The role of organic compounds in modifying functional relationships and morphophysiological traits of wheat under water stress: a correlation network analysis Authors: Seyed Karim Irani Takleh, Morteza Barmaki, Ali Omrani, Seyed Habib Shojaei Published: 2026 ABSTRACT AbstractWheat (Triticum aestivum L.), playing a significant role in providing food security and agricultural economy, is one of the most important cereals in the world. The present experiment aimed to determine the optimal levels of organic compounds (fulvic acid, humic acid, and amino acids) to promote sustainable wheat (cv. Taktaz) cultivation in dry and semi-dry regions while reducing dependency on irrigation water. The experiment was conducted as a split-factorial based on a randomized complete block design (RCBD) with three replications during two cropping years, 2023–2025. Results indicated that three-way interactions were statistically significant at the 0.01 level for all traits except number of fertile spikes and number of grains per spike. According to the polygon view of the traits, applying high amounts of fulvic acid as well as average amounts of the amino acid can have positive effects on controlling stress. In addition, increasing the concentration of humic acid can compensate for low concentrations of compounds like fulvic acid and amino acids. According to the treatment stability diagram, treatments 3 (humic acid 12 L h− 1) and 6 (amino acid 1.5 L h− 1, humic acid 12 L h− 1) under normal irrigation conditions, and 27 (fulvic acid 6 L h− 1, amino acid 3 L h− 1, humic acid 12 L/ha) under drought stress conditions, were selected as desirable treatments. The fact that increasing organic matter, like fulvic acid and amino acids, is helpful in plant metabolism regulation and resistance enhancement against drought stress reveals the effects of compounds included in these treatments on the plant’s resistance to the stress. The treatment with the highest amount of organic compounds (6 L ha− 1 fulvic acid, 3 L ha− 1 amino acid, 12 L ha− 1 humic acid) was identified as the treatment with positive coefficients in relation to the first two components in both years. Under drought stress conditions in both years, grain yield had positive coefficients for the first two principal components. Likewise, flag leaf width had positive effects on the fourth component in both years. Treatments with high fulvic acid and medium humic acid (3 L ha− 1 fulvic acid, 1.5 L ha− 1 amino acid, 12 L ha− 1 humic acid; 6 L ha− 1 fulvic acid; and 6 L ha− 1 fulvic acid, 8 L ha− 1 humic acid) consistently appeared with strong relationships in the correlation network in both years. In general, treatments with moderate fulvic acid and high humic acid can be selected as desirable treatments under both normal and stress conditions. Selecting these treatments indicated that using fulvic acid and humic acid, along with an amino acid that plays a significant role, had the highest positive effects on yield improvement under both normal and irrigation stress conditions.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1038/s41598-026-58518-7.Keywords: Wheat, Drought stress, Fulvic acid, Amino acid, Humic acidSubject terms: Ecology, Ecology, Physiology, Plant sciences Introduction Wheat is one of the most important food crops for humanity, cultivated on approximately 225 million hectares of agricultural land worldwide, with a global production of 750 million tons1. Increasing wheat demand for culinary uses is due to its being a rich source of carbohydrates and gluten2. Being the main food for the Asian population, this plant is a vital cereal product worldwide. Wheat encounters various stresses that restrict its growth, development, and yield. Abiotic stresses have a significant impact on the growth and yield of agricultural crops. This type of stress encompasses any unfavorable environmental conditions that impede proper plant development and, due to severe and persistent adverse conditions, can induce morphological damage in plants3,4. Water deficiency is considered the main limiting factor for Wheat. Climatic extremes such as droughts, salinity, and elevated temperatures, pose serious challenges to food security and crop productivity5,6, and semi-arid regions are projected to face the greatest intensification of water scarcity under future climate change scenarios7. Drought stress is an abiotic stress and the main limiting factor in the growth and productivity of many agricultural products worldwide. This type of stress has caused the drought-induced productivity loss of agricultural products to reach 34% in 20218. Escalating drought stress intensity is harmful to crop growth and leads to dramatic changes in the morphological characteristics of crops9. Drought stress has many negative influences on crop growth, which can decline the trait yield considerably through changing the crop’s key biochemical and physiological systems10. Plants under drought stress exhibit a reduction in photosynthetic rate, which is associated with impaired growth and increased incidence of early senescence11. Drought is one of the most important environmental factors that inhibits photosynthesis and reduces plant growth and productivity. Crop plants such as wheat are particularly sensitive to soil drought, especially during the reproductive stage, which is extremely dependent on the plant’s water status12. One of the most valuable bio-stimulants that can be externally used to increase crop resistance during drought stress is humic acid. This compound consists of plant or animal residues and can play a significant role in crop growth and development as a hormone-like molecule. Due to its low molecular weight, humic acid is absorbed readily by crops13. Fulvic acid is also one of the important factors in metabolism and positively affects wheat seedling growth14,15. This compound is involved in photosynthesis, biochemical transformations of nitrogen, carbon, and sulfur, as well as synthesis and catabolism of protein amino acids and nucleic acids16. Findings have shown that applying fulvic acid and humic acid in low-irrigation regimes can increase yield and water consumption efficiency in wheat17. Humic acid, by inducing the accumulation of proline and soluble sugars, increases cellular osmotic pressure and maintains leaf relative water content under drought stress18. fulvic acid, owing to its low molecular weight, readily crosses the cell membrane and enhances the uptake of nutrients such as iron and zinc19. Amino acids play various important roles in plants. They are rich sources of nitrogen and can reduce stress intensity in plants20. Drought resistance in plants is enhanced by using amino acids, which minimize the level of unfavorable effects and increase the amount of regulated accumulation of salts21. Recommendations based on the results of an experiment on assessing the effects of foliar application of amino acids on decreasing the effects of drought stress reveal that applying these compounds is a valuable approach to reducing the drought stress and improving crop yield in wheat22. In a study conducted to investigate the effects of nitrogen and water stress on five emmer wheat landraces compared with durum and bread wheat, emmer wheat exhibited greater stability under water-deficit conditions but showed a negative response to nitrogen fertilizer application23. Another study demonstrated that drought stress leads to reductions in leaf water potential, relative water content, grain yield, and yield components; however, the application of organic compounds such as salicylic acid can mitigate the adverse effects of drought and enhance the studied traits24. Various research has been done on the effects of humic acid, amino acids, and fulvic acid on yields of different plants, especially wheat, in normal and different irrigation regimes25–27,. However, simultaneous analysis of interactions among these compounds while assessing crop traits in wheat under normal and drought stress conditions provides a novel approach in the cultivation of this crop. The present study aimed to: (1) screening and selecting suitable treatments (amino acid × fulvic acid × humic acid) for wheat cultivation under normal and drought stress conditions; (2) assessing the stability of different treatments in both experimental years; (3) analyzing the effects of treatment × trait; and (4) assessing the relationships among the crop traits. Materials and methods Geographical information of the experimental site The present study was conducted in the research field of the Research Center of Agriculture and Natural Resources in Moghan, Ardabil, Iran, in 2023–2025. This region has a semiarid climate with hot summers and cold winters. According to the meteorological reports of Moghan’s weather station, the average annual maximum and minimum temperatures are 35 °C and 8 °C, respectively. So far, absolute maximum temperature and absolute minimum temperatures have been reached 42.5 °C and − 5.16 °C, respectively. The average amount of annual precipitation has been 251 mm, and annual relative humidity has been around 71% on average. The highest precipitation occurs in fall, winter, and early spring. Figure 1 shows the geographical and climatic information of the experimental site. Geographical and climatic information of the experimental site. (a: The geographical location of the cultivation area, b: Average temperature and precipitation of the months under study in 2024, c: Average temperature and precipitation of the months under study in 2025, d: Comparison of the average temperature of the months under study in 2024 and 2025, e: Comparison of the average precipitation of the months under study in 2024 and 2025). Chemical compounds Chemical compounds used in this experiment, including humic acid, fulvic acid, and the amino acid, were procured from reputable companies. Fulvic acid was bought from Akin Tarim Company, Turkey. Isabion amino acid and humic acid were supplied by Euro Alga Company. The above-mentioned compounds were analyzed and verified by the laboratory department of the Soil and Water Organization. Experimental design This study was conducted as a factorial split-plot based on a randomized complete block design with three replications in two successive crop years. Main plots included normal irrigation regime (N) and irrigation cut-off before flowering onset (D), while foliar application of fulvic acid (at three levels, 3 L ha− 1, 6 L ha− 1 and 0 L ha− 1 as the control treatment), Isabion amino acid (1.5 L ha− 1, 3 L ha−1and control) and humic acid (as seed treatment at a rate of 6 L per 1,000 kg of seed, and as a substance solved in irrigation water at the rate of 8 L ha−1and 12 L ha− 1) were the sub-plots. Seed sowing was recorded in November and crop harvest was recorded in July. Foliar application was applied using a spraying drone and carried out after the end of the cold period and at the tillering stage. the drought stress was also imposed after anthesis and during the grain-filling stage. Prior to the experiment, soil samples taken from 0 to 30 cm depth were sent to an accredited soil laboratory to be analyzed for physical and chemical characteristics. Table 1 reports the results of analyzing the soil samples. The genotype used in this study is one of the newest commercial varieties introduced for the hot and humid climate of Iran, and it is widely cultivated. The commercial variety used is named Tektaz. Wheat seeds of cv. Taktaz (with pedigree: SAUAL/3/MILAN/S87230//BAV92) were disinfected using a suitable fungicide and then planted at a density of 400 seeds per square meter by a grain drill specialized for irrigated cereals. Weed control was done mechanically and, in case of necessity, by applying herbicides, 2,4-D, Granstar and Topik, at dosages recommended by the manufacturer. In order to control pathogens, especially fungal diseases common in Moghan, tilt and folicur fungicides were applied at recommended rates and required times. Finally, yield and agronomic traits were measured after harvest. All stages of planting, field management, and harvesting were carried out mechanized under the supervision of a mechanization specialist. Table 2 shows the codes and characteristics of the treatments and analyzed traits in this study. Table 1Texture and chemical properties of the experimental soil.K(ppm) P (ppm)N Total(ppm)O.C(%)Soil textureSand(%)Clay(%)T.N.V(%)pHEC(dS m− 1)3471.20.060.527Clay19.6460.757.781.129 Texture and chemical properties of the experimental soil. Table 2Codes and characteristics of the treatments and analyzed traits in the study.Treatmentcode.fulvic Acidamino Acidhumic AcidTraitCode.Traittr1 (1)FA0AA0Ha0PLHPlant Heighttr2 (2)FA0AA0Ha8PLPeduncle lengthtr3 (3)FA0AA0Ha12FLLFlag leaf lengthtr4 (4)FA0AA1.5Ha0FLWFlag leaf widthtr5 (5)FA0AA1.5Ha8SLSpike lengthtr6 (6)FA0AA1.5Ha12TGWThousand-grain weighttr7 (7)FA0AA3Ha0GYGrain yieldtr8 (8)FA0AA3Ha8BYBiological yieldtr9 (9)FA0AA3Ha12NSSNumber of seeds per spiketr10 (10)FA3AA0Ha0NFSNumber of fertile spikestr11 (11)FA3AA0Ha8HIHarvest Indextr12 (12)FA3AA0Ha12tr13 (13)FA3AA1.5Ha0tr14 (14)FA3AA1.5Ha8tr15 (15)FA3AA1.5Ha12tr16 (16)FA3AA3Ha0tr17 (17)FA3AA3Ha8tr18 (18)FA3AA3Ha12tr19 (19)FA6AA0Ha0tr20 (20)FA6AA0Ha8tr21 (21)FA6AA0Ha12tr22 (22)FA6AA1.5Ha0tr23 (23)FA6AA1.5Ha8tr24 (24)FA6AA1.5Ha12tr25 (25)FA6AA3Ha0tr26 (26)FA6AA3Ha8tr27 (27)FA6AA3Ha12(FA0: Control, FA3: 3 L ha− 1, FA6: 6 L ha− 1, AA0: Control, AA1.5: 1.5 L ha− 1, AA3: 3 L ha− 1., HA0: Control, HA8: 8 L ha− 1, HA12: 12 L ha− 1, N: Normal Irrigation Condition, D: Drought Stress). Codes and characteristics of the treatments and analyzed traits in the study. (FA0: Control, FA3: 3 L ha− 1, FA6: 6 L ha− 1, AA0: Control, AA1.5: 1.5 L ha− 1, AA3: 3 L ha− 1., HA0: Control, HA8: 8 L ha− 1, HA12: 12 L ha− 1, N: Normal Irrigation Condition, D: Drought Stress). Statistical analysis Mean values obtained from two cropping years were used for analysis of variance (ANOVA) and mean comparison using the Least Significant Difference (LSD) method. By using the data from both years under normal and stress conditions, graphical analysis was performed for evaluation of (1) the treatments (using a polygon view); (2) yields and trait stability of the treatments; (3) relationships among the traits; and (4) PCA. Principal component analysis (PCA) was employed to reduce the dataset of evaluated variables to their essential features. Principal components represent a limited number of linear combinations of the original variables that maximally account for the total variance across all variables. This method can yield graphical outputs in the form of a biplot or present results in numerical format28. Network analysis, by graphically visualizing correlation relationships among treatments or traits, identifies complex interaction patterns and key nodes that play a central role in these associations29. In addition, considering the unit differences of the traits, standardization was done to eliminate the units using the following equation: where Z, X, µ, and σ are the standard score, trait original data (variable raw data), variable mean, and variable standard deviation, respectively. Excell, SAS v9.4, Excelstat 2020, Past.v4.17 and Genstatv12.1 software programs were used for data analysis. Results ANOVA and comparison of means As the results of ANOVA on mean data of both years under normal irrigation conditions indicated, the effect of fulvic acid on plant height, flag leaf length, flag leaf width, grain yield, biological yield, number of fertile spikes, and harvest index was significant at the 0.01 and 0.05 levels. The amino acid had a significant effect on all traits except flag leaf length, thousand grain weight, and biological yield at the 0.01 and 0.05 levels. The effect of humic acid on peduncle length, flag leaf width, thousand grain weight, grain yield, biological yield, number of fertile spikes, and harvest index was statistically significant at the 0.01 and 0.05 levels as well. The interaction effect between fulvic acid and the amino acid was significant for all traits except thousand grain weight. The interaction effect of fulvic acid × humic acid was significant for peduncle length, flag leaf length, flag leaf width, spike length, grain yield, number of fertile spikes, and harvest index. The interaction effect of amino acid × humic acid was significant for all traits except the number of fertile spikes. The three-way interaction effect was significant for all traits except thousand grain weight and number of fertile spikes. The highest value for the coefficient of variation percentage (C.V.%) was recorded for thousand grain weight (27.3), while the least value of that belonged to harvest index (6.3). Based on ANOVA on mean data of both years under stress conditions, fulvic acid affected all traits except spike length significantly at the 0.01 and 0.05 levels. The amino acid had a significant effect on plant height, flag leaf length, flag leaf width, spike length, thousand grain weight, grain yield, number of grains per spike, and harvest index at 0.01 and 0.05 levels. The effect of humic acid was statistically significant at 0.01 and 0.05 levels on plant height, peduncle length, spike length, thousand grain weight, grain yield, biological yield, and number of fertile spikes. Moreover, significant effects were revealed for the interaction effect of fulvic acid × amino acid on all traits, the interaction of fulvic acid × humic acid on plant height, flag leaf length, flag leaf width, spike length, grain yield, number of grains per spike, and harvest index, and the interaction of humic acid × amino acid on all traits. Three-way interaction effect of fulvic acid × amino acid × humic acid was significant for all traits except number of grains per spike. The highest and lowest values of C.V.% were recorded for thousand grain weight (27.3) and peduncle length (7.3), respectively (Table 3). Table 3Analysis of variance on the effect of experimental treatments on evaluated traits in normal and stress conditions.S.O.V.DfPLHPLFLLFLWSLTGWGYBYNSSNFSHI NormalBlock22636**39.8*46.2**14.5**64.1*144.9ns0.01ns35.3**53.2ns0.9ns5.8**fulvic acid(a)2157.8**24.1ns34.1**2.09**10.1ns247.9ns0.9*2.6**7.1ns1.2**0.07**amino acid(b)2990**18.5*1.7ns0.5**2.4*262.8ns0.16*0.38ns41.1*0.2*0.003**a*b4247**11.3*0.9*2.8**12.8*456ns0.21*1.2**50.6*0.13*0.03**Error116132.131.76.091.0614.44670.43.759.31.50.13humic acid (c)2114.7ns29.9*0.25ns0.33*10.5ns178*0.23*0.4*1.6ns0.8*0.03*a*c4120.7ns27.3*2.18*1.01*10.5*420ns0.01*0.06ns7.02ns0.4*0.02*b*c4320.6**22.9*1.8*0.4*17.9*458**0.18*0.58*15ns1.6*0.01*a*b*c8101.9*22.8*2.1*0.33*10.5*214ns0.1*0.4*6.5ns0.6*0.01*Error366510.51.60.411.42310.150.5515.90.340.015CV%--8.97.977.118.821.727.31615.79.327.16.3StressBlock2358**632**18.5ns19.9**11.3*390.8ns3.2ns22.4**37.6ns0.03ns0.34nsfulvic acid (a)261.8*62.8**24.6*3.1**0.8ns3139**0.04*0.14*45.7*1.08**0.5*amino acid (b)2190*1.9ns31.7*0.2*3.5*641.5**0.2*0.7ns0.8*0.07*0.3*a*b496*19.5**11.3*1.9**2.4*334*0.7**2.2**27.3*2.3**0.18*Error11640750.114.41.56.4798.90.42.14170.80.34humic acid (c)2388**25.9*8.7ns0.33ns8.5**254*0.19*0.95*43.4ns0.6*0.34nsa*c492*9.1ns15.1*0.6*1.6*494ns0.14*0.26ns0.28ns0.5*0.14*b*c4414**12.4*25.2*0.7*1.05*1055*0.3*0.7*0.82**1.2**0.08*a*b*c8122**22**5.01*0.8*5.09**847**0.11*0.5**43.3ns1.5**0.2*Error36779.114.70.51.57130.130.422.50.30.18CV%--10.17.323.317.017.827.615.514.811.525.923*,** and ns: Significant (α = 5%), highly significant (α = 1%) and non-significant, respectively. CV: Coefficient of variation.(PLH: Plant Height, PL: Peduncle Length, FLL: Flag Leaf Length, FLW: Flag Leaf Width, SL: Spike Length, TGW: Thousand Grain Weight, GY: Grain Yield, BY: Biological Yield, NSS: Number of Seeds per Spike, NFS: Number Of Fertile Spike, HI: Harvest Index). Analysis of variance on the effect of experimental treatments on evaluated traits in normal and stress conditions. *,** and ns: Significant (α = 5%), highly significant (α = 1%) and non-significant, respectively. CV: Coefficient of variation. (PLH: Plant Height, PL: Peduncle Length, FLL: Flag Leaf Length, FLW: Flag Leaf Width, SL: Spike Length, TGW: Thousand Grain Weight, GY: Grain Yield, BY: Biological Yield, NSS: Number of Seeds per Spike, NFS: Number Of Fertile Spike, HI: Harvest Index). Based on the comparison of means using the LSD method, treatments tr1 and tr4 were selected as the desirable treatments regarding all traits. Treatment tr12 was selected as the undesirable one. Treatments identified as desirable included: tr23 regarding plant height, tr3 and tr5 regarding peduncle length, tr6 in respect of flag leaf length, tr27 for flag leaf width, tr20 for spike length, tr1 for thousand grain weight, tr7 and tr2 for grain yield, tr7 for biological yield, most of the treatments for number of grains per spike, tr23 for number of fertile spikes, tr22, tr25 and tr10 for harvest index (Table 4). According to the comparison of means using data from the stress conditions, treatments tr1 and tr11 were selected as desirable, whereas tr2 and tr7 were considered as undesirable concerning all traits. Treatments identified as undesirable include tr14 and tr23 for plant height, tr3 and tr24 for peduncle length, tr1 for flag leaf length, tr24 for flag leaf width, tr16 and tr26 for spike length, tr11 for thousand grain wight, tr9 for grain yield and biological yield, tr1 and tr17 for number of grains per spike (Table 5). Table 4Comparison of means using the LSD method under normal conditions.TreatmentsRankPLH(cm)PL(mm)FLL(cm)FLW(cm)SL(cm)TGW(g)GY(t ha− 1)BY(t ha− 1)NSSNFSHItr1182.6HI38.7 g-i18.2a-f3.9a-e15.1b67.3a2.62a-d4.78a-f42ab2.4a-d1.81etr21684.8ef40.6 b-i18.3a-e4.1ab16b39.2b2.69ab5.19ab44.6a1.94d-g1.91b-etr31186.4ef41.9a-f18.3a-e3.6b-h15.3b39.2b2.55a-f4.94a-e44a1.94d-g1.91b-etr4293a-d41.7a-g18.8a-c3.3d-h15.9b38.4b2.23ef4.45c-f43.6a1.94d-g1.99bctr5897a44a18.4a-d3.8a-f15.5b37.5b2.52a-f5a-c43.6a2.38a-e1.96c-dtr62195.1a-c39.6d-h19.3a3.1gh15.5b37.8b2.63a-d4.95a-e44a2c-g1.87detr7988.4ef40b-i18.3a-e3.4c-h15.4b38.5b2.7a5.23a42.9a2.25a-f1.94b-dtr82592.4a-d43abc18.9ab3.3d-h14.9b40b2.48a-d4.83a-f41.6ab2.19a-g1.93b-dtr92482.5HI41.3a-h18.4a-d3.2fgh15.3b34.8b2.64a-c5a-c42.3ab2.55a-c1.88c-etr102686.5ef41.1 a-i18.3a-e3.3c-h14.4b38.6b2.54a-f5a-c38.9b1.97d-g2abtr112780.7j39f-i16.4 h3.3 e-h14.9b41.1b2.53a-f4.93a-e41.7ab1.88e-g1.98b-dtr121884.5fj38.6hi17 g-h3.3e-h15.5b35.7b2.37a-f4.61a-f41.8ab2.22a-f1.97b-dtr13687.9ef43.4ab17.4d-h3.5b-h16b36.4b2.32c-f4.49c-f44.6a1.97d-g1.93b-dtr141297a40.4b-i17.9b-g3.1gh15b36.8b2.58a-e4.99a-d44.6a2.33a-d1.96b-dtr152093.2a-d39.5d-i17.2e-h3.1gh14.9b42.3b2.53a-f4.77a-f43.8a1.88e-g1.9b-etr162294.8a-c39.3e-i17.1f-h3.3d-h16b34.6b2.19f4.14f43.6a2.16b-g1.9b-etr171486.3ef40.5 b-i16.9gh3.1gh16.1b41.6b2.27d-f4.28ef43.1a2.11e-g1.88c-etr181093.4a-d42.1a-e17.3d-h3.5 b-h15.4b40.3b2.4a-f4.69a-f41.2ab1.77 fg1.97b-dtr19591.5a-e40.3c-i17.8b-g3.5b-h15b42.6b2.27c-f4.49c-f44a2c-g1.99bctr201990.4a-f40.5 b-i17.8 b-g3.2fgh20.8a40.2b2.39a-f4.54a-f43.6a2.36a-e1.93b-dtr212388.2ef41.1a-h17.5d-h3.1gh15.3b38.8b2.48a-f4.68a-f44.5a2.36a-e1.91b-dtr22483.5 g-j39.1f-i17.5 d-h3.7a-g15b40.4b2.22ef4.35d-f43.3a2.36a-e2abtr231597.9a42a-f18.4a-d3.1gh15.6b40.5b2.3c-f4.46c-f41.9ab2.72a1.97b-dtr24792.2a-e40.75 b-i17.3d-h3.8a-f15.3b41.1b2.35a-f4.51b-f42.9a1.66 g1.93b-dtr25395.9ab38.5hi17.7c-g3.98a-c15.8b44.3b2.33b-f4.82a-f41.3ab2.66ab2.1atr261396.2ab42.4a-d18b-g3.96a-d15.3b43.1b2.38a-f4.71a-f41.2ab2.38a-e1.96b-dtr271789b-f38.1i17.8 b-g4.24a15.3b46.1b2.41a-f4.73a-f42.6ab2c-g1.93b-d(PLH: Plant Height, PL: Peduncle Length, FLL: Flag Leaf Length, FLW: Flag Leaf Width, SL: Spike Length, TGW: Thousand Grain Weight, GY: Grain Yield, BY: Biological Yield, NSS: Number of Seeds per Spike, NFS: Number Of Fertile Spike, HI: Harvest Index). Comparison of means using the LSD method under normal conditions. (PLH: Plant Height, PL: Peduncle Length, FLL: Flag Leaf Length, FLW: Flag Leaf Width, SL: Spike Length, TGW: Thousand Grain Weight, GY: Grain Yield, BY: Biological Yield, NSS: Number of Seeds per Spike, NFS: Number Of Fertile Spike, HI: Harvest Index). Table 5Comparison of means using the LSD method under the drought stress conditions.TreatmentsRankPLH (cm)PL(mm)FLL(cm)FLW(cm)SL(cm)TGW(g)GY(t ha− 1)BY(t ha− 1)NSSNFSHItr1291.97ab39.6d-f21.2a4.44a-e15.52e-g41.5c2.1 fg3.9e45.2a2cdefghi45.27atr22784.49b-f40.1b-f15.8bc3.97de16.9a-c39.4c2.1e-g4e38.9gh1.83efghji38.94ghtr32088.15a-e43.7a16.8bc4.14b-e14.7 g41.6c2.2b-g4.1de39.5e-h1.88defghji39.55d-htr41488.93a-d40.3b-e15.8bc3.86e16.59a-e39.6c2.3b-g4.4b-e43.6a-e1.55ji43.66a-etr51288.85a-d40.4b-f15.9bc3.85e16.8a-d40.5c2.6ab4.7a-c43.5a-f2.33bcde43.55a-ftr61680.98d-f39.6d-f15.2c4.47a-e15.66d-g37.1c2.3b-g4.2b-e40.5a-h2.88a40.5b-htr72685.55b-f39.7c-f16.3bc3.85e16.97a-c40.3c2.3b-g4.3b-e38.9gh1.77fghji38.94ghtr81885.64b-f39.4d-f16bc4de15.55e-g42.4c2.6ab4.8a-c43.7a-d2.27bcdef43.77abcdtr9987.57a-f39.3ef15.5bc4de16.19a-f42.5c2.7a5.2a43.2a-g1.44j43.22a-gtr10886b-f38.3f15.7bc4.10ce16b-f46.8bc2.6a-c4.7a-d44.1a-c2.27bcdef44.11abctr11190.66ab40.3b-f15.7bc4.3a-e15.71d-g85.7a2.3b-g4.3b-e38.1 h1.72ghji38.11 htr121588.52a-d40.7b-f15.8bc3.85e15.83c-f45.1bc2.5a-d4.7a-d41.6a-h2cdefghi41.61a-htr132484.87b-f39.3ef14.3c3.98de16.49a-f45.1bc2.1e-g4e39.5d-h2.22bcdefg39.5d-htr14695.55a42.7ab15.8bc3.81e16.83a-d47.6bc2.4b-f4.5b-e40.3b-h1.9defghij40.33b-htr152580ef41.3a-e16bc4ce15.94b-g45.6bc2 g3.9e40.1c-h2.22bcdefg40.11c-htr161785.88b-f41.7a-e15.9bc4.74a-c17ab46.4bc2.2c-g4.1de41.3a-h2.16bcdefgh41.38a-htr171082.25c-f42.2a-d16.2bc4.6a-d16.27a-f44.7bc2.5a-d4.8ab44.7ab2.5abc44.74abtr18479.44f40.1b-f16.41bc4.19a-e15.5e-g69.1ab2.3b-g4.3b-e38.7 h2.5abc38.77 htr19590a-c42.4a-c19ab4.59a-d15.71d-g45.3bc2.2d-g4.2b-e41.7a-h2.61ab41.75a-htr20789.17a-c42.7ab16.8bc4.55a-d15.91b-g48.5bc2.2c-g4.3b-e41.1a-h2.5abc41.11a-htr211188.14a-e39.9b-f16.2bc4.77ab16.29a-f48.3bc2.3b-g4.3b-e40.3c-h2.61ab40.30c-htr221979.6f40.2b-f16bc4.48a-e15.94b-f49.7bc2.4b-f4.5b-e41.9a-h1.66hji41.98a-htr23394.73a42.5a-c16.8bc4.32a-e15.39 fg43.8c2.4a-e4.4b-e40.9a-h2.66ab40.96a-htr242382.23c-f44.a16bc4.82a15.53e-g42.4c2.3b-g4.4b-e39.8c-h1.72ghji39.85c-htr251394.51a41.5a-e16.5bc4de15.92b-g45.8bc2.3b-g4.3b-e39.7c-h1.94defghji39.72c-htr262286.2b-f42.7ab16.9bc4.15a-e17.35a39.7c2.2c-g4.4b-e39.1f-h1.94defghij39.16fghtr272184.4b-f41.6a-e17.4bc4.35a-e15.71d-g44.9bc2.3b-g4.2c-e39.2e-h2.38abcd39.27efgh(PLH: Plant Height, PL: Peduncle Length, FLL: Flag Leaf Length, FLW: Flag Leaf Width, SL: Spike Length, TGW: Thousand Grain Weight, GY: Grain Yield, BY: Biological Yield, NSS: Number of Seeds per Spike, NFS: Number Of Fertile Spike, HI: Harvest Index). Comparison of means using the LSD method under the drought stress conditions. (PLH: Plant Height, PL: Peduncle Length, FLL: Flag Leaf Length, FLW: Flag Leaf Width, SL: Spike Length, TGW: Thousand Grain Weight, GY: Grain Yield, BY: Biological Yield, NSS: Number of Seeds per Spike, NFS: Number Of Fertile Spike, HI: Harvest Index). Treatment tr1 was selected as desirable in both normal and stress conditions based on the findings of data analysis in both conditions. There were similarities among various traits regarding treatment desirability. The following treatments were selected as desirable in both normal and stress conditions: tr23 for plant height and number of fertile spikes, and tr1 and tr17 for number of grains per spike. Evaluation of treatments using a polygon view Results of the polygon diagram on data derived from the first year under normal conditions showed that the first and second components explained 39.20% and 15.13% (54.33% in total) of the whole variance. Treatments 3, 5, 8, 12, 15, 16, 22, 23, and 27 were selected as desirable treatments regarding all traits (Fig. 2a). According to the findings of the polygon chart on data derived from stress conditions, more than 46% and more than 15% (more than 61% in total) of the variance were explained by the first and second components, respectively. Treatments 8, 15, 19, 21, and 22 were identified as desirable for all traits measured in this study (Fig. 2b). Assessing the polygon chart for the second year under normal conditions indicated that more than 46% and more than 15% (more than 61% in total) of the total variance were explained by the first and second components, respectively. Treatments identified as desirable included treatments 19, 21, 22, 15, and 8 (Fig. 2c). Polygon chart for normal conditions in the second year revealed that the first and second components explained 31.31% and 16.94% (total of 48%) of the whole variance, respectively. Based on this chart, treatments 18, 22, 23, 2, 7, 5, and 24 were known as desirable (Fig. 2d). As was observed in the polygon chart for stress conditions in the second year, the first and second components explained 51.93% and 11.79% (more than 63% in total) of the whole variance, respectively. As the polygon chart based on the simultaneous analysis of data under normal and drought stress in the first year revealed, shares of the first and second components in explaining the whole variance were 41.98% and 12.74% (total of 54.72%), respectively. Treatments D19, D22, and N22 had the best performance concerning all traits (Fig. 2f). Findings of the second year of the study revealed that the first and second components explained more than 41% and more than 12% (total amount of 54%) of the whole variance, respectively. Treatments D24, D14, D13, N22, N23, and D27 were more desirable than other treatments (Fig. 2g). Evaluation of the polygonal chart of the experimental treatments based on the studied traits. (a) normal irrigation conditions in the first year of the experiment; (b) stress conditions in the first year of the experiment; (c) normal irrigation conditions in the second year of the experiment; (d) stress conditions in the second year of the experiment; (e) evaluating the treatments in simultaneous conditions of normal and drought stress in the first year of the experiment; (f) evaluating the treatments in simultaneous conditions of normal and drought stress in the second year of the experiment. Evaluation of treatments in terms of yield and trait stability In assessing the treatments regarding stability and yield in response to the traits in the first year under normal irrigation conditions, treatments 3, 9, 8, 2, and 4 were selected as high-yield treatments, whereas treatments 23, 12, 16, and 27 were considered as low-yield ones regarding the studied traits (Fig. 3a). Regarding the data obtained from the second year, treatments 5, 27, 10, 6, 7, 15, and 26 were identified as high-yield treatments, and treatments 5, 10, 6, 7, 9, 13, 8, and 11 were deemed stable (Fig. 3b). In the evaluation of the data related to the first year of the experiment under drought stress conditions, treatments 22, 13, 21, and 5 were identified as desirable regarding the trait yield efficiency, while treatments 19 and 9 were selected as undesirable treatments (Fig. 3c). According to the evaluation chart for treatment stability and yield in the second year, treatments 27, 10, 11, and 18 were desirable ones with regard to the traits; however, treatments 13, 14, and 1 were undesirable (Fig. 3d). Results obtained from the interaction effects under normal and stress conditions in the first year revealed that all treatments except N3, N6, N8, and D19 were capable of having high yields (Fig. 3e). Studying the interaction effects related to the second year indicated that concerning the traits’ yields, the desirability of D27 outweighed that of other treatments, while D13 was identified as undesirable (Fig. 3f). The evaluation of treatment stability and yield regarding all studied traits in the experimental years. (a) normal irrigation conditions in the first year of the experiment; (b) normal irrigation conditions in the second year of the experiment; (c) drought stress conditions in the first year of the experiment; (d) drought stress conditions in the second year of the experiment; (e) simultaneous evaluation of normal and drought stress conditions in the first year of the experiment; (f) simultaneous evaluation of normal and drought stress conditions in the second year of the experiment. Principal component analysis (PCA) PCA was performed to evaluate the positive effects of the treatments and traits. As the eigenvalues graph based on the data derived from the first year under normal irrigation conditions presented, more than 88% of the total variance was explained by the first five components (Fig. 4a). The value of eigenvector in PCA. (a) normal irrigation conditions in the first year of the experiment; (b) normal irrigation conditions in the second year of the experiment; (c) drought stress conditions in the first year of the experiment; (d) drought stress conditions in the second year of the experiment. Analysis of the trait dispersion plot based on the first two components, which explained more than 39% of the total variance, showed that grain yield, number of fertile spikes, plant height, flag leaf width, and biological yield had positive coefficients for the two components (Fig. 5a). Moreover, the highest positive effects in the first two components were related to plant height, spike length, biological yield, and harvest index. This measure in the second component was recorded for thousand grain weight and number of fertile spikes. The largest share of positive effects in the third component was observed for grain yield, while that in the fourth component was related to peduncle length and number of fertile spikes. This criterion for the fifth component was seen for flag leaf length and flag leaf width (Table 6). Furthermore, treatments 15, 25, 27, and 24 were selected as the ones with positive coefficients of the first and second components based on the treatment dispersion graph (Fig. 6a). Treatments 2, 3, 21, 24, 25, 26, and 27 had the highest positive effects on the first component, while the highest positive effects on the second component were related to treatments 4, 11, 12, 15, and 19. This measure for the third component was recorded for the treatments 14 and 22. Treatments 1, 6, 9, and 23 had the most share of the positive effects on the fourth component, whereas treatments 5, 10, 13, 17, and 18 indicated the highest positive effects on the fifth component (Table 7). Dispersion of the traits tested in the present study based on the first two main components. (a) normal irrigation conditions in the first year of the experiment; (b) normal irrigation conditions in the second year of the experiment; (c) drought stress conditions in the first year of the experiment; (d) drought stress conditions in the second year of the experiment. (PLH: Plant Height, PL: Peduncle Length, FLL: Flag Leaf Length, FLW: Flag Leaf Width, SL: Spike Length, TGW: Thousand Grain Weight, GY: Grain Yield, BY: Biological Yield, NSS: Number of Seeds per Spike, NFS: Number Of Fertile Spike, HI: Harvest Index). Table 6Evaluating the positive effects of the traits on the obtained components in PCA based on the data obtained from normal irrigation conditions in the first year of the experiment.PC1PC2PC3PC4PC5PC6PC7PC8PC9PC10PC11PLH 0.257 0.2060.0050.0180.1400.1460.0420.0780.1010.0070.000PL0.0050.0450.055 0.549 0.0150.0210.2460.0470.0090.0080.000FLL0.0360.1870.1830.053 0.338 0.0900.0000.0680.0140.0300.000FLW0.0310.0030.3300.007 0.341 0.1730.0190.0410.0500.0040.000SL 0.509 0.0300.0230.0630.2100.0020.0000.0860.0000.0770.000TGW0.010 0.326 0.1360.1220.0000.1960.1390.0020.0640.0050.000GY0.0270.329 0.463 0.0040.0120.0100.0820.0060.0610.0010.003BY 0.855 0.0100.0210.0160.0050.0000.0010.0140.0470.0250.005NSS0.0650.0090.209 0.384 0.0390.0670.1420.0120.0630.0100.000NFS0.052 0.441 0.0450.1040.0000.0580.0000.2530.0350.0130.000HI 0.708 0.1500.0400.0030.0010.0010.0440.0290.0000.0170.005(PLH: Plant Height, PL: Peduncle Length, FLL: Flag Leaf Length, FLW: Flag Leaf Width, SL: Spike Length, TGW: Thousand Grain Weight, GY: Grain Yield, BY: Biological Yield, NSS: Number of Seeds per Spike, NFS: Number Of Fertile Spike, HI: Harvest Index). Evaluating the positive effects of the traits on the obtained components in PCA based on the data obtained from normal irrigation conditions in the first year of the experiment. (PLH: Plant Height, PL: Peduncle Length, FLL: Flag Leaf Length, FLW: Flag Leaf Width, SL: Spike Length, TGW: Thousand Grain Weight, GY: Grain Yield, BY: Biological Yield, NSS: Number of Seeds per Spike, NFS: Number Of Fertile Spike, HI: Harvest Index). Dispersion of the studied treatments based on the first and second main components of PCA. (a) normal irrigation conditions in the first year of the experiment; (b) normal irrigation conditions in the second year of the experiment; (c) drought stress conditions in the first year of the experiment; (d) drought stress conditions in the second year of the experiment. Table 7Evaluating the positive effects of the treatments on the obtained components in PCA based on the data obtained from normal irrigation conditions in the first year of the experiment.PC1PC2PC3PC4PC5PC6PC7PC8PC9PC10PC11tr10.2130.1200.030 0.469 0.0070.0380.1010.0200.0000.0000.001tr2 0.511 0.0610.1670.1030.0040.1110.0220.0150.0040.0000.002tr3 0.332 0.1150.2950.0140.0010.0180.1030.0010.0700.0480.002tr40.117 0.313 0.1410.1060.0490.0030.0040.0370.2090.0210.000tr50.0420.0900.0030.079 0.464 0.0920.2090.0030.0170.0000.001tr60.0850.0240.012 0.549 0.2530.0010.0590.0010.0100.0060.000tr70.0320.1330.1640.1300.001 0.301 0.0160.0060.1210.0930.001tr8 0.566 0.1660.0000.0180.0360.0000.0040.0250.0460.1390.000tr90.0470.0700.006 0.395 0.1810.0660.0030.0750.1520.0030.002tr100.0210.0870.0840.201 0.374 0.0700.0840.0750.0000.0030.002tr110.258 0.397 0.0670.0000.0060.0000.1820.0650.0090.0140.002tr120.000 0.547 0.2440.0430.0000.0240.0460.0730.0070.0140.000tr130.0960.0020.0290.260 0.520 0.0090.0580.0120.0120.0020.000tr140.0000.031 0.606 0.0310.0240.0410.0210.2070.0000.0370.003tr150.029 0.404 0.2960.0190.0010.1800.0110.0200.0200.0150.003tr160.0380.0890.1290.0550.103 0.280 0.1460.0000.1570.0010.001tr170.0620.0180.0330.283 0.365 0.1770.0020.0550.0010.0040.000tr180.0370.1080.1130.004 0.498 0.0360.0800.0000.1020.0230.000tr190.024 0.654 0.0370.0190.1050.0060.0000.1480.0010.0050.001tr200.2100.0020.0200.0700.080 0.214 0.1890.1430.0690.0040.000tr21 0.390 0.2570.0960.0130.0190.1360.0070.0010.0650.0150.000tr220.1450.010 0.374 0.1170.0660.0460.0960.1290.0150.0000.001tr230.0330.1820.119 0.190 0.0910.0060.1830.0030.1800.0090.003tr24 0.758 0.1040.0360.0120.0010.0110.0070.0690.0000.0030.000tr25 0.372 0.1730.0910.0440.1050.0190.0840.0770.0220.0130.000tr26 0.589 0.2190.0030.0050.0010.0080.0160.1030.0530.0000.003tr27 0.655 0.0980.0230.0030.0130.0100.0000.1190.0030.0760.001 Evaluating the positive effects of the treatments on the obtained components in PCA based on the data obtained from normal irrigation conditions in the first year of the experiment. Based on the eigenvalues graph plotted for the evaluation of the data obtained from the second year of the experiment, the first five factors explained more than 79% of the variance (Fig. 4b). The trait dispersion based on the first two components revealed that plant height, spike length, grain yield, peduncle length, number of grains per spike, and thousand grain weight owned positive coefficients of the first and second components (Fig. 5b). The highest positive effects were the effect of peduncle length, flag leaf length, thousand grain weight, grain yield, number of fertile spikes, and harvest index on the first component, the effect of flag leaf width, and biological yield on the second component, the effect of spike length on the third component, that of plant height on the fourth component, and the effect of number of fertile spikes on the fifth component (Table 8). As the assessment of trait dispersion indicated, treatments 27, 17, 6, 29, and 16 had positive coefficients with respect to both components (Fig. 6b). The highest positive effects were the effects of treatments 1, 7, 8, 12, 20, 22, 24, 25, and 26 on the first component, that of treatments 4, 21, and 27 on the second component, and the effect of treatments 2, 3, 13, 14, and 16 on the third component. In addition, the effect of treatments 5, 6, 17, 18, and 19 on the fourth component and that of treatments 10, 11, and 23 on the fifth component had the highest positive effects (Table 9). Table 8Evaluating the positive effects of the traits on the obtained components in PCA based on the data obtained from normal irrigation conditions in the second year of the experiment.PC1PC2PC3PC4PC5PC6PC7PC8PC9PC10PC11PLH0.2160.1870.0120.1970.018 0.307 0.0050.0000.0320.0260.000PL 0.443 0.0000.3100.0630.0080.0010.0500.0070.1010.0180.000FLL0.0080.1320.0080.371 0.381 0.0130.0280.0330.0250.0000.000FLW0.0690.0050.1560.312 0.399 0.0010.0020.0080.0440.0030.000SL0.0970.211 0.247 0.0000.0300.1500.1800.0720.0110.0020.000TGW 0.460 0.1060.1810.0060.0280.0550.0320.0760.0150.0420.000GY0.139 0.659 0.1710.0020.0110.0000.0030.0040.0020.0090.000BY0.237 0.689 0.0180.0090.0160.0050.0140.0030.0090.0000.000NSS 0.284 0.1480.0010.0940.0000.2460.1570.0580.0000.0120.000NFS 0.509 0.0040.0120.0990.0390.0090.2120.1010.0020.0140.000HI0.1030.020 0.563 0.1730.0060.0450.0300.0000.0170.0420.000(PLH: Plant Height, PL: Peduncle Length, FLL: Flag Leaf Length, FLW: Flag Leaf Width, SL: Spike Length, TGW: Thousand Grain Weight, GY: Grain Yield, BY: Biological Yield, NSS: Number of Seeds per Spike, NFS: Number Of Fertile Spike, HI: Harvest Index). Evaluating the positive effects of the traits on the obtained components in PCA based on the data obtained from normal irrigation conditions in the second year of the experiment. (PLH: Plant Height, PL: Peduncle Length, FLL: Flag Leaf Length, FLW: Flag Leaf Width, SL: Spike Length, TGW: Thousand Grain Weight, GY: Grain Yield, BY: Biological Yield, NSS: Number of Seeds per Spike, NFS: Number Of Fertile Spike, HI: Harvest Index). Table 9Evaluating the positive effects of the treatments on the obtained components in PCA based on the data obtained from normal irrigation conditions in the second year of the experiment.PC1PC2PC3PC4PC5PC6PC7PC8PC9PC10PC11tr1 0.342 0.1240.0450.1060.1730.0020.0350.1640.0040.0070.000tr20.1510.005 0.420 0.0080.1780.1460.0050.0530.0010.0330.000tr30.1070.145 0.285 0.1880.1010.0030.0240.0280.0250.0940.000tr40.199 0.320 0.0300.2280.1390.0480.0300.0030.0030.0010.000tr50.1350.0170.081 0.465 0.0500.0520.0760.1230.0000.0010.001tr60.0000.1110.201 0.490 0.0380.0000.0030.0900.0070.0600.000tr7 0.320 0.0960.0210.0820.1130.2140.0460.0400.0050.0630.000tr8 0.313 0.1920.2420.0640.0920.0170.0190.0490.0050.0070.000tr90.2020.0760.2410.0410.0310.006 0.326 0.0580.0170.0030.000tr100.0110.0790.1080.060 0.279 0.2490.0380.1600.0080.0090.000tr110.0000.0540.0180.257 0.428 0.0860.0060.1490.0000.0000.000tr12 0.265 0.2410.0550.0410.1620.0290.0020.0010.2010.0010.000tr130.1140.000 0.685 0.0090.0000.1510.0160.0000.0000.0250.000tr140.0000.002 0.530 0.0260.0830.1440.1050.0100.0520.0480.000tr150.1730.1250.0100.0250.0060.0070.0130.167 0.329 0.1450.000tr160.0070.054 0.559 0.0020.1840.0020.0000.0650.1270.0000.000tr170.1120.2490.031 0.279 0.2380.0170.0480.0080.0160.0010.000tr180.0840.2780.049 0.303 0.0710.1280.0380.0440.0050.0000.000tr190.2730.0220.004 0.337 0.0000.2240.1060.0300.0030.0010.000tr20 0.364 0.1650.1270.0860.0190.0100.0620.1260.0080.0330.000tr210.159 0.650 0.0120.0310.0030.0180.0030.0590.0090.0560.000tr22 0.357 0.3330.1380.0150.1280.0010.0040.0020.0130.0080.001tr230.0800.0020.0210.014 0.594 0.0000.1740.0050.0530.0560.000tr24 0.668 0.0420.1460.0130.0250.0730.0200.0020.0090.0020.000tr25 0.682 0.0150.0070.0500.0800.0650.0440.0000.0540.0030.000tr26 0.451 0.0340.0380.1620.0110.0190.2250.0070.0000.0520.002tr270.274 0.535 0.0250.0230.0140.0290.0510.0090.0390.0010.000 Evaluating the positive effects of the treatments on the obtained components in PCA based on the data obtained from normal irrigation conditions in the second year of the experiment. According to the findings of both experimental years under normal conditions, it is reported that grain yield and plant height had positive coefficients for the first two components in both years. Furthermore, plant height and spike length had positive effects on the first component in both years. The treatment evaluation showed that treatment 27 was identified as the treatment with positive coefficients regarding the first two components in both years. In addition, the highest positive effects were the effects of treatments 2, 25, and 26 on the first component, the effect of treatment 4 on the second component, and the effects of treatments 14, 6, and 10 on the third, fourth, and fifth components, respectively. According to the findings of PCA on data obtained from the first year under drought stress conditions, the first four components explained more than 70% of the whole variance based on the eigenvalues and the eigenvectors (Fig. 4c). Trait dispersion based on the first two components indicated that grain yield, biological yield, flag leaf width, harvest index, and peduncle length had positive coefficients for both components (Fig. 5c). Moreover, the greatest share of the positive effects included those of peduncle length, thousand grain weight, number of grains per spike, and number of fertile spikes on the first component, the effects of grain yield and biological yield on the second component, those of spike length and harvest index on the third component, and finally, the effects of flag leaf length and flag leaf width on the fourth component (Table 10). As the trait dispersion graph presented, treatments 22, 16, 15, 24, 13, and 25 had positive coefficients for both components (Fig. 6c). In addition, the highest positive effects were those of treatments 1, 4, 7, 9, 20, 21, and 22 on the first component, the effects of treatments 2, 6, 8, 10, 12, 14, and 23 on the second component, those of treatments 11, 13, 24, 26, and 27 on the third component, and finally, the effects of treatments 15, 16 and 19 on the fourth component (Table 11). Table 10Evaluating the positive effects of the traits on the obtained components in PCA based on the data obtained from drought stress conditions in the first year of the experiment.PC1PC2PC3PC4PC5PC6PC7PC8PC9PC10PC11PLH0.1430.0890.063 0.414 0.0450.0120.1940.0000.0290.0110.000PL 0.362 0.1340.0090.0130.1420.1590.0380.1000.0420.0000.000FLL 0.330 0.0140.1610.0840.0310.2170.0020.1430.0170.0010.000FLW0.052 0.302 0.0080.2950.2010.0400.0110.0000.0900.0020.000SL0.0080.000 0.617 0.0580.0280.1230.1080.0040.0490.0040.000TGW 0.693 0.0010.0750.0330.0010.0190.0550.0000.0190.1040.000GY 0.321 0.2470.1500.1050.0890.0160.0290.0180.0100.0120.001BY0.055 0.726 0.0200.1280.0530.0050.0010.0090.0000.0010.001NSS 0.415 0.0350.1670.0330.0930.0700.0110.1210.0230.0340.000NFS0.0040.0210.3200.085 0.392 0.0010.1480.0110.0010.0180.000HI 0.594 0.2170.0480.0050.0050.0460.0070.0450.0130.0180.001(PLH: Plant Height, PL: Peduncle Length, FLL: Flag Leaf Length, FLW: Flag Leaf Width, SL: Spike Length, TGW: Thousand Grain Weight, GY: Grain Yield, BY: Biological Yield, NSS: Number of Seeds per Spike, NFS: Number Of Fertile Spike, HI: Harvest Index). Evaluating the positive effects of the traits on the obtained components in PCA based on the data obtained from drought stress conditions in the first year of the experiment. (PLH: Plant Height, PL: Peduncle Length, FLL: Flag Leaf Length, FLW: Flag Leaf Width, SL: Spike Length, TGW: Thousand Grain Weight, GY: Grain Yield, BY: Biological Yield, NSS: Number of Seeds per Spike, NFS: Number Of Fertile Spike, HI: Harvest Index). Table 11Evaluating the positive effects of the treatments on the obtained components in PCA based on the data obtained from drought stress conditions in the first year of the experiment.PC1PC2PC3PC4PC5PC6PC7PC8PC9PC10PC11tr1 0.682 0.0150.1280.0300.0690.0260.0040.0410.0010.0050.000tr20.057 0.626 0.0000.0170.0170.1000.0040.0450.0940.0410.000tr30.2580.0840.0400.011 0.281 0.1060.0350.0130.1140.0580.000tr4 0.432 0.1540.0180.1070.0980.0010.0480.1170.0100.0140.000tr50.3940.0860.0010.0070.0150.017 0.430 0.0010.0480.0000.000tr60.043 0.803 0.0960.0000.0350.0030.0130.0040.0020.0010.000tr7 0.626 0.0350.0010.0540.0130.1430.0330.0270.0610.0060.000tr80.325 0.426 0.0020.0370.0020.0340.0350.0480.0400.0520.000tr9 0.512 0.0000.0700.0290.0110.0290.1880.0760.0000.0860.000tr100.131 0.311 0.0960.0770.0660.1520.0190.1450.0010.0020.000tr110.0130.102 0.391 0.1080.0380.1700.0290.0070.0610.0810.000tr120.011 0.628 0.1810.0980.0190.0030.0480.0110.0000.0000.000tr130.1060.003 0.593 0.1100.0340.0960.0460.0070.0030.0020.000tr140.087 0.501 0.0770.0370.0030.1610.0100.0730.0410.0110.000tr150.2110.1760.008 0.256 0.0230.0400.0030.0390.1750.0700.000tr160.3030.0250.009 0.424 0.0250.0750.1180.0160.0000.0040.000tr170.2070.0550.0520.0110.0640.238 0.268 0.0940.0080.0020.000tr180.002 0.352 0.0710.0120.0410.1330.1840.0160.1250.0630.000tr190.0100.1940.001 0.390 0.3420.0090.0420.0060.0050.0000.000tr20 0.388 0.1030.0440.0540.1620.0150.0040.2020.0220.0060.000tr21 0.770 0.0000.0050.0180.1160.0010.0530.0130.0060.0180.000tr22 0.524 0.0910.2960.0150.0000.0000.0510.0080.0150.0000.000tr230.002 0.521 0.3300.0400.0010.0590.0010.0000.0400.0050.000tr240.1560.007 0.367 0.2280.1940.0020.0380.0010.0050.0020.000tr250.0290.0020.2950.1250.051 0.372 0.1170.0050.0010.0040.000tr260.0060.262 0.464 0.0200.1150.0650.0040.0110.0050.0480.000tr270.1390.008 0.308 0.2660.0000.0920.1020.0460.0360.0030.000 Evaluating the positive effects of the treatments on the obtained components in PCA based on the data obtained from drought stress conditions in the first year of the experiment. Analyzing the eigenvectors for the data derived from the second year indicated that the first four components explained more than 66% of the total variance (Fig. 4d). According to the trait dispersion graph based on the first two main components, grain yield had positive coefficients (Fig. 5d). The highest positive effects were recorded for the effects of grain yield, biological yield and harvest index on the first component, those of spike length, thousand grain weight and number of fertile spikes on the second component, the effects of plant height and flag leaf length on the third component, and those of peduncle length and flag leaf width on the fourth component (Table 12). Treatments 7, 8, 9, 10, 13, 14, 15, and 16 had positive coefficients for the first two components according to the evaluation of trait dispersion (Fig. 6d). Furthermore, the greatest share of the positive effects belonged to the effects of treatments 9,10,12, 13, 18, 19, 23, and 25 on the first component, those of treatments 3, 5, 17, and 24 on the second components, the effects of treatments 1, 4, and 8 on the third components, and finally, those of treatments 6, 20, 22, and 26 on the fourth component (Table 13). Table 12Evaluating the positive effects of the traits on the obtained components in PCA based on the data obtained from drought stress conditions in the second year of the experiment.PC1PC2PC3PC4PC5PC6PC7PC8PC9PC10PC11PLH0.0320.019 0.370 0.2210.0990.0360.0040.2050.0120.0020.000PL0.0820.0010.117 0.321 0.0090.1910.2420.0040.0330.0010.000FLL0.0500.085 0.530 0.0200.0200.0260.0680.1630.0230.0140.000FLW0.0820.0190.040 0.540 0.0230.1760.0030.0270.0880.0010.000SL0.007 0.522 0.0030.0320.2800.0610.0160.0030.0590.0160.001TGW0.040 0.412 0.0000.0640.0130.2570.1540.0190.0350.0060.000GY 0.861 0.0000.0090.0550.0020.0040.0140.0000.0270.0060.021BY 0.836 0.0000.0690.0270.0040.0000.0060.0000.0000.0380.018NSS0.0930.2450.0660.000 0.415 0.0690.0000.0800.0280.0020.000NFS0.037 0.354 0.2610.0410.0110.0100.2180.0060.0310.0310.001HI 0.527 0.0170.1400.1000.0520.0210.0090.0340.0110.0880.001 Evaluating the positive effects of the traits on the obtained components in PCA based on the data obtained from drought stress conditions in the second year of the experiment. Table 13Evaluating the positive effects of the treatments on the obtained components in PCA based on the data obtained from drought stress conditions in the second year of the experiment.PC1PC2PC3PC4PC5PC6PC7PC8PC9PC10PC11tr10.0750.121 0.638 0.0150.0190.0180.0350.0690.0070.0020.000tr20.2440.0100.0030.113 0.326 0.0000.2710.0200.0060.0040.001tr30.090 0.459 0.0000.0000.0630.0770.0870.0820.0800.0610.000tr40.0200.012 0.483 0.0280.0020.0110.1260.2500.0450.0170.004tr50.0310.1420.0000.209 0.264 0.1130.1180.0180.0040.0980.001tr60.2650.0300.026 0.439 0.0180.1310.0640.0190.0010.0040.004tr70.0690.0400.0000.2030.1610.221 0.228 0.0620.0020.0150.000tr80.0000.201 0.344 0.0070.0810.0180.2820.0450.0070.0130.003tr9 0.364 0.0590.0200.2210.0320.2130.0010.0160.0010.0710.003tr10 0.306 0.1640.0060.1840.0800.0050.0900.0050.1250.0330.001tr110.0750.1450.1230.0260.047 0.381 0.0980.0100.0820.0090.004tr12 0.448 0.0190.0320.0220.0000.3260.0340.0870.0140.0000.018tr13 0.367 0.0380.0090.1490.0940.1030.1130.0250.0180.0780.005tr140.0710.0100.0620.1480.0390.0850.077 0.274 0.1970.0020.035tr150.081 0.301 0.0870.1210.0860.0000.0820.0090.1380.0780.018tr160.0010.1170.0050.2110.2150.1250.0050.012 0.269 0.0390.001tr170.127 0.491 0.0020.0640.0100.1830.1040.0050.0140.0010.000tr18 0.269 0.0940.1890.0100.1520.2550.0030.0000.0280.0010.000tr19 0.766 0.0170.0770.0700.0100.0150.0100.0110.0150.0080.002tr200.0370.0540.301 0.409 0.0910.0000.0010.0880.0040.0150.000tr210.1200.1030.0040.223 0.289 0.0000.0000.1390.0600.0250.037tr220.2100.0160.006 0.353 0.2310.0430.0890.0080.0070.0110.025tr23 0.282 0.0590.2250.2300.0960.0190.0190.0220.0330.0040.012tr240.001 0.565 0.0180.0300.2770.0020.0380.0610.0040.0000.005tr25 0.621 0.0010.0020.2290.0190.0260.0030.0040.0010.0930.001tr260.0300.1300.087 0.330 0.1440.0010.0100.1160.1130.0380.002tr270.0500.0200.0330.2240.0460.017 0.331 0.2270.0420.0050.005 Evaluating the positive effects of the treatments on the obtained components in PCA based on the data obtained from drought stress conditions in the second year of the experiment. Network plot of treatments on traits Figure 7 presents the Network Plot for the means of the data obtained from both cropping years under normal and drought stress conditions (Fig. 7). Based on this diagram, the similarity among treatments in terms of their responses to traits can be assessed by determining the correlation coefficients between traits. In this diagram, a higher Edge Cutoff value indicates stronger connections, whereas a lower Edge Cutoff includes weaker correlations, resulting in a more complex network. Results related to the normal conditions with 50% Edge Cutoff (Fig. 7a) indicated the main similarities regarding the traits. The strongest inter-treatment relationship was recorded for tr15, tr17, tr18, tr19, tr7, and tr20. Treatment 22 showed no correlation with other treatments. At 60% Edge Cutoff (Fig. 7b), tr18 had the highest relationship with other treatments, while at 70% Edge Cutoff (Fig. 7c), tr18, tr19, tr20, tr23, tr27, tr26, and tr5 indicated the highest relationships. As a result, under normal conditions, treatments like tr18, tr19, and tr20 hold greater importance. On the other hand, under stress conditions, tr14, tr15, tr18, tr19, tr20, tr21, and tr27 showed the highest relationships with other treatments at 50% Edge Cutoff (Fig. 7d). Treatments tr2 and tr3 lacked any correlation with other treatments in respect of the studied traits. At 60% Edge Cutoff (Fig. 7e), relationships of tr15, tr19, tr20, and tr21 exceeded those of other treatments, and at 70% Edge Cutoff (Fig. 7f), tr5, tr15, tr19, tr20, and tr21 indicated more relationships compared to the other treatments. Network Plot of the tested treatments based on the studied traits with various Edge Cutoffs under normal and drought stress conditions. Normal conditions: a: 50%, b: 60%, c: 70%; drought stress conditions: d: 50%, e: 60%, f: 70%. Discussion The polyhedral diagram is used to identify the best treatments among the studied traits. This diagram is constructed by connecting the treatments that are farthest from the origin. Consequently, other treatments are positioned inside this polyhedron. Within each section, treatments exhibiting higher performance and desirability for specific traits are delineated by lines (Fig. 2). The polygon diagram has been used by various researchers on different plants, including wheat, maize30–32, and sunflower33. Based on the graph obtained from normal conditions in the first year of the experiment (Fig. 2a), treatments were selected that primarily had moderate levels of amino acid and fulvic acid. Performance of treatments 7 and 10 outweighed that of other treatments in the cases of grain yield, biological yield, and peduncle length. Furthermore, treatments 24 and 21 performed significantly better than others in terms of spike length and flag leaf width, respectively. Finally, treatments 3, 4, and 9 exceeded others in respect of flag leaf length. Under stress conditions, the selected treatments had high concentrations of humic acid and fulvic acid. Number of grains per spike, peduncle length, biological yield, grain yield, and harvest index were the highest in treatments 9, 6, 7, 3, 10, 4, and 15. Rest of the treatments except tr19 performed better compared to others with regard to all traits except flag leaf length, for which treatment 19 outranked others (Fig. 2b). In the second year of the experiment under normal conditions, the selected treatments contained moderate concentrations of organic compounds. Number of grains per spike, peduncle length, biological yield, grain yield, and harvest index showed the highest records for treatments 9, 6, 7, 3, 10, 4, and 15. Other treatments except tr19 presented better outcomes in terms of traits except flag leaf length, on which treatment 19 had the best effect Fig. 2c. Under stress conditions, the selected treatments had the highest concentrations of humic acid and amino acid among the organic elements used. Treatments 27, 1, 6, 9, and 10 had the best effects on biological yield, grain yield, peduncle length, and plant height. Number of grains per spike showed the highest records for tr19 and tr16, while the highest spike length data were recorded for tr22 (Fig. 2d). In the combined study of normal and stress conditions in the first year of the experiment, the positive effect of fulvic acid in selecting the optimal and desirable treatments was evident. Considering the accumulation of the traits and treatments inside this polygon chart, the balanced trend with regard to traits’ reaction to the treatments was reported. Treatment D19 indicated the highest yield efficiency in respect of flag leaf length compared to other treatments (Fig. 2e). Also, in the second year of the experiment, the high concentration of organic compounds, especially fulvic acid, in the desirable treatments indicated the positive effect of these compounds in controlling humdity stress and improving performance in the traits. Moreover, treatment D18 presented better effects on biological yield and grain yield compared to other treatments, and the performance of D27 outweighed that of other treatments regarding flag leaf length. Since the remaining treatments were observed inside the polygon chart, they had a balanced trend along with other traits in terms of yield efficiency (Fig. 2f). Based on the findings of these charts in different years and normal irrigation conditions, tr22 and tr23 could be selected as suitable for cultivation in the first stage. In the second stage, treatments 5, 8, 13, 14, and 18 could be recommended as desirable. According to the results, applying high rates of fulvic acid as well as an average dosage of the amino acid can have positive effects on controlling the stress in plants. In addition, increasing the concentration of humic acid can compensate for low concentrations of compounds like fulvic acid and amino acids. In the stability diagram (Fig. 3) for the first year of the experiment, treatments with high humic acid concentration and no fulvic acid were selected as stable, while treatments with high fulvic acid concentration showed lower stability under normal conditions. (Fig. 3a). Furthermore, treatments 3, 9, 1, 2, 11, 5, 20, and 6 can be selected as treatments with high stability potential against various traits. Generally, according to the findings of the first year, treatments 3 and 9 were selected as the superior ones. In the second year of the experiment, treatments with high humic acid levels and moderate fulvic acid concentrations positively influenced plant vigor in the desirable treatments. Generally, in the second year of the experiment under normal conditions, treatments 5, 10, 6, and 7 were recognized as treatments with high yield potential and a suitable level of stability in respect of all traits (Fig. 3b). In examining the effect of drought stress during the first year of the experiment, the positive impact of increasing fulvic acid concentration in the treatments was evident in identifying the desirable treatments. Treatments 22, 13, 14, 20, 23, 2, 3, 17, 18, 21, and 24 were selected as stable. Treatments 22, 13, and 21 were identified as high-yield and stable in response to the traits (Fig. 3c). Under drought stress conditions, increasing all treatment levels in the experiment had a positive effect on increasing the stability of the treatments. Moreover, treatments 11, 22, 15, 17, 4, 2, 3, and 27 could be identified as stable treatments, and treatments 27 and 11 could be selected as the ones with high yield and stability in response to the traits (Fig. 3d). In examining the effect of drought stress during the first and second years of the experiment (Fig. 3e and d), the positive impact of increasing fulvic acid concentration in the treatments was evident in identifying the desirable treatments. Regarding stability, all treatments except D6, D8, N22, D22, D11, D15, and N15 had satisfying levels of stability based on their proximity to the mean axis in the chart (Fig. 3e). In the second year, treatments except N18, D14, N16, D21, N24, N23, and D18 had acceptable levels of stability. Generally, D27 could be selected as the desirable treatment (Fig. 3f). According to the plotted charts for different conditions in both experimental years, treatments 3 and 6 under normal irrigation conditions and treatment 27 under drought stress conditions were identified as desirable treatments. The fact that increasing organic matter, like fulvic acid and amino acids, can enhance plant resistance against drought stress through regulating the plant metabolism obviously shows the effect of compounds in this treatment on plant resistance. A study conducted to assess the reduction of drought stress effects by humic acid reported that a dosage increase in humic acid could decline drought stress-induced oxidative effects through the induction of non-enzymatic and enzymatic antioxidants at different and significant levels34. Another study evaluating the effect of humic acid on physiological characteristics under drought stress revealed that humic acid can enhance plant growth under drought stress conditions through improving osmotic regulation, antioxidant capacity, and photosynthesis35. In a study conducted to investigate the effects of organic compounds on wheat under water-deficit conditions, these compounds played a significant role in mitigating drought-induced damage by improving morphological traits and effectively suppressing the production of reactive oxygen species (ROS)36. The improvement in water-use efficiency in response to the application of organic compounds is primarily attributed to increased photosynthesis and reduced transpiration. The use of these compounds enables wheat to utilize water more efficiently under drought stress conditions37. The principal component analysis (PCA) conducted over two experimental years under both normal and drought stress conditions provided valuable insights into the key traits influencing treatment performance and highlighted the most effective treatments. Under normal irrigation conditions, the first five principal components explained a substantial portion of variance-over 88% in the first year and more than 79% in the second year. This underscores that the measured traits collectively capture the majority of variation in treatment effects. Grain yield and plant height consistently showed positive coefficients for the first two components across both years, indicating their pivotal role in differentiating treatment performance. Traits such as spike length and biological yield also had strong positive influences on the primary components, emphasizing their importance for overall productivity and plant vigor. The evaluation of treatments based on PCA revealed that certain treatments, particularly treatment 27, consistently exhibited positive coefficients across the first two components in both years under normal conditions. This suggests their robustness and favorable performance for essential agronomic traits. Moreover, treatments 2, 25, and 26 demonstrated strong positive effects on the first component, while treatment 4 showed the highest effect on the second component, highlighting their specific contributions to certain trait variations. Under drought stress conditions, the PCA explained over 70% and 66% of total variance in the first year and second year, respectively. Grain yield, biological yield, flag leaf width, harvest index, and peduncle length were among the traits with positive loadings on the main components, reflecting their critical roles in stress adaptation and yield stability. Traits such as peduncle length and thousand grain weight also had notable importance, signifying their relevance in sustaining productivity under drought. Treatments 15, 16, and 13 emerged with positive coefficients for the primary components in both years under stress, suggesting their potential as stress-resilient candidates. Treatment 9 showed the highest positive effect on the first component across both years, reinforcing its suitability for humidity-deficient conditions. The consistency of grain yield’s positive loadings on the first two components across all conditions and years indicates its strong association with the overall performance and stress tolerance of the treatments. Additionally, the recurring positive impact of flag leaf width on the fourth principal component points to its possible role in compensating for stress effects and contributing to physiological efficiency. Results related to the drought stress conditions in both years showed that grain yield had positive coefficients for the first two main components in both years. Flag leaf width had the highest positive effect on the fourth component in both years. Treatments 15, 16, and 13 had positive coefficients for both first components, and treatment 9 had the highest positive effect on the first component in both years. Based on the network plot of treatments on traits under normal conditions, treatments like tr18, tr19, and tr20 hold greater importance. On the other hand, under stress conditions, tr14, tr15, tr18, tr19, tr20, tr21, and tr27 showed the highest relationships with other treatments at 50% Edge Cutoff and under stress conditions, tr15, tr19, tr20, and tr21 are considered more desirable with stronger relationships and correlations. Treatments tr15, tr19, and tr20 that have shown up consistently with strong relationships under both normal and stress conditions played a noteworthy role in the correlation network. Treatments like tr22 under normal conditions, as well as tr2 and tr3 under stress conditions, do not have considerable relationships with other treatments. This could be a consequence of their different and weaker performance for the studied traits. Conclusions and future prospects The findings of this two-year study under normal and drought-stress conditions highlight the pivotal role of traits such as grain yield, plant height, panicle length, biological yield, harvest index, and flag leaf width in determining treatment performance. Among the evaluated treatments, numbers 27, 15, 19, and 20 demonstrated superior yield and greater stability, making them promising candidates for cultivation under diverse environmental conditions. Moreover, folic acid and humic acid significantly contributed to mitigating drought stress and improving yield performance, emphasizing their importance in enhancing stress tolerance. Network analysis further indicated that treatments with stable trait associations across environments represent the most reliable options for sustainable crop management under variable conditions. This research demonstrates that organic compounds act as physiological modulators, reinforcing and harmonizing the interrelationships among traits under stress, rather than merely inducing an isolated increase in individual traits. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Acknowledgements Many thanks to colleagues at the Faculty of Agriculture, University of Mohaghegh Ardabili for providing facilities and laboratory and assistance in publishing this manuscript. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Author contributions Conceptualization, S.H.S.; data curation, A.O; S.H.S; formal analysis, S.H.S; A.O.; software, S.H.S.; methodology, M.B; A.O.; writing—review and editing, S.I.T; S.H.S; A.O.; project administration, M.B; supervision, M.B; All authors have read and agreed to the published version of the manuscript. Data availability The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request.