SELECTION FOR DAYS TO HEADING UNDER NORMAL IRRIGATION AND DROUGHT STRESS CONDITIONS WITH MONITORING TaELF3 GENE EXPRESSION IN BREAD WHEAT
Rasha E. Mahdyaand Ameer E. Elfarashb
a Agron. Dept., Fac. Agric., Assiut Univ., Assiut, Egypt
b Genetics Dept., Fac. Agric., Assiut Univ., Assiut, Egypt
E: mail:
ABSTRACT
The present article was carried out for four seasons to study the relative merits of pedigree selection for days to heading (DH) restricted by grain yield/plant (GY/P) under drought stress and normal irrigation environments to compare three cycles of selection started in the F2 to the F5 with one cycle started in the F4-generation, and to study the sensitivity of the selected families to environments,in addition to monitoring TaELF3 gene expression in bread wheat.The reduction % in DH in the F2 caused by drought stress was 5.1 compared to 2.8 and 3.29% for the two parents Sids4 and Giza164; respectively. The phenotypic (PCV)and genotypic(GCV) coefficients of variability decreased from 5.15 and 4.08% in the F2 to 1.72 and 1.64% after cycle3 under drought stress, and from 6.14 and 2.1% in the F2 to 3.73 and 3.71% after cycle3 under normal irrigation; respectively. The realized heritability increased from 68.15% in cycle1 to 70.5% in cycle3 under drought stress, and from 80.45 to 88.40% under normal irrigation for the respective cycles. The observed genetic gain (OG) in DH of C3 drought selected families was -5.31 and -5.06% (P≤0.01) of the bulk sample, compared to -7.29 and -6.24% for normal irrigation selected families, when evaluation was practiced under drought stress and normal irrigation; respectively.Realized heritability of DH under normal irrigation (antagonistic selection) increased from 68.15 % in cycle1 to 70.5% in cycle3 under drought stress, and from 80.45 to 88.40% under normal irrigation for the respective cycles. The observed genetic grain (OG)in DH of C3 drought selected families was -5.31 and -5.065 (P≤0.01) of the bulk sample, compared to -7.29 and -6.24% for normal irrigated selected families, when evaluation practiced under drought stress and normal irrigation; respectively.Selection for DH under normal irrigation (antagonistic selection) was better than under drought stress (synergistic selection). The results indicated that one cycle of selection for DH started in F4was equal or better than three cycles started in F2, especially under normal irrigation. The results suggested delaying selection for DH to the F4 or F5 till the recombinants reached acceptable level of homozygosity to save efforts and expenses. The correlated gain in GY/P after three cycles under drought stress was 10.92 and 10.0%, and was 14.94 and 12.36% after one cycle from the F4, while it was 22.18 and 21.20% after three cycles under normal irrigation, and 24.97 and 23.97% after one cycle from F4, from the bulk sample and the better parent; respectively. The other correlated gains were discussed. The relative merits of selection for DH indicated that selection for DH under normal irrigation (antagonistic selection) was better than selection under drought stress (synergistic selection) in changing the mean.qRT-PCR was used to test the expression levels of the TaELF3 gene, responsible for the early flowering in T. aestivum, under both normal irrigation and drought stress conditions, confirming that drought stress enhanced the expression TaELF3 gene and reducedthe required days to heading.
Key words: Pedigree selection, Days to heading, Drought stress, Triticum aestivum, Realized heritability, Sensitivity to environment,qRT-PCR.
INTRODUCTION
Earliness in field crops is an important goal in breeding programs. Early matured cultivars are preferred to escape drought, heat, diseases, pests and other stress injuries that occur at the end of the growing season (Menshawy 2007). In Egypt, planting cotton after wheat necessitate development of early wheat cultivars.Early cultivars could be suitable for northern coast cultivation, in which precipitation of rainfall ranged from 150 to 200mm/year with supplement of one irrigation. Heading date in wheat is an early character to identify, and can be modified by selection (Allard and Harding 1963, Avery et al 1982 and Frederickson and Kronstad 1985). Time of anthesis and subsequent period from anthesis to maturity are the main determinants of maturity in wheat (Duguid and Brule-Babel 1994). The duration and rate of grain filling also determines final grain weight. Iqbal et al (2007) found positive genetic correlation between grain and maturity (0.69), rate (0.78) and duration (0.99) of grain fill and harvest index (0.55). Early-heading cultivars completed a greater fraction of the grain filling earlier in the season when air temperatures were lower and generally more favorable (Tewoldeet al 2006).
Keim and kronstad (1979) proposed that an ideal cultivar for stress-prone environments should have high yield in the most severely stressed environments expected, and a strong response (b<1) to more favorable environments.
Jinks and Connolly (1973 and 1975) studied the relationship between selection environment and environmental sensitivity (stability) in Schizophyllum commune. They concluded that the sensitivity of selection to environment was reduced if selection and environment effects were in opposite direction. The effect of selection environment on environmental sensitivity was generalized by Jinks and Pooni (1982) on the basis that, the genetic control of environmental sensitivity of a selected character and its mean performance were at least partly independent. They reported that the relationship between selection environment and environment sensitivity for plant height in Nicotiana rustica L. was in agreement with Jinks and Connolly model.
Falconer (1990) suggested that antagonistic selection was significantly better than synergistic selection for changing the mean, but there is no theoretical justification for this expectation. Ceccarelli and Grando (1991a) found that, selection environment affects the performance of barely material. The higher stability genotypes were selected under low yielding environment (Ceccarelli and Grando 1991 b). They indicated that, the probability of a crop failure of genotypes selected for high grain yield under high yielding conditions was between 1.8 and 2.7 times higher than for genotypes selected for high grain yield under low yielding conditions.
Kheiralla and El-Defrawy (1994) found that selection for days to heading under water stress environment increased sensitivity, which does not conform to Jinks-Connolly model.
After two cycles of pedigree selection for days to heading under favorable and stress environments Ali (2011) achieved realized gain of -2.19 and -1.85% from the bulk sample under favorable and stress environments; respectively, which conform the Jinks-Connolly model. Mahdy (2012) found that antagonistic selection was better than synergistic selection for days to heading. Mahdyet al (2012) achieved significant (P≤0.01) direct observed gain in days to heading of -7.12% in a population and -7.30% in another population of bread wheat under normal irrigation. Under normal irrigation and water stress, two cycles of selection for days to heading achieved significant observed genetic gain from the bulk sample of -16.63 and -10.17%; respectively, which conform the Jinks Connolly model (Mahdy et al 2015). Zikhaliet al (2015) made the first direct fine mapping of Eps genes. He studied its effect in bread wheat and defined the Eps-D1 genes, and showed their responsibility for the early flowering phenotype in bread wheat. The ELF3 gene was found to be a repressor of flowering time in at least six species and the loss of its function by mutations which result in an early flowering phenotype for wheat. Deletion of large fragments of chromosomes, single genes, or portions of genes caused variation in flowering time (Yan et al 2003; Fu et al 2005; Wilhelm et al 2009; Shitsukawaet al 2007; Distelfeld and Dubcovsky 2010 and Faure et al 2012). Zikhali et al (2015) have used qRT PCR to investigate the expression levels of TaELF3 to confirm that the loss of the D copy did indeed result in reduced total expression. The objectives of the present article were to study; 1) the relative merits of pedigree selection for days to flowering under normal irrigation and drought stress environments. 2) the efficiency of three cycles of selection for days to heading started in the F2 compared with one cycle started in the F4-generation. 3) the sensitivity of the selected lines to drought stress.And 4)theTaELF3 gene expression in bread wheat.
MATERIALS AND METHODS
The experiments were carried out during the four seasons; i.e.2012/13, 2014/15, 2015/16, 2016/17 at Faculty of Agriculture Experimental Farm, Assiut University, Egypt (latitude=27.178̊, Longitude=31.185̊). The soil texture is clay. The genetic materials were F2, F3, F4 and F5 generations of bread wheat population (Triticum aestivum L.) stemmed from the cross Giza164 X Sids 4 cultivars. The pedigree of the parents and origin were as follows:
Parent / Pedigree / OriginGiza164 / KVZ/Buha’’S’’//K al/Bb / Egypt
Sids4 / Maya’’S’’Man(S)//CMH74.AS92/3/Giza157-2 / Egypt
Planting date
Season / Date / Generation / Experimental designIrrigation / Drought
2012/13 / 5/12/2012 / F2 / Non-replicated plots / Non-replicated plots
2014/15 / 11/12/2014 / F3 / RCBD with 3-Reps / RCBD with 3-Reps
2015/16 / 17/12/2015 / F4 / " / "
2016/17 / 5/12/2016 / F5 / " / "
Irrigation
The experiment under normal irrigation in the four seasons received planting irrigation and four irrigations throughout the growing season. However, the experiment under drought stress received planting irrigation and only one irrigation three weeks later.
Fertilization
In all experiments, super phosphate (, 15.5%) was added during land preparation at a rate of 150 kg/fed (feddan=4200m²). Nitrogen fertilization in the form of ammonium nitrate (33.5% N) was added at a rate of 80kgN/fed in one dose before the first irrigation.
Two types of selection for days to heading were performed. The first was three cycles of selection started from F2 to F5 generation under normal irrigation and stressed environments. The second was one cycle started in the F4 under both environments and evaluated in the F5 generation.
A-Season 2012/13; the F2 seeds were grown under normal irrigation and under drought stress in non-replicated plots, in rows 3m long, 30 cm apart and 15cm between seeds with a row. The parents were grown in separate plots.
The data were recorded on 440 guarded plants for each environment. At harvest, the bulk sample was formed from mixing equal number of grains from each F2 plant. The recorded data in all generations were days to heading (DH), plant height; cm(PH),spike length; cm(SL), number of spikes/plant (NS/P), biological yield/plant; g(BY/P), grain yield/plant; g(GY/P), harvest index(HI%) and 100-grain weight; g(100-GW). The earliest 40 plants gave equal or more than the population mean in GY/P from each environment were saved for the next generation. Furthermore, bulk grains from each 440 plants were saved to grown in pedigree rows of non-replicated experiment in the F3 an F4 in each environment.
B-Season 2014/15(F3-generation); grains of the 40 selected plants from each environment along with the parents and the bulk sample were grown in rows under normal irrigation and drought stress environments in RCBD with three replications. The plot size was single row, 3m long, 30 cm apart and 10 cm between grains within a row. Data were recorded on 20 guarded plants for each family. At the end of the season, the best 20 plants in grain yield from the earliest 20 families were saved from each experiment for the next season. Grains of 440 F3-families were grown in non-replicated experiment under both of irrigation and drought stress environments. At the end of the season, bulk grains from each row (family) were saved without selection.
C- Season 2015/16(F4-generation). Seeds of the selected plants (20plants for each environments) were grown as the previous season. At the end of the season, grains of the best 10 plants in GY/P from the 10 earliest families were saved from each experiment. Furthermore, the best ten plants in grain yield from 10 earliest families under each environment of the 440 F4-non-replicated families were saved for the next generation.
D- Season 2016/17(F5-generation);the ten early plants selected under irrigation + the ten selected under stress environment + 10 selected plants from the non-replicated F4-trial under irrigation + 10 selected plants from the non-replicated F4-trial under stress environment were evaluated under both environments along with the two parents and the bulk sample. Each replication under irrigation or under stress environment included 43 families. Each family was represented by 30 pants in one row 3m long. The data were recorded as the previous seasons.
Statistical analysis
Statistical analysis on plot mean basis and separation mean analysis were performed according to Steel and Torrie(1980). Two analyses of variance were done, the first was for the selected families to calculate PCV%, GCV% and heritability, the second was for the selected families, parents and the bulk sample to perform the significance tests.Heritability in broad sense “H” was estimated as the ratio of genotypic (to phenotypic (p) variance (Walker 1960). Realized heritability was calculated as: h²=R/S (Falconer, 1989), where R=response to selection and S=selection differential.The phenotypic (PCV %) and genotypic (GCV%)coefficients of variability were calculated as outlined by Burton(1952).Drought susceptibility index (DSI) was calculated according to Fischer and Maurer (1978). The sensitivity test was performed as outlined by Falconer (1990). The relative merits of selection in changing the mean was performed according to Falconer (1990) as the ratio:
(change of mean by antagonistic selection) / (change of mean by synergistic selection)
Synergistic selection: selection upward in a good environment or downwards in bad.
Antagonistic selection: selection upward in a bad environment or downwards in good, selection and environment acting in opposite direction on the character.
Gene expression
Expression studies of the TaELF3 gene were carried out as described by Zikhaliet al (2015). Samples were collected from3-week-old plants grown under long days (16h light) at 16–18 °C in the light and 20 °C in the dark period. Two families (early family N0. 84, and late family N0.273) were selected to test TaELF3 gene expression under normal and drought stress conditions. RNA was extracted by using the high pure RNA Isolation Kit (Roche). The purity and concentration of the RNA was determined by gel electrophoresis and spectrophotometry (NanoDrop, Ijsselstein, Netherlands). First-strand cDNA was obtained using First-strand cDNA Synthesis Kit (Amersham Biosciences, GE Healthcare). The qRT-PCR was performed at the Molecular Biology Research Center (MBRU), Assiut University in a Bio-Rad iCycler. The qRT-PCR was performed with Bio-Rad iQ SYBR Green Supermix and adding TaELF3 targeting primers (Fw:GTGGGATCGACAGACCTC, and Rv:CGACGCGTTCCTTCC)
The following cycling parameters were used: 94°C for 3 min followed by 40 cycles of 94°C for 60 s, 55°C for 45 s and 72°C for 60 s, followed by 72°C for 7 min. norm2 (house-keeping gene control) was used to normalize gene expression. The Relative Quantification (ΔΔCT) method was utilized to calculate fold change.
RESULTS AND DISCUSSION
Description of the base population; F2generation
Summary of the characteristics of the F2generation and the two parents under normal irrigation and drought stress environments is shown in Table1. Mean number of days to heading (DH) of the earlier parent Sids4 was 71 and 69 days compared to 88 and 85 days for Giza164 under normal irrigation and drought stress environments; respectively. The difference between the two parents in DH was large (16-17 days). The F2-mean lies between the two parents, indicating nearly no-dominance. The reduction% in DH in the F2 caused by drought stress was 5.1% compared to 2.82 and 3.29% for Sids4 and Giza164; respectively indicating stability of the parents in DH with respect to drought stress.
The coefficients of variability in DH under both environments were low, indicating doubtful response to selection. However, selection for DH depended mainly upon that several plants in the F2 generation showed minimum values of DH near or equal the earlier parent Sids4, indicating feasibility of selection for this character.
Heritability in broad sense of DH was low (21.15%)under normal irrigation, and moderate (62.84%) under drought stress. The predicted genetic advance under selection of 10% earliest plants was 1.70% under normal irrigation, and 5.67% of the mean under drought stress. Ali (2011) noted PCV of 12.2% for DH and broad sense heritability of 95%, Mahdyet al (2012) found PCV of 10.88 and 10.22%, and heritability in broad sense of 89.66 and 67.49% in two base populations. Mahdyet al (2015) reported PCV of 6.47% for DH and heritability of 98.49% with 11.72% predicted genetic advance in percentage of the mean.
The coefficients of variability in plant height were low either under normal irrigation or drought stress conditions. Low variability was obtained in plant height in the F2 accompanied with low heritability (41.67 and 29.98%) and expected genetic advance (5.51 and 4.08%) under normal irrigation and drought stress; respectively. The reduction % in plant height was 4.09, 11.35 and 2.05% for the F2, sids4 and Giza164; respectively. Mohamed (1999) found reduction in plant height under water stress condition. Mahdy (2007) reported reduction in plant height of 20 cultivars of 14.21% evaluated across two years.
Table 1. Means of the studied traits in the F2, parents, heritability in broad sense(H) and genetic advance(GA) under selection of 10%superior plants.Item / Normal irrigation
DH / PH;cm / SL,cm / NS/P / GY/P;g / 100GW;g / HI%
F2
Mean±SE / 81.92
±0.18 / 89.030
±0.32 / 12.57
±.08 / 9.57
±0.20 / 22.56
±0.52 / 5.213
±.026 / 35.0
±0.2
max / 92.00 / 115.00 / 20.00 / 25.00 / 71.20 / 7.48 / 41.95
min / 72.00 / 62.00 / 8.00 / 2.00 / 5.38 / 3.14 / 24.83
PCV% / 6.14 / 6.71 / 6.59 / 25.11 / 48.31 / 7.44 / 11.58
GCV% / 2.10 / 5.30 / 1.64 / 4.08 / 28.59 / 0.44 / 0.58
H % / 21.15 / 41.67 / 61.39 / 66.95 / 35.01 / 42.81 / 1.36
GA / 1.39 / 4.92 / 1.84 / 4.80 / 6.70 / 0.41 / 0.26
GA/Mean% / 1.70 / 5.51 / 14.65 / 50.18 / 38.61 / 7.85 / 1.15
Sids4
Mean±SE / 71
±0.95 / 79.09
±1.16 / 14.67
±0.21 / 5.57
±0.31 / 20.11
±1.71 / 5.67
±0.09 / 36.38
±0.91
max / 75.00 / 90.00 / 17.00 / 8.00 / 34.90 / 6.34 / 42.89
min / 61.00 / 70.00 / 13.00 / 3.00 / 8.83 / 4.58 / 29.07
PCV% / 6.14 / 6.71 / 6.59 / 25.11 / 39.07 / 7.44 / 11.58
G164
Mean±SE / 88.08
±0.36 / 92.76
±0.99 / 10.64
±0.23 / 11.6
±0.60 / 30.77
±1.92 / 5.03
±0.08 / 35.96
±0.77
max / 91.00 / 100.00 / 13.00 / 18.00 / 51.70 / 6.12 / 46.90
min / 85.00 / 85.00 / 9.00 / 6.00 / 14.35 / 4.34 / 28.70
PCV% / 2.02 / 5.35 / 10.81 / 25.98 / 31.28 / 7.99 / 9.41
Drought stress
F2
Mean±SE / 77.74
±0.2 / 85.02
±0.33 / 11.92
±0.1 / 7.18
±0.2 / 15.79
±0.6 / 4.93
±0.05 / 33.60
±0.3
Reduction% / 5.10 / 4.09 / 5.17 / 24.97 / 42.88 / 5.37 / 4.00
max / 90.00 / 100.00 / 18.00 / 20.00 / 50.00 / 5.10 / 37.50
min / 69.00 / 59.00 / 8.00 / 3.00 / 5.00 / 2.10 / 23.20
PCV% / 5.15 / 7.76 / 16.78 / 55.71 / 76.00 / 20.28 / 17.86
GCV% / 4.08 / 4.25 / 12.93 / 41.45 / 45.81 / 17.57 / 12.43
H % / 62.84 / 29.98 / 59.38 / 55.36 / 36.34 / 75.00 / 48.47
GA / 4.41 / 3.47 / 2.08 / 3.89 / 7.65 / 1.32 / 5.10
GA/Mean / 5.67 / 4.08 / 17.48 / 54.13 / 48.46 / 26.70 / 15.19
Sids4
Mean±SE / 69
±0.6 / 70.11
±1.2 / 14.17
±0.20 / 4.81
±0.35 / 18.21
±1.71 / 5.1
±0.10 / 35.48
±0.91
Reduction% / 2.82 / 11.35 / 3.41 / 13.64 / 9.45 / 11.18 / 2.47
max / 70.00 / 85.00 / 16.00 / 6.00 / 30.50 / 6.10 / 39.50
min / 60.00 / 68.00 / 13.00 / 3.00 / 10.20 / 3.90 / 30.20
PCV% / 4.35 / 8.56 / 7.06 / 36.38 / 46.95 / 9.80 / 12.82
G164
Mean±SE / 85.18
±0.34 / 90.86
±01.0 / 10.4
±0.30 / 8.6
±0.67 / 25.81
±2.1 / 4.5
±0.10 / 33.86
±0.81
Reduction% / 3.29 / 2.05 / 2.26 / 25.86 / 16.12 / 10.54 / 1.33
max / 88.00 / 95.00 / 12.00 / 16.00 / 50.40 / 5.50 / 38.50
min / 85.00 / 85.00 / 9.00 / 6.00 / 15.00 / 4.10 / 28.20
PCV% / 2.00 / 5.50 / 14.42 / 38.95 / 40.68 / 11.11 / 11.96
SE; standard error, PCV and GCV%; phenotypic and genotypic coefficients of variability; respectively.
The reduction% caused by drought stress was low for spike length (5.17%), 100 grain weight (5.37%) and harvest index (4%). However, it was high for NS/P (24.97%) and GY/P (42.88%). Kazmi et al (2003) found that ear length was reduced by 36% under water stress. Kheirallaet al (2004) found that skipping irrigation at any stage reduced spike length. Mahdy(2007) noted average reduction in spike length of 6.30%. Mahdyet al(2012) found reduction% of 6.38% for NS/P and 12.10% for grain yield/plant.
Variability and heritability estimates
Family mean squares of DH and the other traits after three cycles of selection from F2 to F5- generation, and one cycle of selection from F4 to F5-generation was significant (P≤0.05 to P≤0.01), either selection practiced under normal irrigation or drought stress, and evaluated under both environments, except for harvest index under drought stress (Table2).The PCV and GCV of days to heading of the selected families decreased rapidly cycle after cycle (Table3). The PCV and GCV decreased from 5.15 and 4.08% in the base population to 3.75 and 3.01% for cycle1, 1.84 and 1.77% for cycle2, and 1.72 and 1.64 for cycle3 under drought stress; respectively. Likewise, under normal irrigation, the PCV dropped from 6.14% in the base population to 5.43, 4.14 and 3.73% for C1, C2 and C3; respectively. The phenotypic and genotypic coefficients of variability were higher under normal irrigation than under drought stress. This could be due to that water stress induced flowering, and minimized the differences among families; and the early heading plant/family is easily to identify in the field, then; the differences among selected families decreased cycle after cycle. The little discrepancies between PCV and GCV resulted in high estimates of broad sense heritability of 91.09% under stress and 98.92% under normal irrigation after three cycles of selection.Likewise, the realized heritability increased from 68.15% in C1 to 70.5% in C3 under drought stress, and from 80.45 to 88.40% under normal irrigation for the respective cycles. Furthermore, heritability estimates after one cycle from F4/F5 generation were high under both environments. The very large magnitude of broad sense heritability could mainly be due to that the evaluation of the selected families at one site for one year inflated the families mean squares by the confounding effects of families x locations, families x years and families x locations x years interactions. Consequently, a very large genotypic variance could be obtained from the expected mean squares.These results are in agreement with many researchers; Kheiralla 1993, Ismail1995, Ismail et al 1996, Zakaria et al 2008, Ali2011, Mahdyet al 2012 and Mahdy2012.