Differential Responses to Yellow-Rust Stress Assist in the Identification of Candidate Wheat (Triticum aestivum L.) Genotypes for Resistance Breeding

Abstract

Yellow rust (YR) globally affects wheat crops. It may turn into an epidemic, resulting in significant yield losses if the environment is suited to YR spread. The provision of resistant wheat cultivars is a sustainable protection strategy against YR. The current study aimed to use a combination of classical analytical tools to identify potential wheat lines through screening under YR stress for utilization in YR-resistance breeding. A total of 14 parents, which included 9 lines and 5 testers, were formed into 45 cross combinations via line × tester mating format. The tested germplasm was grown in a triplicate randomized full-block design, under optimal and yellow rust (YR)-stress conditions. Data were recorded on different morphological, physiochemical, yield and component traits at appropriate crop stages. A pre-combining ability analysis revealed significant inter-genotype variations. A combining ability study identified non-additive gene action in the inheritance of most of the investigated traits. Four potential parents (PR128, AN179, KS17 and WD17) and two crosses (PR128 × WD17 and AN179 × KS17) retained higher combining ability values for yield traits under YR-stress. A cluster analysis based on the overall performance found divergent classes among the screened genotypes. The clustering of different genotypes was shifted under YR-stress, which suggests variable genotypic response. Through factor analysis, we assessed and confirmed genotypes performing consistently under YR-stress. The identified genotypes may be used for disease-resistance breeding in wheat. Based on their positive correlation with grain yield, we suggest the use of peduncle length and tillers per plant as phenotypic markers for wheat selection and breeding. The knowledge base generated through the current study will add to the ongoing research on sustainable wheat breeding program.

1. Introduction

Yellow rust (YR) caused by Puccinia striiformis is one of the most challenging diseases affecting long-term wheat production [1]. YR-stress incurs higher damages to cereal crops worldwide as compared to leaf and stem rust [2]. Yellow rust can decrease wheat harvest by 10–70% in case of susceptible wheat variety, and if environmental conditions are conducive for disease development [3]. Under significant disease stress, grain-yield losses in wheat are reported to be 45–60%, with losses reaching 100% if infection develops early in the growth season and if conditions remain favorable [4]. It has also been claimed that for every 10 days of increased disease pressure, grain yield may drop by 6% [3].

YR affects the wheat morphology, and yield components at different crop stages. YR-stress may also reduce wheat-grain yield and affect other yield component traits such as the number of grains per spike, grain size, grain weight, plant biomass, vigor, and plant growth [5]. Plant developmental phases such as tillering, anthesis, grain filling, and maturity are also affected by YR stress [6]. Early tillering and jointing anthesis stage, reduced grains number per spike, low grain weight and reduced kernel quality are the characteristic features in development of YR-stress [7,8,9].

High YR-stress can also damage several physiological processes in wheat, including photosynthetic rate, transpiration activity, and the grain development rate. The most important physiological processes that determine wheat-grain production is photosynthesis, which is inextricably associated with plant pigments i.e., chlorophyll A and B concentrations at required stage. The host YR-resistance against the prevalent strain is also influenced by the pathotypes’ activeness, which determines disease valuation of the stress genotype [10]. The reduced plant efficiency during YR-stress is mainly due to its inability to photosynthesize in the scarcity of chlorophyll with aberrations in the leaves structure, which is the apparent YR-stress signal in wheat [5,11]. Resultantly, canopy temperature of the plant increases, leading to reduced grain yield with small and shriveled grains [12].

Host-plant resistance is the most efficient and environmentally friendly method to control YR. The improvement of YR-resistant and agronomically superior genotypes need genetically diverse and compatible parental lines and best crosses for selection [12]. It is essential that the genetic diversity of cultivated germplasm for YR stress tolerance be evaluated in order to develop YR resistant wheat genotypes. Different genotypes at their successive developmental stages of the crop respond differently to the disease pressure. Genotypic variance in response to stress is mainly due to the variations in resistant genes. Gene introgression requires an understanding about the nature of such genes’ action and their heritability estimation in advance lines and their early generations. Heritability estimates provide accurate information about the transmissibility of traits and aids breeders in making desirable selections [13]. Significant genetic resistance to YR disease stress has been documented in different wheat genotypes [14]. However, the development of resistant cultivars is being threatened by continuous mutations in and adaptation of P. striiformis affecting sustainable crop production [15].

Kempthorne created the line-tester mating design in 1957, which offers dependable data regarding the general and specific combining abilities of parents and their hybrid combinations in resultant generations [16]. Due to its importance in identifying and choosing parents who have the greatest chance of passing on desirable characteristics to the following generations, the design has been widely employed and is still used in quantitative genetic studies in wheat [14,17,18]. It is crucial to carefully investigate the breeding material in autogamous crops such as wheat, when the goal is to develop pure-line cultivars in order to assess variations. Combining ability estimation also establishes the breeding value of parental lines in generating successful wheat-cross combinations for yield, yield-associated traits, and physiological and biochemical parameters under stress [19]. In this regard, line × tester is a modified top cross scheme, which utilizes numerous testers as opposed to the conventional top cross scheme that utilizes only one tester.

The main aim of this study was to develop resistant wheat germplasm against yellow rust. Line × tester mating design was used to forecast the performance of wheat lines in cross combinations for grain yield and component qualities under optimal and YR-stress conditions. We also screened wheat genotypes for morphological and physiochemical diversity. The estimation of correlation among different plant parameters and their association with genotypes were also assessed.

2. Materials and Methods

2.1. Breeding Material, Experimental Design and Study Location

The current study was conducted during years 2019–2020 and 2020–2021 at the Cereal Crops Research Institute (CCRI), Pirsabak Nowshera, Khyber Pakhtunkhwa, Pakistan (latitude 34.01681757, Longitude: 72.04362582) at an elevation of 945 feet from the mean sea level. The climate of Nowshera during wheat-growth season (Oct-May) is mostly hot and warm, with an average precipitation of about 2.9 mm during the length of growing season. Nowshera’s weather is an ideal screening spot for yellow rust. The soil texture is sandy loam (sand 30%, loam 70%) with pH 8.1. The experimental material consisted of 14 wheat parents and 45 crosses were evaluated under disease free, hereafter referred as “optimal” and yellow rust inoculated conditions, hereafter referred as “YR-Stress”, for their morphological, physiological, biochemical, yield and its component traits (

Table 1. List of wheat breeding material i.e., lines (L), testers (T) and their crosses.

	S. No.	Parentage	Abbr.	S. No.	Parentage	Abbr.	S. No.	Parentage	Abbr.
Parents	1	PR123	L1	6	PR129	L6	11	PS15	T2
	2	PR125	L2	7	PR130	L7	12	PK15	T3
	3	PR126	L3	8	AN179	L8	13	KS17	T4
	4	PR127	L4	9	AN837	L9	14	WD17	T5
	5	PR128	L5	10	PS13	T1
Crosses	1	PR123 × PS13	L1 × T1	16	PR127 × PS13	L4 × T1	31	PR130 × PS13	L7 × T1
	2	PR123 × PS15	L1 × T2	17	PR127 × PS15	L4 × T2	32	PR130 × PS15	L7 × T2
	3	PR123 × PK15	L1 × T3	18	PR127 × PK15	L4 × T3	33	PR130 × PK15	L7 × T3
	4	PR123 × K17	L1 × T4	19	PR127 × KS17	L4 × T4	34	PR130 × KS17	L7 × T4
	5	PR123 × WD17	L1 × T5	20	PR127 × WD17	L4 × T5	35	PR130 × WD17	L7 × T5
	6	PR125 × PS13	L2 × T1	21	PR128 × PS13	L5 × T1	36	AN179 × PS13	L8 × T1
	7	PR125 × PS15	L2 × T2	22	PR128 × PS15	L5 × T2	37	AN179 × PS15	L8 × T2
	8	PR125 × PK15	L2 × T3	23	PR128 × PK15	L5 × T3	38	AN179 × PK15	L8 × T3
	9	PR125 × KS17	L2 × T4	24	PR128 × KS17	L5 × T4	39	AN179 × KS17	L8 × T4
	10	PR125 × WD17	L2 × T5	25	PR128 × WD17	L5 × T5	40	AN179 × WD17	L8 × T5
	11	PR126 × PS13	L3 × T1	26	PR129 × PS13	L6 × T1	41	AN837 × PS13	L9 × T1
	12	PR126 × PS15	L3 × T2	27	PR129 × PS15	L6 × T2	42	AN837 × PS15	L9 × T2
	13	PR126 × PK15	L3 × T3	28	PR129 × PK15	L6 × T3	43	AN837 × PK15	L9 × T3
	14	PR126 × KS17	L3 × T4	29	PR129 × KS17	L6 × T4	44	AN837 × KS17	L9 × T4
	15	PR126 × WD17	L3 × T5	30	PR129 × WD17	L6 × T5	45	AN837 × WD17	L9 × T5

).

For the YR-stress experiment, fresh yellow rust inoculum was procured from the disease screening center at the Cereal Diseases Research Institute Murree (CDRI-Murree), NARC Islamabad, Pakistan. The inoculation mix of one part YR-spores reconstituted with 100 parts thickening agent (talcum powder) was sprayed using a turbo-air sprayer. Tween-20 surfactant was applied to break the surface tension of the spore mixture for to maintain spore germination and successful penetration of the developing fungus hyphae inside the host tissue. Following comprehensive screening during 2019–2020, the best-performing advanced lines, i.e., three each YR-resistant, moderately resistant, and susceptible advance lines were selected as the female parents; whereas, two lines each of highly resistant, two moderately resistant, and one YR-susceptible were selected as the male testers. Hybridization of selected breeding material, i.e. lines and testers was undertaken using line × tester mating format in 2020–2021.

Parents and their respective crosses were planted under randomized complete block design in triplicates under the optimal and YR-disease stress conditions. Fresh inoculation was used as a proxy for the development of high disease pressure because of the yellow-rust conducive environment at the early tillering phase, which also coincides with the early jointing stage of wheat with maximum humidity and low temperatures during February to March. The experimental plot area was 4 m² per entry, with four rows of 1.5 m length and 0.3 m row-row distance. All the experiments were replicated thrice, and space planted using wooden dibbler with 2 seeds per hill. Morocco spreaders were planted around the disease stress experiment for yellow-rust disease encouragement. The experiment was conducted using standard agronomic and cultural procedures. Nitrogen (120 kg ha⁻¹), Phosphorus (90 kg ha⁻¹), Potassium (60 kg ha⁻¹), Zinc (33 percent) 10 kg ha⁻¹, and Boron (17 percent) 7 kg ha⁻¹ were the fertilizers used. Except for nitrogen, which was applied in four splits, all fertilizers were given during the time of plantation, while with the first irrigation zinc and boron were applied. Post germination thinning was undertaken to ensure uniform plant population.

2.2. Collection of Data

The data were collected for different morphological (plant height, flag leaf area, and spike and peduncle lengths), physiological (canopy temperature at vegetative stage and normalized differential vegetative index), bio-chemical (chlorophyll A, B, total chlorophyll and carotenoids contents), yield (grain yield per plant) and its component traits (tillers per plant, spike length, peduncle length, grains per spike, and thousand grains weight). All the parameters were measured in each of the experimental blocks separately for parents and crosses, grown under optimal and YR-stress conditions.

2.3. Yellow Rust Co-Efficient of Infection

The coefficients of infection (C.I.) were calculated as previously described [20] by multiplying the disease infection percentage with the infection class as per modified Cobb scale [21].

2.4. Morphological, Physiological and Yield Component Parameters

Data related to the morphological traits, i.e., plant height (PHT), tillers per plant (TPP), peduncle length (PDL), spike length (SPL), grains per spike (GPS), thousand grain weight (TGW) and grain yield per plant (GYD) were recorded. The canopy temperature (CTV) was measured at the grain filling stage using a GM-320 digital infrared thermometer L.C.D 50–38 °C non-contact I.R Laser Gun Pyrometer Temperature Thermometer (Shenzhen Jumaoyuan Science and Technology Co. Ltd., Shenzhen, China). The flag leaf (FLA) was measured using a hand-held leaf area meter CI-203 (C.I.D Bioscience-Camas, WA USA). The normalized differential vegetative index (NDVI) was measured at the late vegetative crop and early anthesis stage using Green-Seeker handheld Crop Sensor device (Trimble Agriculture, SUNV, CA, USA).

2.5. Determination of Chlorophyll and Carotenoids Content

About 0.2–0.3 g of fresh leaf sample was sliced and kept in the test-tube comprising 10 mL of 80 percent acetone for around 24 h to determine chlorophyll concentration [22]. Optical densities (OD) of the supernatants were determined at 663 nm, 645 nm, 643 nm and 480 nm, for estimation of the chlorophyll A, chlorophyll B, total chlorophyll, and total carotenoid concentrations, respectively.

• Chlorophyll-A (CHA) = (12.7 × OD663) − (2.69 × OD645) × 1000 mL × Shoot fresh weight (g)
• Chlorophyll-B (CHB) = (22.9 × OD645) − (4.69 × OD663) × 1000 mL × Shoot fresh weight (g)
• Total-Chlorophyll (TCH) = (2.02 × OD643) + (8.02 × OD663) × 1000 mL × Shoot fresh weight (g)
• Carotenoid content (CAD) = OD480 × 4

OD_n is optical density at different absorbance levels, i.e., n = 663 nm/645 nm/643 nm/480 nm.

2.6. Statistical Analysis

2.6.1. Statistical Analysis of YR CI Data

The triplicate CI data for each parent and cross were analyzed for means and standard errors using IBM SPSS version 23.0. Tukey’s post hoc test under a one-way ANOVA was performed to analyze pair-wise comparisons among pairs of parents and crosses.

2.6.2. Estimation of General and Specific Combining Abilities

To quantify the genotype-environment interactions, a two-way analysis of variance (ANOVA) across the two test conditions (optimal and YR-Stress) for the above-mentioned parameters was undertaken using a mixed-effects model in MS Excel and Statistix ver. 8.1 packages.

To evaluate the significance of the conditions and genotype main effects, mean squares pertaining to genotype conditions interaction were utilized as an error term. Additionally, genotype conditions interactions were tested using error means squares. The significantly different trait interaction effects were further analyzed via line × tester analysis for assessment of general combining ability (GCA) and specific combining ability (SCA) effects under optimal and YR-Stress conditions (

Table 2. Line by tester analysis of 59 wheat genotypes (including 14 parental and 45 crosses) under each test conditions (optimal vs. YR-stress).

Source of Var.	Degree of Freedom	Mean Square	F. Calculated-Value
Replicates (r)	(r − 1) = 2		M1/M3
Lines (L)	L − 1 = 8	MSq1	σ2e + r(co-var. F-S − 2 × co-var. H-S) + (rT × co-var. H-S)
Testers (T)	T − 1 = 4	MSq2	σ2e + r(co-var. F-S − 2 × co-var. H-S) + (rL × co-var. H-S)
L × T	(L − 1)(T − 1) = 32	MSq3	σ2e + r(co-var. F-S − 2 × co-var. H-S)
Error	(r − 1)(L + T + LT) − 1 = 116	MSq4	σ2e
Total	(r)(L + T + LT) − L = 162

Key: co-var. F-S = co-variance full-sibs, co-var. H-S = co-variance of half-sibs.

).

The replicate-wise averages of each trait for the 45 cross genotypes were subjected to line × tester analysis, using the recommended model developed by Kempthorne in 1957 [16].

2.6.3. Variances Components Estimation

The variances of general combining ability (σ2gca) and specific combining ability (σ2sca) were calculated under both test conditions using previously described methods, based on the estimates of Co-var.(H-S) and Co-var.(F-S) [23,24]. The estimation of the effects of combining abilities, Standard errors for combining abilities and significance test for GCA and SCA effects were also undertaken as previously described.

2.6.4. Cluster Analysis for Classification of Genotypes

Both parents and crosses grown under optimal and YR-stress conditions were classified based on parametric data. The distribution of the genotypes into different classes was performed in IBM-SPSS ver. 23.0 using the Hierarchical cluster analysis with Agglomeration schedule [25,26]. A dendrogram was constructed using between-group linkage clustering method.

2.6.5. Among Group Variation of Hierarchical Classes

The hierarchical classes obtained from the above analyses were further evaluated for means and standard errors for each parameter. The magnitude of variance among different classes within parents and crosses was calculated using a one-way ANOVA in IBM-SPSS [25,27]. Pair-wise comparison of the classes was performed using post hoc Tukey’s test at significance level of 0.05 [28].

2.6.6. Factor and Correlation Analysis

Association among parameters and with genotypes under optimal and YR-stress conditions was tested using principal component analysis (PCA) in IBM-SPSS [25,29]. Biplots were constructed using varimax rotation separately for parents, crosses, optimal and YR-stress conditions [30]. Coefficient values less than 0.3 were suppressed for better readability of the plots. Correlation among the parameters was calculated using the factor analysis with Mineigen criteria at an iteration value of 25 [31]. To construct the correlation matrix, the KMO REPR AIC extraction rotation method was used.

3. Results

3.1. Yellow Rust Co-Efficient of Infection

Diverse disease severity as depicted by the co-efficient of infection (CI) was observed among parents and crosses (Figure 1). Since, under optimal conditions, no notable YR disease could be observed, data for YR CI are presented only for the YR-stress conditions. Among parents, the highest CI was noted for PR130 (33.0 ± 0.79) and AN179 (35.0 ± 0.71), whereas, the lowest CI was noted for parent PR125, PR127, PR128, KS17 and WD17 (2.7–5.3). Among crosses, the most severe infection was seen in cross AN179 × PK15 with the highest CI (56.3 ± 0.2), whereas, the lowest level infection with least CI range (1.2–3.8) was noted in crosses PR128 × PS15, P126 × PS15, PR126 × KS17, and PR126 × WD17. The results further revealed that both parents of the worst affected cross AN179 × PK15 were also highly susceptible to YR in independent screening, On the contrary, of the two parents of the cross AN179 × KS17 (5.5 ± 0.79), which had the lowest disease severity, one parent (AN179) had the highest disease severity; however, KS17 had the lowest disease coefficient of infection.

3.2. Diversity among Genotypes

Highly significant genetic differences (p < 0.05) were noted among the tested wheat genotypes for morphological, yield and component traits under optimal and YR-stress conditions (

Table 3. Analysis of variance for physiological, yield and yield components, bio-chemicals traits under optimal and YR-stress conditions.

MSE (Optimal)

SOV

DF

CTV

NDVI

PHT

FLA

TPP

SPL

PDL

GPS

TGW

GYD

CH A

CH B

TCH

CAD

Replicates

2

0.53 ns

0.031 ns

25.8 ns

25.3 ns

3.3 ns

1.6 ns

7.8 ns

57.2 ns

11.5 ns

612.6 ns

1.07 ns

0.94 ns

3.64 ns

0.04 ns

Lines (L)

8

1.50 **

0.036 **

379.8 **

26.6 **

3.8 **

4.5 **

111.3 **

219.0 **

126.7 **

41.3 **

196.7 **

56.6 **

454.5 **

7.8 **

Testers (T)

4

0.14 **

0.081 **

500.3 **

15.4 **

3.2 **

2.4 **

168.9 **

178.0 **

57.3 **

98.2 **

60.2 **

27.9 **

166.0 **

4.7 **

L × T

32

0.51 ns

0.009 **

40.4 **

3.8 **

2.5 **

0.3 **

21.9 **

55.6 **

55.2 **

50.3 **

30.7 **

10.4 **

68.9 **

1.3 **

Error

116

0.75

0.0001

9.0

26.6

1.2

0.12

2.2

18.47

10.78

21.50

0.7

0.2

0.8

0.01

MSE (YR-Stress)

SOV

DF

CTV

NDVI

PHT

FLA

TPP

SPL

PDL

GPS

TGW

GYD

CHA

CHB

TCH

CAD

Replicates

2

1.06 ns

0.022 ns

41.8 ns

38.7 ns

7.47

1.24 ns

9.1 ns

106.8 ns

48.7 ns

1252.6 ns

0.83 ns

0.17 ns

0.59 ns

0.03 ns

Lines (L)

8

12.4 **

0.050 **

179.6 **

211.1 **

1.88 **

3.66 **

90.1 **

140.2 **

87.3 **

84.3 **

134.0 **

40.2 **

314.9 **

6.4 **

Testers (T)

4

5.3 **

0.005 **

413.3 **

122.5 **

1.57 **

1.93 **

136.8 **

113.9 **

39.5 **

200.4 **

68.7 *

19.3 **

157.8 **

3.9 **

L × T

32

7.72 **

0.001 **

65.60 **

30.08 **

1.22 *

0.24 **

17.77 **

35.59 **

38.05 **

102.7 **

23.85 **

7.06 **

50.08 **

1.02 **

Error

116

1.36

0.022

6.97

7.77

1.10

0.15

1.76

10.51

13.92

45.42

0.52

0.18

0.93

0.03

Key: * = Significant at 5% probability level, ** = Significant at 1% probability level, ^ns = Non-significant, MSE = Mean squared error, SOV = Source of variance, DF = Degree of freedom, CTV = Canopy temperature at vegetative stage, NDVI = Normalized Differential Vegetative Index, PHT = Plant height, FLA = Flag leaf area, TPP = Tiller Plant⁻¹ PDL = Peduncle Length, SPL = Spike Length, GPS = Grains Spike⁻¹, TGW = Thousand Grain Weight, GYD = Grain Yield Plant⁻¹), CHA = Chlorophyll A, CHB = Chlorophyll B, TCH = Total Chlorophyll, CAD = Carotenoids.

). Consistent trends of higher genetic variances were seen among tested genotypes for all biochemical traits. Contrastingly, L × T interaction effects of the canopy temperature also demonstrated statistically non-significant (ns) variations under the YR stress.

3.3. General Combining Ability Effects (GCA)

Diversity among the parent material is a dynamic source of genotypic evaluations. To measure the genetic potential of parental lines to combine and give better performance, GCA effects of parental lines are estimated. In the present study, the GCA effects of parental lines showing that the lines PR125 (−3.53 **) and tester PS15 (−4.94 **) were the best general combiners for plant height under optimal conditions; while line PR123 (−2.19 **) and tester WD17 (−2.28 **) were good general combiners under YR-stress conditions (

Table 4. Summary table of good general combiners for different physiological, grain yield, yield components and biochemical traits under optimal and YR-stress conditions.

Condition	Type	CTV	PHT	FLA	TPP	SPL	PDL	GPS	TGW	GYD	CHA	CHB	TCH	CAD
Optimal	Lines	PR125	PR125							PR123
										PR126	PR126
	Testers	PS15	PS15				KS17				PS15	PS13		PS15
Overall	Lines		PR130	PR128	PR123	PR123	PR128	PR123	PR127		PR128	PR128	PR128	PR128
			AN837	PR129	AN179	PR126	PR129	PR126	AN179		AN179	AN179	AN179	AN179
				PR130				PR128	AN837		AN837	AN837	AN837	AN837
								PR130
	Testers		PS13	PS15		PK15		PS15	PS13		PS13		PS13	PS13
								KS17			KS17	KS17	KS17	KS17
					WD17		WD17			WD17		WD17	WD17	WD17
YR-stress	Lines	PR126	PR123							PR128
		PR130								AN179
		AN837
	Testers	PK15	WD17							KS17	WD17

Key: CTV = Canopy temperature at vegetative stage, NDVI = Normalized Differential Vegetative Index, PHT = Plant height, FLA = Flag leaf area, TPP = Tiller Plant⁻¹ PDL = Peduncle Length, SPL = Spike Length, GPS = Grains Spike⁻¹, TGW = Thousand Grain Weight, GYD = Grain Yield Plant⁻¹), CHA = Chlorophyll A, CHB = Chlorophyll B, TCH = Total Chlorophyll, CAD = Carotenoids.

and Table S1). Conversely, lines PR130 (−3.70 **, −2.24 **), and AN837 (−5.33 **, −3.27 **), and tester PS13 (−2.28 **, −4.39 **) were found to be the best general combiner for plant height under both optimal and YR-Stress conditions. Similarly, lines PR128 and AN179 demonstrated good general combining ability for yield component traits, i.e., tillers per plant and grain per spike; as well as for some biochemical traits such as chlorophyll-A, chlorophyll-B, total chlorophyll and carotenoids under both optimal and YR-stress conditions. For grain yield plant⁻¹, lines PR123 and PR126 under optimal conditions and tester KS17 (3.84 **) under YR-Stress demonstrated the best general combining ability estimates. However, PR128 and AN179 were found as overall good general combiners for grain yield plant⁻¹. For 1000 grains weight, under respective conditions (optimal, YR-stress), lines PR127 (3.11 **, 2.59 **), AN179 (4.81 **, 4.00 **) and AN837 (2.74 **, 2.28 **), and tester PS13 (2.48 **, 2.06 **) proved as good general combiners.

3.4. Specific Combining Ability Effects (SCA)

The specific combining ability (SCA) defines the ability of parents to combine in a specific combination. The SCA ability effects of 45 crosses for morphological, physiochemical and yield component traits under optimal and YR-stress conditions are given (

Table 5. Specific combining ability (SCA) estimates of crosses for physiological, grain yield, yield components and biochemical traits under optimal and YR-stress conditions.

Condition	CTV	NDVI	PHT	FLA	TPP	SPL	PDL	GPS	TGW	GYD	CH A	CH B	TCH	CAD
Optimal		L7 × T4	L5 × T2			L5 × T3			L2 × T5	L3 × T1	L3 × T1	L1 × T4	L2 × T3
		L9 × T2	L7 × T2						L5 × T5	L4 × T2	L8 × T4	L8 × T1	L5 × T3
			L9 × T1							L8 × T5	L8 × T5		L6 × T3
Overall	L2 × T4	L1 × T4	L2 × T3	L1 × T2	L2 × T4	L2 × T4	L1 × T4	L5 × T3	L3 × T4	L1 × T2	L1 × T3	L1 × T3	L1 × T3	L1 × T3
		L2 × T2		L5 × T3	L3 × T1	L3 × T5	L3 × T1	L6 × T4	L4 × T3	L1 × T3	L2 × T1	L2 × T1	L2 × T1	L2 × T1
		L2 × T4		L8 × T1		L8 × T1	L5 × T1	L8 × T1	L6 × T1	L2 × T4	L3 × T2	L2 × T3	L3 × T2	L3 × T2
		L3 × T2					L5 × T3	L8 × T4	L8 × T2	L5 × T3	L3 × T4	L3 × T2	L3 × T4	L3 × T4
		L3 × T3					L9 × T4		L9 × T1	L9 × T5	L4 × T3	L3 × T4	L4 × T2	L4 × T3
		L4 × T1					L9 × T5				L5 × T3	L4 × T3	L4 × T3	L5 × T4
		L5 × T2									L5 × T4	L5 × T3	L5 × T4	L6 × T5
		L5 × T4									L6 × T5	L5 × T4	L6 × T5	L9 × T2
		L6 × T4									L7 × T3	L6 × T4	L7 × T3	L9 × T4
		L6 × T5									L8 × T1	L6 × T5	L9 × T2
		L7 × T1									L9 × T2	L7 × T3	L9 × T4
		L8 × T1									L9 × T4	L9 × T2
		L8 × T3										L9 × T4
		L9 × T5
YR-stress	L5 × T1		L3 × T4	L4 × T3	L8 × T5	L9 × T2		L2 × T4	L2 × T3	L8 × T4		L7 × T1
	L5 × T3		L5 × T5	L9 × T5
	L6 × T2		L7 × T1
	L7 × T5		L9 × T2
	L8 × T3

Key: CTV = Canopy temperature at vegetative stage, NDVI = Normalized Differential Vegetative Index, PHT = Plant height, FLA = Flag leaf area, TPP = Tiller Plant⁻¹ PDL = Peduncle Length, SPL = Spike Length, GPS = Grains Spike⁻¹, TGW = Thousand Grain Weight, GYD = Grain Yield Plant-1), CHA = Chlorophyll A, CHB = Chlorophyll B, TCH = Total Chlorophyll, CAD = Carotenoids, L1 = PR123, L2 = PR125, L3 = PR126, L4 = PR127, L5 = PR128, L6 = PR129, L7 = PR130, L8 = AN179, L9 = AN837, T1 = PS13, T2 = PS15, T3 = PK15, T4 = KS17, T5 = WD17.

and Table S2). Few crosses demonstrated significant desirable SCA effects for traits under study under optimal and/or YR-stress conditions. Five crosses i.e., PR123 × PS15 (7.22 **, 4.96 **), PR125 × KS17 (7.54 **, 4.52 **), PR128 × PK15 (6.74 **, 5.09 **), AN179 × WD17 (7.48 **, 5.62 **) and AN837 × WD17 (10.05 **, 6.54 **) showed positive SCA effects for grain yield plant⁻¹ under all conditions. Two crosses i.e., PR126 × PS13 (8.58 **) and PR127 × PS15 (4.85 **) demonstrated positive SCA for grain yield plant⁻¹ only under optimal conditions. In contrast, under YR-stress, three crosses, i.e., PR128 × KS17 (7.32 **), PR129 × PS13 (5.25 **), and AN179 × KS17 (9.46 **) showed significant positive SCA effects for grain yield per plant. For canopy temperature, cross PR125 × KS17 was best specific combiner under optimal, while crosses AN179 × PK15, PR130 × WD17, PR127 × KS17, PR128 × PS13 and PR129 × PS15 had the best SCA under YR-stress. About 45% of crosses showed positive SCA effects for NDVI under optimal and YR-stress. Cross AN179 × PS13 was the best specific combiner for physiological and the most yield associated under optimal and YR stress. Cross AN837 × WD17 was found to be good specific combiner for flag leaf area, NDVI, tillers per plant, peduncle length and grain yield plant⁻¹ under optimal and YR-stress. In addition, four crosses (PR125 × PS15, PR125 × KS17, PR126 × WD17 and AN179 × PS13) were found to be the best specific combiners for spike length, AN837 × KS17 for chlorophyll-A, chlorophyll-B and total chlorophyll under optimal and YR-stress. However, cross PR123 × PK15 showed best SCA effects for carotenoids under both conditions.

3.5. Variances and Broad Sense Heritability Estimation

The estimates of variation due to general combining ability (σ²gca), specific combining ability (σ²sca) and their ratio (σ²gca/σ²sca) revealed that for most of studied traits σ²sca was greater than σ²gca, indicating non-additive gene action for most of studied traits under both test conditions (

Table 6. Estimation of related variances (σ²) and heritability for physiological, biochemical, yield and component traits under optimal and YR disease stress conditions.

	Traits	σ2gca (L)	σ2gca (T)	σ2gca (Av.)	σ2sca (C)	σ2A	σ2D	σ2gca/σ2sca	(σ2sca/σ2gca)1/2	H2BS
Variances estimation (optimal)	CTV (°C)	0.07	0.02	0.03	0.08	0.03	0.08	0.19	0.44	0.30
	NDVI %	0.018	0.027	0.02	0.03	0.05	0.029	0.83	0.91	0.90
	FLA (cm2)	14.58	4.13	7.86	8.91	15.72	8.91	0.88	0.94	0.76
	TPP	0.09	0.03	0.05	0.43	0.10	0.43	0.11	0.34	0.40
	SPL (cm)	0.28	0.08	0.15	0.06	0.30	0.06	2.63	1.62	0.86
	PDL (cm)	5.96	5.44	5.63	6.59	11.25	6.59	0.85	0.92	0.90
	GPS	10.90	4.53	6.81	12.38	13.61	12.38	0.55	0.74	0.68
	TGW(g)	4.76	0.08	1.75	14.82	3.50	14.82	0.12	0.34	0.69
	GYP (g)	1.23	3.62	1.89	19.09	3.78	19.09	0.10	0.31	0.78
	PHT(cm)	22.63	17.03	19.03	10.48	38.06	10.48	1.82	1.35	0.91
	CHA	11.07	1.09	4.65	10.01	9.31	10.01	0.47	0.68	0.90
	CHB	3.08	0.65	1.52	3.38	3.04	3.38	0.45	0.67	0.95
	TCH	25.70	3.60	11.49	22.7	22.99	22.7	0.51	0.71	0.96
	CAD	0.44	0.128	0.24	0.41	0.48	0.41	0.58	0.76	0.90
Variances estimation (YR-Stress)	CTV (°C)	0.31	0.09	0.19	2.12	0.11	2.12	0.03	0.16	0.62
	NDVI %	0.011	0.017	0.029	0.02	0.03	0.017	0.85	0.92	0.95
	FLA (cm2)	12.07	3.42	0.88	7.43	13.02	7.43	0.88	0.94	0.77
	TPP	0.04	0.013	0.024	0.23	0.048	0.23	0.10	0.32	0.48
	SPL (cm)	0.23	0.06	0.12	0.05	0.24	0.05	2.54	1.59	0.89
	PDL (cm)	4.83	4.41	4.56	5.34	9.12	5.34	0.85	0.92	0.92
	GPS	6.97	2.90	4.36	8.36	8.71	8.36	0.52	0.72	0.71
	TGW(g)	3.28	0.05	1.21	8.04	2.41	8.04	0.15	0.39	0.51
	GYP (g)	−0.60	1.77	0.93	11.05	1.85	11.05	0.08	0.29	0.65
	PHT(cm)	7.60	12.88	10.99	19.54	21.99	19.54	0.56	0.75	0.91
	CHA	7.34	1.66	3.69	7.78	7.38	7.78	0.47	0.69	0.97
	CHB	2.21	0.45	1.08	2.29	2.16	2.29	0.48	0.70	0.95
	TCH	17.66	3.99	8.87	16.38	17.74	16.38	0.54	0.74	0.98
	CAD	0.36	0.11	0.20	0.33	0.39	0.33	0.59	0.77	0.97

Key: L = Lines, T = Testers, AV = Average, C = Crosses, A = additives, D = dominance, CTV = Canopy temperature at vegetative stage, NDVI = Normalized Differential Vegetative Index, PHT = Plant height, FLA = Flag leaf area, TPP = Tiller Plant-1 PDL = Peduncle Length, SPL = Spike Length, GPS = Grains Spike⁻¹, TGW = Thousand Grain Weight, GYD = Grain Yield Plant⁻¹), CHA = Chlorophyll A, CHB = Chlorophyll B, TCH = Total Chlorophyll, CAD = Carotenoids, h² = heritability, BS = Broad sense.

). The higher broad-sense heritability was observed for total chlorophyll (0.96), chlorophyll B (0.95), followed by NDVI, plant height (each 0.91), peduncle length, chlorophyll A, and carotenoids (each 0.90) under both optimal and YR stress conditions. Greater heritability estimates indicate comparatively greater selection effectiveness for these traits. For grain yield per plant, moderate heritability was observed in magnitude under optimal (0.78), while low heritability was observed under YR-stress (0.65) conditions. Under optimal conditions, for flag leaf area (0.76), spike length (0.86), thousand grain weight (0.69), grains per spike (0.68) moderate heritability were observed and for tillers per plant (0.40) and canopy temperature (0.30) low heritability were observed. Thousand grain weight (0.69 vs. 0.51) and grain yield per plant (0.78 vs. 0.65) depicted disease-stress effects, with moderate h² under optimal and low h² under stress conditions.

3.6. Classification of Parents and Crosses

Cluster analysis classified parents and crosses into different classes based on their mean performance in all traits measured (Figure 2). A separate analysis was performed for the specimen grown under optimal and YR-stress conditions. In parent genotypes two distinct classes formed under optimal condition, while under YR-stress the parents were categorized into three classes. Under optimal condition two parent lines (PR128 and PR125), and three parent testers (KS17, WD17, and PS15) grouped in one class, while the rest made a different class. However, under YR-stress, one parent line (PR128), and the three testers (KS17, WD17, and PS15), made a third separate cluster on dendrogram.

Similarly, under optimal condition, crosses were classified into three distinct classes; however, under YR-stress the crosses made four different clusters (Figure 2). Under optimal conditions, 12 crosses were grouped under class-2, one under class-3 and the remaining 32 crosses were grouped under class-1. Under YR-stress one cross distinctly clustered under class-4, eight crosses were grouped in class-3, five under class-2 and remaining 36 crosses were grouped into single large cluster (class-1). One cross AN179 × KS17 distinctly clustered under both optimal and YR-stress condition.

3.7. YR-Stress Significantly Affects Wheat Parameters under Study

The genotype classes identified through cluster analyses were further evaluated for statistical differences between classes (

Table 7. Mean values of different parameters in (a) parents and (b) crosses based on distance classes in disease free and YR stress conditions.

(a) Parents
Traits	Optimal				YR-Stress						F-Value
	Class-1		Class-2		Class-1		Class-2		Class-3
CTV	18.6 ± 0.17 a		18.7 ± 0.17 ab		20.6 ± 0.59 b		20.4 ± 0.77 ab		20.9 ± 1.08b		4.6 **
NDVI	0.7 ± 0.02 bc		0.8 ± 0.02 c		0.5 ± 0.02 a		0.6 ± 0.02 a		0.6 ± 0.02 ab		14.2 ***
PLH	87.5 ± 4.41 ab		104.3 ± 3.01 b		84.3 ± 5.19 ab		75.4 ± 6.27 a		93.6 ± 4.65 ab		3.6 *
FLA	37.1 ± 2.60		34.9 ± 3.01		36.7 ± 2.45		27.1 ± 1.71		33.6 ± 3.27		1.4 ns
PDL	31.2 ± 1.58 ab		38.0 ± 2.09 b		30.8 ± 0.46 ab		24.6 ± 3.45 a		35.8 ± 2.22 b		5.8 **
TPP	6.9 ± 0.50 b		6.9 ± 0.50 b		6.0 ± 0.34 ab		4.6 ± 0.64 a		5.4 ± 0.51 ab		3.0 *
SPL	10.9 ± 0.46 bc		11.5 ± 0.47 c		9.4 ± 0.21 b		7.4 ± 0.31 a		9.8 ± 0.20 bc		10.1 ***
GPS	51.3 ± 2.73 bc		57.8 ± 5.15 c		43.2 ± 1.55 ab		31.1 ± 0.87 a		49.3 ± 2.38 bc		7.4 **
TGW	45.3 ± 1.84 b		42.0 ± 3.66 bc		35.5 ± 0.80 ab		45.4 ± 1.06 b		31.9 ± 1.09 a		7.8 ***
GYP	30.2 ± 2.79 b		34.6 ± 3.97 b		25.9 ± 1.10 ab		15.6 ± 3.72 a		28.0 ± 3.31 ab		3.8 *
CHA	6.53 ± 0.78 a		15.7 ± 0.99 b		4.7 ± 0.62 a		8.2 ± 2.73 a		13.6 ± 1.44 b		19.0 ***
CHB	3.8 ± 0.52 a		8.6 ± 1.10 c		2.6 ± 0.40 a		4.8 ± 1.05 ab		7.4 ± 0.93 bc		11.9 ***
TCH	10.1 ± 1.14 a		24.3 ± 1.76 b		7.2 ± 0.81 a		13.0 ± 3.27 a		21.0 ± 2.26 b		22.0 ***
CAD	1.6 ± 0.31		2.5 ± 0.38		1.4 ± 0.35		2.1 ± 0.58		2.2 ± 0.44		1.4 ns
(b) Crosses
Traits	Optimal				YR-Stress						F-value
	Class-1	Class-2		Class-3	Class-1	Class-2		Class-3		Class-4
CTV	18.4 ± 0.09 a	18.5 ± 0.08 a		18.8 ± 0.58a	20.9 ± 0.27 b	20.1 ± 0.57 ab		21.5 ± 0.82 b		20.3 ± 1.02 ab	15.6 ***
NDVI	0.6 ± 0.01 c	0.6 ± 0.02 c		0.5 ± 0.01b	0.5 ± 0.01 b	0.5 ± 0.01 b		0.5 ± 0.02 b		0.3 ± 0.01 a	25.3 ***
PLH	102.0 ± 1.16 b	101.8 ± 2.01 b		87.6 ± 2.14a	91.2 ± 0.98 a	93.0 ± 1.71 ab		92.3 ± 3.53 ab		84.5 ± 3.76 a	12.7 ***
FLA	38.8 ± 0.99	35.8 ± 1.26		35.8 ± 1.63	35.0 ± 0.86	33.3 ± 3.32		33.1 ± 1.29		32.6 ± 1.48	2.8 ns
TPP	7.3 ± 0.16 d	7.0 ± 0.30 cd		5.8 ± 0.31 bc	5.1 ± 0.11 ab	4.5 ± 0.13 ab		5.2 ± 0.31 ab		4.0 ± 0.21 a	31.4 ***
SPL	12.1 ± 0.10 d	11.4 ± 0.16 cd		11.5 ± 0.22 cd	10.8 ± 0.09 abc	10.9 ± 0.16 bc		10.0 ± 0.18 a		10.3 ± 0.19 ab	25.7 ***
PDL	36.7 ± 0.69 b	35.4 ± 1.04 b		25.1 ± 0.47 a	32.7 ± 0.54 b	33.3 ± 3.12 b		32.2 ± 0.83 b		22.6 ± 0.42 a	15.0 ***
GPS	52.2 ± 0.92 e	51.4 ± 1.34 de		33.5 ± 2.66 ab	41.6 ± 0.69 c	44.8 ± 1.91 cd		39.4 ± 1.24 bc		26.8 ± 2.13 a	37.6 ***
TGW	46.0 ± 0.79 bc	46.2 ± 1.12 bc		61.1 ± 6.40 d	38.2 ± 0.66 a	36.3 ± 1.06 a		39.7 ± 1.14 ab		50.7 ± 5.31 c	20.9 ***
CHA	5.8 ± 0.29 a	13.3 ± 1.33 b		15.3 ± 0.08 b	4.5 ± 0.27 a	8.2 ± 0.51 a		13.5 ± 1.00 b		13.2 ± 0.17 b	43.3 ***
CHB	2.5 ± 0.22 ab	6.7 ± 0.71 d		6.3 ± 0.01 cd	1.8 ± 0.15 a	4.2 ± 0.50 bc		6.7 ± 0.56 d		4.8 ± 0.23 cd	32.3 ***
TCH	8.3 ± 0.47 ab	20.0 ± 1.95 c		21.7 ± 0.08 c	6.4 ± 0.38 a	12.3 ± 0.61 b		20.2 ± 1.38 c		18.1 ± 0.37 c	45.9 ***
CAD	0.9 ± 0.07 ab	2.6 ± 0.22 c		2.9 ± 0.04 c	0.8 ± 0.06 a	1.5 ± 0.31 b		2.7 ± 0.11 c		2.6 ± 0.06 c	46.8 ***
GYP	42.9 ± 0.95 d	41.9 ± 1.83 d		28.3 ± 2.03 c	29.0 ± 0.76 c	28.1 ± 1.72 ab		36.2 ± 2.55 cd		19.8 ± 1.42 a	31.0 ***

Key: * = p < 0.05, ** = p < 0.01, *** = p < 0.001, ns = non-significant, CTV = Canopy temperature at vegetative stage, NDVI = Normalized Differential Vegetative Index, PHT = Plant height, FLA = Flag leaf area, TPP = Tiller Plant-1 PDL = Peduncle Length, SPL = Spike Length, GPS = Grains Spike-1, TGW = Thousand Grain Weight, GYD = Grain Yield Plant-1), CHA = Chlorophyll A, CHB = Chlorophyll B, TCH = Total Chlorophyll, CAD = Carotenoids, superscripts on the values show pairwise comparisons in a row.

). The analysis revealed significant differences among classes grouped under optimal and YR-stress. The classes of parental lines under optimal conditions demonstrated significantly higher performance compared to the corresponding classes under YR-stress. However, parental class-3 (under YR-stress) showed comparable performance with the parental classes under optimal conditions. The YR-stress class-2 had significantly lower values for NDVI (p < 0.05), tillers per plant (p < 0.05), spike length (p < 0.001), grains per spike (p < 0.01), and grain yield per plant (p < 0.05) as compared to both the optimal classes. Additionally, NDVI for YR-stress class-1 was also found to be significantly lower than classes under optimal conditions. The thousand grains weight was lower (p < 0.05) in YR-stress class-3 compared to the optimal classes.

Under optimal conditions, both parental classes performed similar except for chlorophyll-A, B and total chlorophyll, which were significantly higher (p < 0.001) in class-2 genotypes. Under YR-stress, classes 1 and 2 showed similar performance, except for the spike length, which was significantly higher in class-1 (p < 0.05). Under YR-stress, class-3 had significantly higher chlorophyll-A (p < 0.001; class-1), B (p < 0.01; class-1) and total chlorophyll (p < 001; class-1) contents as compared to class-1. Moreover, under YR-stress parental class-3 as compared to class-2 showed significantly higher (p < 0.05) performance for spike length and grains per spike. However, the thousand grains weight was significantly higher (p < 0.05) in class-2 as compared to class-3 under YR-stress.

In contrast to the parents, three and four classes were found among crosses under optimal and YR-stress, respectively. Overall consistency was noted for different parameters across optimal and YR-stress conditions. Minor differences were observed among classes under optimal and YR-stress conditions. However, the reduction in NDVI, tillers per plant, spike length, and thousand grains weight were noted for classes under the YR-stress as compared to classes under optimal conditions. In an exception, significant increase was noted for canopy temperature among YR-stress classes 1 and 3 as compared to the optimal classes.

As mentioned above (Section 3.5. Classification of parents and crosses), cross AN179 × KS17 was distinctly classified under class-3 under optimal, and in class-4 under YR-stress conditions. A means analysis showed that for this specific cross the values for NDVI, tillers per plant, spike length, thousand grains weight, and grain yield per plant reduced significantly (p < 0.05) under YR-stress. Other parameters for cross AN179 × KS17 remained similar among the two conditions.

3.8. Association Study through Principal Component Analysis

The PCA revealed an association among different parameters and with genotypes. For parents under both optimal and YR-stress conditions, different parameters formed unique clusters showing mutual associations. Chlorophyll-A, B, and total chlorophyll contents formed single cluster under optimal as well YR-stress conditions revealing narrow angles between their leading lines (Figure 3). Similarly, other parameters also formed independent PCA clusters, revealing association under both optimal and YR-stress conditions, including among NDVI and carotenoid; tiller per plant and flag leaf area; and spike length, grains per spike, plant height, and the peduncle length.

Other parameters showed divergence under YR-stress in contrast to optimal conditions. For example, canopy temperature was re-clustered with spike length, grains per spike, plant height, and the peduncle length as opposed to its positioning with tiller per plant and flag leaf area under optimal condition. Under YR-stress, grain yield per plant demonstrated weakening of its association with peduncle length, and plant height in comparison to its stronger association with these parameters under optimal conditions. The thousand grains weight had a negative association with yield and component traits under both optimal and YR-stress, demonstrated by very wide angle in between.

The clustering of parents and crosses resulting from PCA (Figure 3) was as predicted through the dendrogram (Figure 2) described in Section 3.5. Among the parents, PR126 demonstrated stronger association with grain yields under optimal condition. Parents PR128, KS17, and WD17 under optimal conditions had strong positive association with grains per spike, while PR125 showed association with tillers per plant as depicted by their co-clustering. Parents PR123, PR130, and PS13 were positioned together with chlorophyll-A, B, total chlorophyll contents on PCA plot. Parent AN179 clustered away from all the parameters under both optimal and YR-stress conditions, indicative of poor performance. However, YR-stress resulted in weakened association of KS17 with grains per spike. Similarly, grain yield per plant improved in PR125 and AN837 under YR-stress.

PCA of the crosses showed genotypes association with different plant parameters under YR-stress. Cross combinations AN179 × PS15, PR127 × PS15, and PR128 × PS13 showed strong positive association with grain yield per plant under YR-stress compared to optimal conditions. The cross PR125 × PS13 outperformed AN179 × WD17 and AN837 × (PS13, PS15, and KS17) for grains per spike under YR-stress. AN179 × (PS13, KS17 and WD17) and AN837 × (PS15 and WD17) demonstrated a shift in clustering and association with parameters under YR-stress.

3.9. Correlation among Different Parameters

Under optimal conditions, among parents, strong positive correlation of plant height was noted with spike length, peduncle length, and grains per spike (Figure 4). Moreover, spike length was positively correlated with peduncle length and grains per spike. However, total grains weight had negative correlation with spike length and grains per spike. Under YR-stress, a shift in correlation among parameters was observed for grain yield per plant, which showed a strong positive correlation with tillers per plant in contrast to peduncle length under optimal condition. Correlation under optimal conditions of plant height with other three parameters weakened under YR-stress. Similarly, correlation between spike length and peduncle length also weakened under YR-stress. Inter-parameter correlation among crosses remained consistent under optimal and YR-stress, except grain yield per plant, which only had moderate correlation with tillers per plant under YR-stress as opposed to a strong correlation under optimal condition. Thousand grains weight showed strong negative correlation with grains per spike both under optimal and YR-stress. Correlation among other parameters did not show drastic change between optimal and YR-stress conditions.

4. Discussion

The different YR severity pattern noted in the current study among the crosses compared to their parents may possibly be due to the rearrangement of YR resistance loci in the recombinant lines as reported previously [32]. This is also supported by observations for the cross AN179 × KS17, which showed increased resistance as opposed to one of its highly susceptible parents, i.e., AN179. AN179 × KS17 contains YR resistance associated novel quantitative trait loci (QTL) [33].

Combining ability tests are frequently adapted by wheat experts to identify the best parents in various cross combinations [34]. Lines (PR128 and AN179) and testers (WD17 and KS17) showed favorable GCA effects under YR stress with additive gene effects for grain yield and other traits are recommended for targeted YR resistance breeding. Moreover, it is expected that inter-mating of selected parents should reveal hidden genetic variability in F₂ and F₃ progeny via disruption of the unfavorable linkages [34]. The differential GCA response of wheat genotypes under optimal and YR-stress conditions may be due to changeable multi-gene control of grain yield. The presence of additive and dominant gene action for grain yield and its associated traits in wheat under YR-stress has been reported previously [35].

In self-pollinated crops like wheat GCA estimates alone are not effective for desirable line selection method because of similarity in pure lines and low genetic diversity [36]. However, GCA in combination with SCA estimates may aid in predicting desirable cross combinations. It is interesting to note from the current study that at least one parent (PR128 and WD17, respectively) of each of the specific crosses (PR128 × PS13 and AN179 × WD17 with good SCA estimates) had the best GCA effects. Prediction of the prolific cross combinations via SCA is required to produce transgressive segregants for use as effective pure lines [37]. Target trait improvement may be aided by selecting better cross-combinations, with good SCA estimates, and with high GCA for at least one of the parents [34,38,39].

The observation of higher SCA estimates compared to the GCA is potentially indicative of non-additive gene action for most of the studied traits [35]. Good combiners identified through GCA and SCA estimates clustered closely on dendrogram based on their mean performance. YR impacted most of the plant parameters, which was confirmed by the resultant shift of classes as revealed by the cluster analysis. These findings indicate differential genotype response to YR stress. Other studies have also reported negative effects of YR-stress on different yield and component traits in 44 Pakistani wheat genotypes [40], as well as in 16 Spanish wheat lines [9]. The sole cross AN179 × KS17 (having YR-resistance QTLs identified in the previous study [33]) retained its distinct class under both optimal and YR-stress conditions in the current study.

PCA scatter plot backed up outcomes of the GCA, SCA and cluster analysis and identified YR-resistant genotypes with consistent performance under YR-stress. The best general and specific combiners that clustered in close relationship also showed their association with yield and component traits. The known resistant parent KS17 [33] performed well under YR-stress, whereas parent AN179, identified as a susceptible line in the above-mentioned analyses, was found to be the most affected by YR-stress through PCA. Among crosses, PR128 × PS13, which was identified as good specific combiners was the most productive under YR-stress too. Outcomes from the current study also confirm the power of PCA as a useful tool to identify candidate genotypes for wheat breeding programs.

The direct selection of genotypes based on grain yield is not a viable option since grain yield is a complex, polygenic trait influenced by genetics and the environment [41]. Therefore, the development of superior genotypes should be informed by the correlation between yield and associated parameters [42]. If plant’s physiology is stressed, it will compensate by enhancing another trait. The current study revealed that under YR-stress grain yield correlated strongly with tillers per plant, rather than peduncle length, thus reduced photosynthetic activity under YR-stress is compensated for by an increase in productive tillers [43]. Previous studies have demonstrated that wheat grain yield can be increased by improving yield component traits, such as grains per spike, tillers per plant, and peduncle length [44]. In the current study and before [45], these parameters were mutually correlated. Previous research [33,46] suggests that YR-stress weakens the connection between plant height, spike length, peduncle length, and grains per spike. Consistent with prior studies [47], thousand grains weight correlated negatively with grains per spike. The trade-off between morphological and yield components may stabilize grain yield under YR-stress.

5. Conclusions

This study identified genotypes that performed consistently under YR-stress. Combining ability estimates will increase the diversity of YR-resistance genotypes to counter evolving P. striiformis pathotypes. The peduncle length and tillers per plant can be used as phenotypic markers for developing high-yielding YR-resistant wheat cultivars. This study will lead to improved wheat agronomic performance, and sustainable production with increased YR resistance. Good general combiners should be incorporated in a YR-resistance breeding program, whereas consistent specific combiners are recommended for multi-location trials after preliminary testing of transgressive segregants.