Research on the Effects of the Relationship between Agronomic Traits and Dwarfing Genes on Yield in Colored Wheat

This research focuses on 72 approved varieties of colored wheat from different provinces in China. Utilizing coefficients of variation, structural equation models, and correlation analyses, six agronomic traits of colored wheat were comprehensively evaluated, followed by further research on different dwarfing genes in colored wheat. Using the entropy method revealed that among the 72 colored wheat varieties, 10 were suitable for cultivation. Variety 70 was the top-performing variety, with a comprehensive index of 87.15%. In the final established structural equation model, each agronomic trait exhibited a positive direct effect on yield. Notably, plant height, spike length, and flag leaf width had significant impacts on yield, with path coefficients of 0.55, 0.40, and 0.27. Transcriptome analysis and real-time fluorescence quantitative polymerase chain reaction (RT-qPCR) validation were used to identify three dwarfing genes controlling plant height: Rht1, Rht-D1, and Rht8. Subsequent RT-qPCR validation clustering heatmap results indicated that Rht-D1 gene expression increased with the growth of per-acre yield. Rht8 belongs to the semi-dwarf gene category and has a significant positive effect on grain yield. However, the impact of Rht1, as a dwarfing gene, on agronomic traits varies. These research findings provide crucial references for the breeding of new varieties.


Introduction
Colored wheat is a valuable genetic resource with distinctive grain colors such as black, blue, purple, and green.Due to the abundance of natural pigments in its grains, colored wheat has drawn attention for its unique quality characteristics and development potential [1,2].Colored wheat can play a significant role in the prevention of various diseases associated with oxidative stress [3][4][5].However, colored wheat germplasm resources, especially those with both optimal quality and yield-related traits and availability for large-scale planting, are still lacking [6].Thus, breeding colored wheat with both superior quality and superior yield features has become a popular objective to increase the value of processed wheat products.This study selected suitable colored wheat varieties for local conditions, compared different agronomic traits and yields, and ultimately identified dwarfing genes that increase yield.These genes were then validated, providing theoretical support for the breeding of high-yield colored wheat.This research aimed to provide valuable insights for the development of new varieties of colored wheat resources.Consequently, this study provides theoretical support for breeding high-yielding colored wheat and aims to offer valuable insights and guidance for the development of new varieties within the colored wheat resources.This endeavor is expected to inject new momentum and direction into the future development of the colored wheat breeding field.
Different yield components can lead to variations in yield under different environmental conditions.The spike number, grain number per spike, and thousand-grain weight determine the individual plant yield of colored wheat and are crucial yield components.Plant height and spike length are related to the lodging resistance of colored wheat and influence its overall yield.Therefore, to breed high-yielding, stable, and high-quality colored wheat varieties, it is essential to investigate these yield components and conduct in-depth analyses.Currently, there are numerous reports on the contributions of the three yield components to grain yield.However, due to variations in the use of varieties, sample sizes, and experimental conditions, the conclusions of previous research are inconsistent.Under specific conditions, the thousand-grain weight [7][8][9], grain number per spike [10], or spike number per unit area may all become the major contributors to yield.Furthermore, for some varieties, plant height is the primary determining factor for yield [11].The deployment of Reduced height (Rht) semi-dwarfing genes (Rht-B1 and Rht-D1) was a crucial component of the 'Green Revolution' for improving yield potential due to an increased harvest index and lodging resistance.Due to these factors, more than 70% of wheat cultivars grown globally nowadays incorporate at least one of these semi-dwarfing genes [12].
The genes controlling plant height include the Rht series of genes, each of which has a different impact on wheat plant height.Among them, the most widely used dwarf genes in wheat dwarfing breeding are Rht-B1b and Rht-D1b from the Agronomy 10 wheat cultivar, as well as Rht 8 from Japanese red wheat (AKAK-OMUGi) [13].To further clarify the relationship between colored wheat traits and yield, this study utilized structural equation modeling, which is sometimes referred to as covariance structure analysis, in addition to covariance structure modeling and the analysis of covariance structures [14], as a multivariate analysis method.Initially developed for use in the social sciences, structural equation modeling has also been applied to address complex issues in ecology [15].
Although more than 20 dwarfing genes have been documented in wheat, only a few genes are presently utilized in wheat breeding.More loci or genes that regulate plant height without influencing yield components need to be investigated further.In this study, we identified the Rht1, Rht-D1, and Rht8 genes that influence plant height.This finding holds significant implications for enhancing wheat yield.This study employed statistical methods to screen 72 varieties of colored wheat in China, selecting those suitable for local conditions.The selected varieties were further analyzed to determine their agronomic traits, and three dwarfing genes controlling plant height, namely Rht1, Rht-D1, and Rht8, were validated.This research aimed to provide valuable insights for the development of new varieties of colored wheat resources.

Experimental Materials
The experiment was conducted at the Yangchangzi Experimental Base (46 • 05 ′ N, 227 • 03 ′ E) of the Hinggan League Agricultural and Animal Husbandry Technology Extension Center during 2022-2023 (Figure 1).A total of 72 germplasm resources were utilized in this study (Table A1).These materials are all genetically stable varieties bred by various breeding units.The experimental plots were designed, managed in the field, and characterized based on the new standards of the "Technical Regulations for Regional Trials of Crop Varieties (Wheat)" issued by the Ministry of Agriculture.A total of six traits were investigated, including plant height, spike length, flag leaf length, flag leaf width, stem thickness, and yield.
by various breeding units.The experimental plots were designed, managed in the field, and characterized based on the new standards of the "Technical Regulations for Regional Trials of Crop Varieties (Wheat)" issued by the Ministry of Agriculture.A total of six traits were investigated, including plant height, spike length, flag leaf length, flag leaf width, stem thickness, and yield.

Data Analysis
Data statistical analysis was performed using SPSS 20.0 software, and the "R package" was utilized for structural equation modeling.The entropy method was employed to screen out varieties.For index measurement, in each repetition, 10 plants were randomly harvested, and the parameters of plant height, ear length, flag leaf length, flag leaf width, stem thickness, and yield were statistically analyzed.

Initial Structural Equation Model
A structural equation model is an a priori model [15], requiring the establishment of hypothesized paths based on practical experience and theory.Subsequently, through structural equation modeling, the direct and indirect relationships among various factors are analyzed, validated, and optimized.Ultimately, the rational causal relationships among the factors are determined [16].The initial model for this study is illustrated in

Data Analysis
Data statistical analysis was performed using SPSS 20.0 software, and the "R package" was utilized for structural equation modeling.The entropy method was employed to screen out varieties.For index measurement, in each repetition, 10 plants were randomly harvested, and the parameters of plant height, ear length, flag leaf length, flag leaf width, stem thickness, and yield were statistically analyzed.

Initial Structural Equation Model
A structural equation model is an a priori model [15], requiring the establishment of hypothesized paths based on practical experience and theory.Subsequently, through structural equation modeling, the direct and indirect relationships among various factors are analyzed, validated, and optimized.Ultimately, the rational causal relationships among the factors are determined [16].The initial model for this study is illustrated in Figure 2. Based on the literature [16] and years of breeding experience, the initial model assumes that plant height, flag leaf length, flag leaf width, ear length, and stem thickness have direct effects on yield.Therefore, it was hypothesized that greater plant height leads to more tillers, resulting in a greater number of ears.After setting up the initial model, the estimation of 16 relationships between parameters was required (Figure 2), which was fewer than the total number of equations that could be established [6 (6 + 1)/2 = 21].Thus, the model was identifiable.
assumes that plant height, flag leaf length, flag leaf width, ear length, and stem thickness have direct effects on yield.Therefore, it was hypothesized that greater plant height leads to more tillers, resulting in a greater number of ears.After setting up the initial model, the estimation of 16 relationships between parameters was required (Figure 2), which was fewer than the total number of equations that could be established [6 (6 + 1)/2 = 21].Thus, the model was identifiable.

Model Verification
The overall evaluation of the model was conducted using the χ 2 test [16], where a smaller χ 2 value indicates a better fit of the model.If p > 0.05, it indicates that the model does not omit important parameters.In this study, there were a total of 432 (72 × 6) data points, meeting the minimum recommended sample size for the χ 2 test in structural equation modeling, which is at least five times the number of model parameters [17].The fit of the model was further assessed using the root mean square error of approximation (RMSEA) and the Akaike information criterion (AIC) [18].For RMSEA, a value less than 0.05 indicates an excellent fit; 0.05 ≤ RMSEA < 0.08 suggests a good fit; 0.08 ≤ RMSEA < 0.10 indicates an acceptable fit; and RMSEA ≥ 0.10 suggests a poor fit.For the AIC, a lower value indicates a better fit of the model.

Entropy Method and Indicator Selection
Entropy method: The entropy method is a technique for objectively assigning weights to indicators based on the magnitude of their information entropy.A smaller value of information entropy indicates a greater degree of dispersion for the indicator, implying more information content and therefore a higher weight.To overcome the limitations of the entropy method in dealing only with cross-sectional data and not being able to compare different varieties, this study adopted the panel entropy method [19].

Model Verification
The overall evaluation of the model was conducted using the χ 2 test [16], where a smaller χ 2 value indicates a better fit of the model.If p > 0.05, it indicates that the model does not omit important parameters.In this study, there were a total of 432 (72 × 6) data points, meeting the minimum recommended sample size for the χ 2 test in structural equation modeling, which is at least five times the number of model parameters [17].The fit of the model was further assessed using the root mean square error of approximation (RMSEA) and the Akaike information criterion (AIC) [18].For RMSEA, a value less than 0.05 indicates an excellent fit; 0.05 ≤ RMSEA < 0.08 suggests a good fit; 0.08 ≤ RMSEA < 0.10 indicates an acceptable fit; and RMSEA ≥ 0.10 suggests a poor fit.For the AIC, a lower value indicates a better fit of the model.

Entropy Method and Indicator Selection
Entropy method: The entropy method is a technique for objectively assigning weights to indicators based on the magnitude of their information entropy.A smaller value of information entropy indicates a greater degree of dispersion for the indicator, implying more information content and therefore a higher weight.To overcome the limitations of the entropy method in dealing only with cross-sectional data and not being able to compare different varieties, this study adopted the panel entropy method [19].

Transcriptome Sequencing and Real-Time Fluorescence Quantitative Polymerase Chain Reaction (RT-qPCR) Validation of Candidate Genes
Based on the analysis of the above agronomic traits, 10 varieties of colored wheat were selected (Table A2).The total RNA was extracted using the TRizol method, and the transcriptional sequencing was performed by Beijing Genomics Company (BGI) (the instruments are listed in Table A3).The DNBSEQ high-throughput platform was utilized for sequencing, followed by subsequent bioinformatics analysis.
First, sample detection involves taking a certain amount of RNA samples, denaturing their secondary structures at an appropriate temperature, and enriching mRNA using oligo(dT) magnetic beads.In the subsequent step, mRNA obtained from the previous step is subjected to fragmentation by adding a disruption reagent and reacting at an appropriate temperature for a certain duration, resulting in fragmented mRNA.The fragmented mRNA is then subjected to a one-step synthesis reaction system previously prepared, and a reaction program is set up to synthesize single-stranded cDNA.A two-step synthesis reaction system (containing dUTP) is prepared, and a reaction program is set up to synthesize double-stranded cDNA.A reaction system is prepared, and a reaction program is set up to repair the ends of the double-stranded cDNA and add an A base at the 3 ′ end.A reaction system for adapter ligation is prepared, and a reaction program is set up to ligate the adapters to the cDNA.
Next, a PCR reaction system is prepared, and a reaction program is set up to amplify the products.According to the requirements of the product, appropriate detection methods are selected for library quality control.After denaturing the PCR products into single strands, a ligation reaction system is prepared, and a reaction program is set up to obtain single-stranded circular products, digesting the linear DNA molecules that are not circularized.The single-stranded circular DNA molecules undergo rolling circle amplification, forming DNA nanoballs (DNBs) containing multiple copies.The obtained DNBs are loaded into the mesh-like pores on a high-density DNA nanochip, and sequencing is performed using combinatorial probe-anchor synthesis (cPAS) technology.Each material had three biological replicates.The expression levels of three candidate dwarfing genes, Rht1, Rht-D1, and Rht8, were obtained through transcriptome analysis (specific method, the 2 −∆∆CT (CT, cycle threshold) method was employed to calculate the relative expression levels of the candidate genes in different varieties), and a heat map was generated.The expression characteristics of different materials were determined using the RT-qPCR method.The primer sequences are shown in Table A2.

RNA Extraction
First, 1 mL of Trizol was added to a tissue sample that was approximately the size of a green bean, homogenized for 1 min using a Tissue Ruptor, and then allowed to stand at room temperature for 10 min.Then, 200 µL of chloroform was added, thoroughly mixed, and centrifuged at 15,000 rpm for 5-7 min.After this, the supernatant was transferred to a 1.5 mL Eppendorf tube, 600 µL of chloroform was added and mixed in, and the mixture was centrifuged at 15,000 rpm for 5 min.
Next, the supernatant was transferred to a 1.5 mL Eppendorf tube, 500 µL of isopropanol was added and mixed, and the mixture was then centrifuged at 15,000 rpm for 10 min.The supernatant was then discarded, and the remaining RNA pellet was washed with 1 mL of 75% ethanol before being centrifuged at high speed for 5 min.Subsequently, the supernatant was again discarded, and the RNA pellet was air dried for 2-3 min, after which the dried RNA pellet was washed in RNA-free water.Finally, 1 µL of the total RNA was taken for OD260 measurement and quantification (for reagents, assay kits, and instruments, see Tables A4 and A5).

Reverse Transcription
The following reagents were sequentially added to a 200 µL PCR tube, totaling 12 µL; (10 − x) µL of DEPC (Diethylpyrocarbonate) water; 2 µL of Random primers/Oligo dT (50 pM/µL); and x µL of RNA (2 µg).The tube was then incubated at 65 • C for 5 min in a PCR machine, followed by immediate transfer to an ice bath and centrifugation at a high speed (above 5000× g) for 5 s.Subsequently, the following reagents were added to the PCR tube in the specified order, resulting in a total volume of 20 µL after their addition: 1 µL of RiboLock RNase Inhibitor; 4 µL of 5×Reaction Buffer; 2 µL of dNTP (Deoxyribonucleotide Triphosphate) Mix; and 1 µL of RevertAid Reverse Transcriptase.
The ingredients in the tube were mixed well, after which they were incubated at 25 • C for 5 min and then at 42 • C for 60 min (for samples with a high GC (gas chromatography) content, the temperature could be increased to 45 • C).Then, the tube was incubated at 70 • C for 5 min and subsequently centrifuged at a high speed (above 5000× g) for 5 s.The contents were then stored at −20 • C, but for longer storage, they could be stored at −80 • C.

Fluorescent Quantitative RT-qPCR Amplification
Sequence and primer design: sequences were referenced from the Gene Bank database for each target gene, and primers were designed using the National Center for Biotechnology Information (NCBI) Primer-Blast and synthesized by Suzhou Jinweizhi (Table A2).
RT-qPCR amplification, for sequence and primer design, sequences referred to the sequence of the target gene in the GenBank database, while primers were designed using the NCBI's Primer-BLAST and synthesized by Suzhou Jinwei Zhi (Table A2).
Specific method, the 2 −∆∆CT (CT, Cycle Threshold) method was employed to calculate the relative expression levels of the candidate genes in different varieties.Total RNA extraction: the total RNA was extracted using the Trizol method and quantified.

Basic Statistical Characteristics of Colored Wheat
The coefficients of variation for plant height, flag leaf length, flag leaf width, ear length, stem thickness, and yield of colored wheat ranged from 2.86% to 56.14%, with yield exhibiting the highest variability at 56.14%.As shown in Table 1, the highest yield was 5744.1 kg hm −2 for variety 70, while the lowest yield was 382.05 kg hm −2 , resulting in a range of 2709.15 kg hm −2 (Table 1).

Correlation Analysis of Main Agronomic Traits and Yield in Colored Wheat
A correlation analysis was performed to determine the correlations between the main agronomic traits and yield in colored wheat.The correlation between plant height and flag leaf length, as well as ear length, reached a highly significant level, with correlation coefficients of 0.306 and 0.404, respectively.The correlation between plant height and yield was significant, with a coefficient of 0.257.The correlation between flag leaf length and ear length was highly significant, with a coefficient of 0.334.Yield showed a significant correlation with ear length, with a coefficient of 0.45 (Table 2).

Using Structural Equation Modeling to Analyze the Relationship between the Main Agronomic Traits and Yield in Colored Wheat
The fit results of the initial model showed that χ 2 was 5.626 with four degrees of freedom (df), resulting in a χ 2 /df ratio of 1.4065 and p < 0.05.However, both the root mean square residual (RMR) and RMSEA were greater than 0.05 (Table 3).Therefore, the model required revision.During the revision, the significant path coefficients were retained and the non-significant path coefficients were removed.Structural equation models, as a priori models, allow for the retention of some statistically insignificant paths with biological significance based on the actual growth of crops.In the final revised model, the path from stem thickness to flag leaf width was eliminated.The revised model parameters are shown in Table 3, with a p-value of 0.025 and both RMR and RMSEA values of less than 0.05.In the final structural equation model established, all agronomic traits had positive direct effects on yield.Among them, plant height, ear length, and flag leaf width had significant impacts on yield, with path coefficients of 0.55, 0.40, and 0.27, respectively.The path coefficient of flag leaf length on yield was 0.17, while the influence of stem thickness on yield was relatively small.Each trait also exerted an effect on the final yield through their mutual effects (Figure 3).This indicates that random variations in the environment where colored wheat is grown have a significant impact on each agronomic trait, and once these trait characteristics are determined, the final yield is essentially determined.

Using the Entropy Method to Screen Colored Wheat Seed Resources
Structural equation modeling was used to determine the impacts of agronomi on yield.The next step involved using the entropy method to screen locally su varieties of colored wheat based on these agronomic traits (Tables 4 and A2), facil Figure 3. Effects of agronomic traits on the yield of colored wheat and relationships among agronomic traits as determined by structural equation modeling.Note: Red lines show significant paths, with greater line thickness reflecting the greater size of the absolute path coefficient.The dashed line shows that the path is not significant, but this path needs to be conserved based on the physiological process of colored wheat.The values beside the lines are standard path coefficients, and the values beside the rectangles and ovals are determination coefficients.

Using the Entropy Method to Screen Colored Wheat Seed Resources
Structural equation modeling was used to determine the impacts of agronomic traits on yield.The next step involved using the entropy method to screen locally suitable varieties of colored wheat based on these agronomic traits (Tables 4 and A2), facilitating promotion and demonstration planting for farmers and herders.The agronomic trait panel data for 72 colored wheat germplasm resources in the 2022-2023 trial base, including the plant height, flag leaf length, flag leaf width, main spike length, main spike stem thickness, and yield per hectare, represented attributes in six dimensions.The entropy method was utilized to calculate the weights of the six indicators and the final overall scores.The entropy method comprehensively evaluated the agronomic traits of the 72 wheat varieties, resulting in the calculation of composite indices for each variety.Finally, 10 colored wheat varieties that were suitable for the local conditions were selected (Table 4).The specific classification weights are shown in Table 5.

Screening for Dwarfing Genes in Colored Wheat
This research employed transcriptome sequencing (conducted by BGI Genomics) and RT-qPCR validation (sequence and primer design in Table A2).The estimated expression patterns for each biosynthetic gene were validated through RT-qPCR (Figure 4).Similar expression patterns were observed in the transcriptome data for randomly selected biosynthetic genes.Notably, Rht1, Rht8, and Rht-D1 showed high expression levels in colored wheat.The RT-qPCR reports for 10 different colored wheat varieties indicated the presence of three dwarfing genes that control plant height: Rht1, Rht-D1, and Rht8.The clustering heatmap revealed that semi-dwarf genes, such as Rht-D1 and Rht8, exhibited consistent trends in both plant height variation and the RT-qPCR reports (Figure 5), while the trend for Rht1 gene expression differed slightly.Analyzing the results from a yield perspective (Table 4), variety 6, characterized by dwarf stature, showed high expression in both the clustering heatmap (Figure 5) and RT-qPCR reports and ranked second in grain yield.This suggests that dwarf varieties not only resist lodging but also have high yields.

Discussion
Dwarfing genes have been a major driver of improved adaptation and performance with breeding and domestication across all the major cereals [20].In colored wheat, the identification and deployment of genes for semi-dwarf stature have promoted the commercial release and global adoption of wheat cultivars with greater yield potential Figure 5. Heatmaps depicting the relative expression of the studied reduced dwarfing genes in different colored wheat lines.The heatmaps were generated from the transcriptome sequencing of colored wheat leaf blades.The transcriptional sequencing was performed by Beijing Genomics Company (BGI).The DNBSEQ high-throughput platform was utilized for sequencing.The level of normalized gene expression from high to low is indicated by the color scheme from red to white to blue.Numbers 1~10: different colored wheat.dwarfing genes: Rht1, Rht8, and Rht-D1.

Discussion
Dwarfing genes have been a major driver of improved adaptation and performance with breeding and domestication across all the major cereals [20].In colored wheat, the identification and deployment of genes for semi-dwarf stature have promoted the commercial release and global adoption of wheat cultivars with greater yield potential and stability [21].Selection for reduced crop height remains a key objective of colored wheat breeding programs worldwide, owing to semi-dwarfs being less prone to lodging and producing greater amounts of grain to increase their harvest index and crop yields [22].In this study, three dwarfing genes, Rht1, Rht-D1, and Rht8, were identified in 10 colored wheat varieties (Figure 4).The RT-qPCR markers for each dwarfing gene were specific, allowing for gene differentiation.Sequencing of different wheat varieties for the Rht1, Rht-D1, and Rht8 alleles was used to create a heatmap, which indicated that semi-dwarfing gene varieties such as Rht1 and Rht8 roughly corresponded with the changes in plant height reported by RT-qPCR (Figure 4).However, the trend of the Rht1 gene differed somewhat from the RT-qPCR reports, possibly due to environmental variations.Molecular detection and transcriptome sequencing findings were generally consistent with the RT-qPCR results.Sequence-tagged site markers based on RT-qPCR can be used for the identification of the Rht1, Rht-D1, and Rht8 genes in colored wheat varieties and for screening in breeding progeny.
The discovery of Rht genes played a crucial role in the development of modern high-yielding wheat varieties due to their capacity to allocate more energy to grain production [23].With regards to the impact of the Rht1 and Rht8 mutations on plant height, the effects observed were consistent with those reported by Flintham et al. [24].Regarding leaf morphology traits, we reported an increasing trend in leaf length and width caused by dwarfing allele genes.This trend can be explained by an increased allocation of photosynthetic products in developing leaves [25].Additionally, higher length fertility led to a higher harvest index, supporting the views on semi-dwarfing genes Rht1 and Rht8.
The results of structural equation modeling indicated that the effects of each agronomic trait on yield were positive.Among them, plant height, ear length, and flag leaf width had significant impacts on yield.Our study suggests that the random variations in the environment where colored wheat is grown have a significant impact on each agronomic trait.However, once these agronomic trait characteristics are determined, the final yield can be essentially predicted.Plant height, as a crucial agronomic trait in cereal crops, has a significant impact on crop yield potential and stability.Studies have shown that reducing plant height can enhance a plant's lodging resistance [26,27].In the 1960s, dwarfing genes were first applied in wheat breeding, triggering the first "Green Revolution" [28].The positive role of the Green Revolution gene Rht1 that led to higher grain yield potential has been extensively reported worldwide [24].In agreement with our results, Sherman et al. [29] and Tang et al. [30] indicate that the semi-dwarfed plants carrying these genes are associated with an increase in productive tillers and grains per spike, leading to higher GN and grain yield.
To date, a total of 25 dwarfing genes (Rht genes) have been identified in wheat, and each Rht gene has different effects on wheat plant height.Among them, the most widely used dwarfing genes in wheat breeding for reduced plant height are Rht-D1b from Norin 10 and Rht8 from Japanese red wheat (AKAK-OMUGi) [31], respectively.Studies have shown that when the Rht-D1b gene is present alone, the dwarfing effect is approximately 24%, and when both genes are present simultaneously, there is a cumulative effect with a dwarfing effect of nearly 60% [32,33].The dwarfing effect of Rht8 is 11% [34], indicating a weaker dwarfing ability compared to Rht-D1b [35].This is consistent with the high yield of the cultivars carrying dwarf genes in this study.Variety 6 is a short-stemmed variety with high expression in cluster heatmaps and RT-qPCR reports, ranking first in grain yield.This indicates that short-stemmed varieties can not only resist lodging but also have high yields.

Conclusions
This study identified that each agronomic trait has a positive direct effect on yield.Additionally, each trait also influences the final yield through its intercorrelated effects.This suggests that random environmental variations during the production of colored wheat significantly impact various agronomic traits.Regarding leaf morphology traits, we observed an increasing trend in leaf length and width induced by dwarfing allele genes.This phenomenon can be attributed to an enhanced allocation of photosynthetic products in developing leaves.Analyzing the results from a yield perspective reveals that shortstemmed varieties not only exhibit lodging resistance but also demonstrate an increasing trend in leaf length and width, contributing to higher yields.Therefore, researching dwarfing genes is of great significance, and our study provides valuable insights for the development of new high-yielding varieties of colored wheat.They hold potential for local cultivation and promotion.

Figure 1 .
Figure 1.Yangchangzi Experimental Base of the Hinggan League Agricultural and Animal Husbandry Technology Extension Center (Yangchangzi village, Yilelite Town, Ulan Hot City, Hinggan League, Inner Mongolia).(a) Photographs illustrating the maturity stage of wheat; (b) photographs depicting the wheat flowering stage; (c) photographs illustrating the emergence stage of wheat seedlings; (d) Yang chang zi Experimental Base.

Figure 1 .
Figure 1.Yangchangzi Experimental Base of the Hinggan League Agricultural and Animal Husbandry Technology Extension Center (Yangchangzi village, Yilelite Town, Ulan Hot City, Hinggan League, Inner Mongolia).(a) Photographs illustrating the maturity stage of wheat; (b) photographs depicting the wheat flowering stage; (c) photographs illustrating the emergence stage of wheat seedlings; (d) Yang chang zi Experimental Base.

Figure 2 .
Figure 2. Initial modeling of the relationship between yield and agronomic traits in colored wheat.The initial model contains 16 relationships between parameters, which are shown by arrows, where e1 through e6 indicate errors.GY: grain yield; LI: leaf length; LW: leaf width; HL: head length; ST: stem thickness; PH: plant height.

Figure 2 .
Figure 2. Initial modeling of the relationship between yield and agronomic traits in colored wheat.The initial model contains 16 relationships between parameters, which are shown by arrows, where e1 through e6 indicate errors.GY: grain yield; LI: leaf length; LW: leaf width; HL: head length; ST: stem thickness; PH: plant height.

Genes 2024 ,Figure 3 .
Figure 3. Effects of agronomic traits on the yield of colored wheat and relationships agronomic traits as determined by structural equation modeling.Note: Red lines show sign paths, with greater line thickness reflecting the greater size of the absolute path coefficie dashed line shows that the path is not significant, but this path needs to be conserved based physiological process of colored wheat.The values beside the lines are standard path coeffi and the values beside the rectangles and ovals are determination coefficients.

Figure 4 .Figure 4 . 18 Figure 5 .
Figure 4. RT-qPCR results of three dwarfing genes in colored wheat.RT-qPCR: real-time fluorescence quantitative polymerase chain reaction.Values are the mean ± SE of three biological replicates, and different letters indicate significant differences among different samples using Duncan's test (p < 0.05).

Table 1 .
The coefficient of variation for various agronomic traits of different colored wheat varieties.
Note: SD, standard deviation; CV, coefficient of variation.

Table 2 .
Correlation analysis of agronomic characteristics of colored wheat.

Table 3 .
Initial and final fit indices of structural equation models.

Table 4 .
A list of the top 10 ranked wheat varieties determined by the entropy method using 72 colored wheat.

Table 5 .
The weights of the top 10 ranked wheat varieties determined by the entropy method using 72 color-coded samples.