on the 7A linkage group remained stable in both experiments (172.9 cM); the QTL allele was donated by Opata 85. At the same time, QTL alleles localized on chromosomes 3А (39.5 сМ) and 6В (67.9 сМ) were contributed by W7984. In spite of relatively low LOD value, the maximum level of phenotypic variability was 16.04% in both experiments, which indicates a significant contribution of detected QTLs into the manifestation of the associated trait. Such traits as ear texture, beardedness and fragility, the form of ear and floral glumes, the color of kernel, and hard threshing were stable in both experiments concerning the identified QTLs and their genetic distances (Table 1). Alleles of QTLs, associated with the ear texture and fragility, and the QTL allele, detected on chromosome 5A and associated with the hard threshing, were donated by W7984. The alleles of other QTLs associated with the ear beardedness, ear form of ear and floral glumes, and the color of kernel, and also the QTL allele, detected on chromosome 2D for the trait associated with hard threshing were donated by the Opata 85 variety. One should note the maximum values of LOD (6.39 and 10.18 for experiments 1 and 2, respectively) and level of phenotypic variability (42.00 and 38.01% for experiments 1 and 2, respectively) for the QTL localized on chromosome 3D and associated with the color of kernel. One QTL associated with the ear length was detected on chromosome 4A; the genetic distance in experiments 1 and 2 was 124.4 and 240.7 cM, respecNo. 1

2017

54

CHESNOKOV et al.

tively. Similar results were obtained for the trait controlling the number of spikelets in the ear; however, in this case, two QTLs were detected on chromosome 2D in each experiment, and additional QTLs were detected on the 2D and 4D linkage groups. The QTL allele detected on chromosome 4D was contributed by Opata 85, whereas all alleles detected on chromosome 2D were contributed by W7984. In both experiments, the character of manifestation of traits controlling the number of grains per spikelet and number of grains per ear was similar. As a result, for these traits, we identified QTL alleles localized on chromosomes 6D (196.2 cM for both traits in experiment 1) and 1D (161.5 cM for both traits in experiment 2). Note that alleles identified for both traits in experiments 1 and 2 were contributed by W7984 and Opata 85, respectively. Two other pairs of related traits (grain weight per ear and weight of 1000 grains) also demonstrated a similar character of their manifestation. In experiment 2, we detected a QTL allele associated with these traits and localized on chromosome 7D (60.6 cM). At the same time, we detected in experiment 1 a QTL localized on the same chromosome and at the same distance but associated only with the weight of 1000 grains. In the case of the trait controlling the grain weight per ear, only one QTL was detected in experiment 1; this QTL was localized on chromosome 2B (216.9 cM), and its LOD value was 3.91. Note that we additionally detected a QTL associated with the weight of 1000 grains and located on chromosome 3D (145.6 сМ) in experiment 1, whereas an additional QTL associated with the grain weight per ear was identified on chromosome 4A (80.2 сМ) in experiment 2. The maximum level of phenotypic variability for the grain weight per ear was observed in experiment 1 (46.29%), whereas it was observed in experiment 2 (45.93%) for the weight of 1000 grains. Such traits as the grain glassiness and the number of ears were stable in both experiments concerning the number and localization of detected QTLs. For grain glassiness, both experiments revealed one associated QTL on chromosome 6A (35.2 cM for both experiments), whereas the associated QTL was revealed on chromosome 5A (74.4 and 63.8 cM for experiments 1 and 2, respectively) for the number of ears. The maximum level of phenotypic variability for the grain glassiness was observed in experiment 1 (31.90%), whereas for the number of ears the maximum value was observed in experiment 2 (28.36%). The performed one-factor ANOVA analysis of the obtained experimental variances (Table 2) revealed some traits that did not vary and remained stable in both experiments. Note that, according to common rules [25], even though the number of analyzed lines was different for each trait due to the difference in the trait expressivity in the studied RILs under different growing conditions, the significance at p < 0.05 indi-

cates that traits with such significance parameters vary; the closer the significance level to zero, the higher the variability. Other traits for which the calculated significance level р of the variance ratio F was equal or exceeded 0.05 were stable. Differences in the trait expressivity also explain the dispersion of the degrees of freedom of the remaining variation. In general, according to performed calculations, the manifestation of only 9 of 30 evaluated traits varied depending on the experimental conditions. It is interesting that these nine traits are associated mainly with the productivity, i.e., they provide the best possible conditions for the plant reproduction and support of further generations. For example, the trait controlling the duration of the “sprouting−booting” period is responsible for a very important phase of the initial plant development, and the manifestation of this phase will depend on the ecogeographical conditions under which this trait will be realized. The number of spikelets in the ear, the grain weight per ear, weight of 1000 grains, and the number of ears are directly associated with the grain productivity that provides not only a high agronomical value of genotypes, which are stable concerning these traits under adverse conditions, but also the possibility of plants to realize their productivity potential required for the reproduction. The ear fragility, grain glassiness, and wax coating on axial and abaxial leaf sides represent morphophysiological traits probably associated with protective and adaptive functions realized in the course of plant vegetation. None of the other 21 traits studied depended on the experimental conditions, and their manifestation was stable in both experiments (Table 2), which indicates that the manifestation of these traits does not depend (physiologically or genetically) on the illumination and temperature regimes used in this study. DISCUSSION The ascertainment of the nature of the ecogenetical and ecophysiological interaction between the genotype and environment is one of those fundamental tasks whose solution is possible via the ascertainment of the mechanisms of inheritance of genetic components of the systems providing the individual and population fitness. The phenotypic manifestation of quantitative traits obviously depends on both genetic and physiological components influencing on the formation of the M−V (morphology−viability) system of the individual or population; in the “genotype−environment” interaction, this system is determined by a genotype. A correlative ranking of genotypes usually varies under different environmental conditions, and their interaction can be complex [26]. Many quantitative traits in wheat, including the grain yield, flowering period, and resistance to various diseases demonstrate a reliable variability of their manifestation in the “genotype−environment” interaction [5, 10–12]. Classic investigations of quantitative traits usually average a


Vol. 64

No. 1

2017


55

Table 2. Results of a single-factor analysis of variance* Trait designation** VSB*** VSH VSM StWB PH StLuI StNS LFP LFL LFW LWBi*** LWBo*** LLC AuC ChF SpT SpWB SpAwn SpBR*** GS KC DifThC SpL NSpt*** NseSpt NseSp GMSp*** TGW*** GrT*** NStS***

Number of degrees of freedom, d.f.

Mean square deviation (variance), MS

F—variance ratio

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

516.73

9.79

33.95 65.59 0.01 75.26 12.28 0.37 6.88 5.33 0.10 9.26 9.30 0.02 0.04 0.62 1.99 3.09 0.04 1.77 0.98 0.40 3.15 0.04 53.30 0.13 35.24 205.94 802.78 2.50 24.86

0.30 0.47 0.00 1.06 0.74 2.34 3.13 0.45 2.19 5.03 4.80 0.12 0.06 0.57 1.12 0.59 0.07 9.37 0.47 1.82 0.93 0.02 10.55 0.86 0.70 128.75 28.03 4.68 5.01

Remaining variation (error) p—significance 0.002 0.587 0.496 0.970 0.305 0.391 0.128 0.078 0.501 0.141 0.026 0.030 0.729 0.800 0.453 0.292 0.442 0.798 0.003 0.493 0.179 0.336 0.892 0.001 0.357 0.403 0.000 0.000 0.032 0.026

d.f.

MS

227 200 188 203 184 184 184 200 200 200 205 205 205 203 200 200 200 200 192 192 155 183 183 183 146 146 146 146 146 183

114.95 140.90 7.52 71.21 16.58 0.16 2.20 11.75 0.05 1.84 1.94 0.19 0.70 1.10 1.78 5.22 0.65 0.19 2.08 0.22 3.38 2.07 5.05 0.15 49.99 1.60 28.64 0.53 4.96

52.81

* Factor represents the experimental variant. ** Designation of traits is described in detail in Table 1. *** Traits manifesting the variation depending on the experimental cultivation conditions.

physiological and genetic interaction between the genotype and environment and take into account mainly the action of the whole genome than its individual loci or QTLs [27]. At the same time, the majority of modern molecular-genetic studies of the above-mentioned interaction and QTL mapping are confined by the measurement of the effect of seasonal influences on a mapping population and do not cover their study in different ecogeographical areas [28, etc.]. An alternative variant includes the study of different mapping populations in similar growing conditions, as it was RUSSIAN JOURNAL OF PLANT PHYSIOLOGY

Vol. 64

done for barley [29]. There are also some studies in which the QTL mapping was performed for the influence of any chemical additives (for example, fertilizers [30]) on the growth and development of plants that determines the ecogenetic “genotype−environment” interaction from another point of view, since it does not consider the basic physical factors, such as illumination, temperature, humidity, etc. In our studies, we first tried to determine the quantity and exact chromosomal localization of QTLs involved in the physiological and genetic process of No. 1

2017

56

CHESNOKOV et al.

realization of complex agronomically valuable traits in bread wheat (Triticum aestivum L.) manifesting under controlled conditions of the agroecobiological testing ground. The distinctive feature of the regulated agroecosystem is the possibility to change individual components of the environment, while other influencing factors remain invariable. Obviously, the phenotypic manifestation of a trait depends on both physiologogenetic components and environmental factors influencing on the plant development. According to the earlier published model of the ecogenetic control of productivity traits [1], the number and range of genes determining the average value and genotypic dispersion of quantitative traits is determined by a limiting factor (LIM factor) of the environment. Any change of the LIM factor causes a change in the range of genes controlling a quantitative trait. It seems that such change in the range of genes determining a complex trait represents a “top floor” of epigenetic phenomena in the ontogenesis of plants. Note that the role of a physiological component providing the morphophysiological manifestation of a trait is very important. The obtained results clearly demonstrate that the localization of QTLs determining the manifestation of the studied traits remains stable under fixed environmental conditions; at the same time, any changes in certain environmental conditions result in changes in the localization of QTLs involved in the realization of some traits. In our experiments, according to the results of one-factor ANOVA analysis, the change of the temperature and illumination regimes caused variations in the manifestation of only nine out of 30 traits (30%). Note that, among these nine traits, four were directly associated with the grain productivity that determined not only the agronomical value of these traits but also their importance for the survival and expansion of a species. Four other traits are associated with protective and adaptive functions realizing during a vegetation, and one more trait is realized at the stage of initial growth and development initiating the realization of a whole cascade of physiological and genetic mechanisms providing the maximum possible plant productivity under given environmental conditions. The performed QTL mapping allowed us to establish the distribution of identified loci on different linkage groups (Table 1). For example, main QTLs associated with the grain productivity formed clusters on chromosomes 2B, 2D, 3D, 6D, 7D, and 5A under conditions of experiment 1 and on chromosomes 1D, 2D, 4D, 7D, 4A, and 5A under conditions of experiment 2. QTLs associated with the wax coating of axial and abaxial leaf sides were localized on chromosomes 2D and 3D, which correlated with the localization of QTLs associated with the wax coating on stems and ears. Obviously, these blocks of coadapted genes associated with wax coating were donated by Aegilops tauschii Coss., a carrier of the D genome. This species also was the donor of a hard-threshing trait, whose QTL was detected on chromosome 2D at a position close to

that of QTLs associated with the wax coating on the stems, ears, and axial /abaxial leaf sides. The contribution of the 2D linkage group, including the block of agronomically valuable genes involved in the formation of some physiological and morphophysiological traits (such as the wax coating), associated with the realization of a potential productivity is significant, since genetic determinants of traits manifesting at the late stages of development of higher plants are often linked with genes determining the growth and viability of the organism at the early developmental stages (M−V system). Such systems were revealed in homologous chromosomal segments in related species of the genera Gossypium, Lycopersicon, Triticum, Phaseolus, etc. [11, 12]. In addition, as was shown earlier, some genes controlling the same or correlated traits can be linked in a block or localized in different linkage groups (or chromosome arms), and their activation can be controlled by a coordinating gene. Therefore, we should not consider chromosomal loci as a mechanical linkage of genes; they rather represent a certain degree of the organic ordering of genes, a group of functionally associated genes, or a block of coadapted genes [10–12]. It seems that the same concept can be used in relation to QTLs associated with the grain productivity traits whose manifestation is controlled mainly by the loci of chromosomes 5A and 7D and with traits determining the duration of the certain periods of vegetative development (“sprouting−booting,” “sprouting−earing,” and “sprouting−ripening”) whose QTLs are localized not only on chromosome 5A but also on 6A or 7D linkage groups (depending on the experimental conditions). To reveal the effects of interaction between the genotype and environment, we used complementary QTL and one-factor ANOVA approaches. Note that, though the QTL analysis evaluated the stability of the identification of chromosomal loci, which determine the manifestation of the studied traits, on linkage groups, whereas the one-factor ANOVA analysis evaluated the variances of the manifestation of the same traits, both methods used in this study demonstrated complementary results. It seems that the reason for such a phenomenon is explained by the use of the same initial data for both statistical approaches and also by the use of the maximum likelihood criterion and the determination of a statistical reliability in each of the approaches used. However, QTL analysis has some advantages as compared with the one-factor ANOVA, especially because of the use of the RILs from the ITMI mapping population as the experimental material. Since these lines contain a lot of molecular markers [3], their use provides not only the possibility to detect positions of QTLs responsible for the manifestation of any trait and to reveal the level of phenotypic variability determined by each of identified QTLs but also the possibility to reveal molecular markers genetically linked with detected QTLs [5, 10– 12]. Such information provides the possibility to


Vol. 64

No. 1

2017


screen collections of breeding or any other material, to control the QTL introgression into other species, varieties, and lines, and, finally, to obtain the required QTL combinations responsible for a certain level of a trait manifestation. Since the QTL manifestation can be dependent or independent of the environmental conditions, then, if a researcher is interested in environmentally dependent QTLs, he can choose the growing conditions providing the manifestation of the required traits. On the other hand, if a researcher is interested in independent QTLs, this provides a chance to obtain a new variety or line with a high plasticity and, therefore, to expand the area of a possible use of this line/variety. Thus, the results of our study allow us to conclude that (1) manifestation of QTLs detected under controlled conditions of a regulated agroecosystem can depend or be independent on environmental conditions, but the evaluated quantitative traits interact and correlate between themselves, (2) one chromosomal locus can contain several QTLs determining two or more traits and forming the blocks of evolutionary coadapted genes, and (3) the obtained results can be used for the study of physiological and genetic mechanisms of realization of the studied traits and also for the marker-assisted selection of wheat. The detailed understanding of QTL effects with allowance for their stability for a certain ecological zone and the revealed correlations between the manifestation of some QTLs and temperature and illumination regimes create the necessary prerequisites for the further analysis of the identified correlations and revealing of the interaction of the “QTL−environment” type under natural growing conditions (in relation to controlled conditions of regulated agroecosystems). The further use of such information is possible under certain ecogeographical conditions of realization of genetic determinants controlling agronomically valuable traits in bread wheat. ACKNOWLEDGMENTS This study was partially supported by the Russian Foundation for Basic Research, project no. 16-0400311-a.

5.

6.

7.

8.

9.

10.

11.

12.

REFERENCES 1. Dragavtsev, V.A., Litun, P.P., Shkel’, I.M., and Nechiporenko, N.N., The model of ecological and genetic control of quantitative traits of plants, Dokl. Akad. Nauk SSSR, 1984, vol. 274, no. 3, pp. 720–723. 2. Tanksley, S.D., Mapping polygenes, Annu. Rev. Genet., 1993, vol. 27, pp. 205–233. 3. Ganal, M.W. and Röder, M.S., Microsatellite and SNP markers in wheat breeding, Genomics-Assisted Crop Improvement, Vol. 2: Genomics Applications in Crops, Varshney, R.K. and Tuberosa, R., Eds., New York: Springer-Verlag, 2007, pp. 1–24. 4. Nelson, J.C., van Deynze, A.E., Sorrells, M.E., Autrique, E., Lu, Y.H., Merlino, M., Atkinson, M., RUSSIAN JOURNAL OF PLANT PHYSIOLOGY

Vol. 64

13.

14.

57

and Leroy, P., Molecular mapping of wheat. Homoeologous group 2, Genome, 1995, vol. 38, pp. 516–524. Börner, A., Schumann, E., Fürste, A., Göster, H., Leithold, B., Röder, M.S., and Weber, W.E., Mapping of quantitative trait loci determining agronomic important characters in hexaploid wheat (Triticum aestivum L.), Theor. Appl. Genet., 2002, vol. 105, pp. 921– 936. Faris, J.D., Anderson, J.A., Francl, L.J., and Jordahl, J.C., RFLP mapping of tan spot resistance genes in wheat, Proc. 5th, 6th Public Workshop Int. Triticeae Mapping Initiative, Genetic Resources Conservation Program, Division of Agriculture and Natural Resources, McGuire, P.E. and Qualset, C.O., Eds., University of California, 1996, p. 179. Singh, R.P., Nelson, J.C., and Sorrels, M.E., Mapping Yr28 and other genes for resistance to stripe rust in wheat, Crop Sci., 2000, vol. 40, pp. 1148–1155. Pshenichnikova, T.A., Ermakova, M.F., Chistyakova, A.K., Shchukina, L.V., Börner, A., and Röder, M., Molecular mapping of loci associated with indicators of the quality of soft wheat, S.-kh. Biol., 2006, no. 5, pp. 41–47. Pshenichnikova, T.A., Ermakova, M.F., Chistyakova, A.K., Shchukina, L.V., Berezovskaya, E.V., Röder, M., and Börner, A., Mapping of the quantitative trait loci (QTL) associated with grain quality characteristics of the bread wheat grown under different environmental conditions, Russ. J. Genet., 2008, vol. 44, pp. 74–84. Chesnokov, Yu.V., Pochepnya, N.V., Börner, A., Lowasser, U., Goncharova, E.A., and Dragavtsev, V.A., Ecological–genetic organization of plant quantitative traits and mapping of the loci determining agronomically important traits in soft wheat, Dokl. Biochem. Biophys., 2008, vol. 418, pp. 36–39. Chesnokov, Yu.V., Pochepnya, N.V., Kozlenko, L.V., Sitnikov, M.N., Mitrofanova, O.P., Syukov, V.V., Kochetkov, D.V., Lowasser, U., and Börner, A., Mapping of QTLs determining the expression of agronomically and economically valuable features in spring wheat (Triticum aestivum L.) grown in environmentally different Russian regions, Russ. J. Genet.: Appl. Res., 2013, vol. 3, no. 3, pp. 209–221. Chesnokov, Yu.V., Sitnikov, M.N., Schumlyanskaya, N.V., Kocherina, N.V., Goncharova, E.A., Kozlenko, L.V., Syukov, V.V., Kochetkov, D.V., Lohwasser, U., and Börner, A., Catalogue of Recombinant Inbred Lines of Mapping Population ITMI of Soft Spring Wheat Triticum aestivum L. (Ecological and Geographical Trials and QTL Mapping), St. Petersburg: Vavilov Inst. Plant Genet. Resour., 2014. Chesnokov, Yu.V., Goncharova, E.A., Sitnikov, M.N., Kocherina, N.V., Lohwasser, U., and Börner, A., Mapping QTL for water regime in spring bread wheat, Russ. J. Plant Physiol., 2014, vol. 61, pp. 834–841. Paterson, A.H., Damon, S., Hewitt, J.D., Zamir, D., Rabinowitch, H.D., Lincoln, S.E., Lander, E.S., and Tanksley, S.D., Mendelian factors underlying quantitative traits in tomato: comparison across species, generations, and environments, Genetics, 1991, vol. 127, pp. 181–197.

No. 1

2017

58

CHESNOKOV et al.

15. Stuber, C.W., Lincoln, S.E., Wolff, D.W., Helentjaris, T., and Lander, E.S., Identification of genetic factors contributing to heterosis in a hybrid from two elite maize inbred lines using molecular markers, Genetics, 1992, vol. 132, pp. 823–839.

22.

16. Ungerer, M.C., Halldorsdottir, S.S., Purugganon, M.D., and Mackay, T.F., Genotype-environmental interactions at quantitative trait loci affecting inflorescence development in Arabidopsis thaliana, Genetics, 2003, vol. 165, pp. 353–365.

23.

17. Eriksen, L., Borum, F., and Jahoor, A., Inheritance and localisation of resistance to Mycosphaerella graminicola causing septoria tritici blotch and plant height in the wheat (Triticum aestivum L.) genome with DNA markers, Theor. Appl. Genet., 2003, vol. 107, pp. 515– 527. 18. Singh, A., Knox, R.E., de Pauw, R.M., Singh, A.K., Cuthbert, R.D., Campbell, H.L., Shorter, S., and Bhavani, S., Stripe rust and leaf rust resistance QTL mapping, epistatic interactions, and co-localization with stem rust resistance loci in spring wheat evaluated over three continents, Theor. Appl. Genet., 2014, vol. 127, pp. 2465–2477. 19. Habash, D.Z., Bernard, S., Schondelmaier, J., Weyen, J., and Quarrie, S.A., The genetics of nitrogen use in hexaploid wheat: N utilisation, development and yield, Theor. Appl. Genet., 2007, vol. 114, pp. 403–419. 20. Parent, B., Shahinnia, F., Maphosa, L., Berger, B., Rabie, H., Chalmers, K., Kovalchuk, A., Langridge, P., and Fleury, D., Combining field performance with controlled environment plant imaging to identify the genetic control of growth and transpiration underlying yield response to water-deficit stress in wheat, J. Exp. Bot., 2015, vol. 66, pp. 5481–5492. 21. Ermakov, E.I., Makarova, G.A., and Nerusheva, G.V., Programmed obtaining of transgressive on term earing wheat lines in controlled agroecosystem, Metodicheskie rekomendatsii (Methodical Recommendations),

24.

25. 26. 27. 28.

29. 30.

St. Petersburg: Agrofiz. Inst., Ross. Akad. S-kh. Nauk, 2002. Panova, G.G., Dragavtsev, V.A., Kanash, E.V., Arkhipov, M.V., and Chernousov, I.N., Scientific and technical basics of optimization of the production process in a controlled agroecosystem, Agrofizika, 2011, no. 1, pp. 29–37. Haldane, J.B.S., The combination of linkage values, and the calculation of distance between linked factors, J. Genet., 1918, vol. 8, pp. 299–309. Lander, E.S., Green, P., Abrahamson, J., Barlow, A., Daly, M.J., Lincoln, S.E., and Newburg, L., MAPMAKER: an interactive computer package for constructing primary genetic linkage maps of experimental and natural populations, Genomics, 1987, vol. 1, pp. 174–181. Fisher, R.E., Statisticheskie metody dlya issledovatelei (Statistical Methods for Researchers), Moscow: Gosstatizdat, 1958. Allard, R.W. and Bradshaw, A.D., Implications of genotype-environment interactions in applied plant breeding, Crop Sci., 1964, vol. 4, pp. 503–508. Merezhko, A.F., Genetic analysis of quantitative traits for problem solving in plant selection, Genetika, 1994, vol. 30, pp. 1317–1325. Bezant, J., Laurie, D., Pratchett, N., Chojecki, J., and Kearsey, M., Mapping QTLs controlling yield and yield components in spring barley (Hordeum vulgare L.) cross using marker regression, Mol. Breed., 1997, vol. 3, pp. 29–38. Thomas, W.T.B., Prospects for molecular breeding of barley, Ann. Appl. Biol., 2003, vol. 142, pp. 1–12. Hirel, B., Le Gouis, J., Ney, B., and Gallais, A., The challenge of improving nitrogen use efficiency in crop plants: towards a more central role for genetic variability and quantitative genetics within integrated approaches, J. Exp. Bot., 2007, vol. 58, pp. 2369–2387.


Translated by N. Statsyuk

Vol. 64

No. 1

2017