Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Scientific Reports volume  12, Article number: 7359 (2022 ) Cite this article

Trigonella foenum-graecum is widely distributed worldwide and grown under a wide range of climatic conditions. The current research was conducted to study the effects of the environmental variables on yield related traits and metabolite contents of 50 different Persian fenugreeks at various geographical locations. Accordingly, multivariate statistical techniques including canonical correspondence analysis (CCA), hierarchical clustering on principal components, and partial least squares regression (PLSR) were applied to determine important proxy variables and establish a relevant model to predict bioactive compounds in fenugreeks. The interrelation of clustered groups emphasized the importance of functional groups of bioactive compounds and several yield related traits. The CCA indicated that two climatic variables of temperature and solar irradiation contributed prominently to 4-hydroxyisoleucine accumulation. The predicted model based on PLSR revealed climatic variables such as temperature, solar, and rain. The precursor of isoleucine was the predictive power for 4-hydroxyisoleucine accumulation while seed weight predicted trigonelline content. The current study's findings may provide helpful information for the breeding strategies of this multipurpose crop.

Fenugreek belongs to the Trigonella genus with around 260 species with broad geographical habitats, ranging from the Mediterranean, Asia, Australia, Europe, and Macaronesia to North and South Africa1. Thus, Fenugreek is likely to change its chemotype contents, including phenol, flavonoid, alkaloid, protein2,3,4, and its morphometric characteristics for surviving in different ecological conditions. Plant plasticity is an important feature helping the plants adapt to the surrounding situation changes, survive, and produce another generation5. It has been well-documented plants’ chemical composition, and morphometric characteristics are more influenced by the environmental conditions and the genetic variation6,7. A study showed that the geographical locations had a major role in the variations of the morphometric characteristics among different Pinus tabulaeformispopulations8. It has been reported that the chemical composition of the fenugreek genotypes is significantly affected by various environmental variables9. Abiotic stresses, such as salinity10,11,12 and drought13,14, significantly alter the accumulation of bioactive compounds in Fenugreek under controlled experimental conditions.. Evidence indicates that the content of amino acids and secondary metabolites of different Fenugreek genotypes is modified in different geographical locations like Africa15, and Asia15. Another report also showed that genotypic variations and environmental conditions mainly control the content of bioactive compounds16. Furthermore, there is evidence that different chemical components' abundance is predicted at various environmental conditions in a wild population of Pilocarpus pennatifolius7. Fenugreek seeds are the most important and well-studied part of the fenugreek plant17. However, to date, no study indicates to what extent different environmental variables alter the content of the bioactive compounds with their precursors and yield related traits in the fenugreek seeds in their natural habitats. To this end, the seeds of T. foenum-graecum L. were collected from 50 different geographical locations in Iran and their yield related traits were measured using a standard descriptor. Then, the primary metabolites and bioactive compounds of the seeds were measured by the High-performance Liquid Chromatography (HPLC) and Ultraviolet (UV)/Visible (VIS) spectrometry. A series of multivariate statistical analyses were applied to the obtained data for addressing some important issues: (i) to estimate the variations and data grouping as well as to determine the most critical variables in each group; (ii) to understand the relationship and effect of the environmental variables on the fenugreek seeds; (iii) to define the heritability of the plant traits, and finally (iv) to determine the desired prediction model for two bioactive compounds of 4-hydroxyisoleucine and trigonelline and to measure the direct and indirect effects of the traits on these two metabolites.

The present study complies with relevant institutional, national, and international guidelines and legislation. Seeds of the Trigonella foenum-graecum L were collected with permission to collect the seeds of the plant material by the Agricultural Biotechnology Research Institute of Iran (ABRII), project No: ABRII14-05-05-9256.

The analysis of variance was conducted to determine the variation in yield related traits and metabolites of the T. foenum-graecum L populations (Table 1). The results indicated that 21 traits, such as three environmental variables, seven yield related traits, and 11 metabolites, displayed a significant difference (P < 0.001) in the fenugreek’s seeds. Among the environmental variables, the region of Mirash (50.731482°, 36.182436°) was the rainiest location with annual average precipitation of 34.28 mm, while the annual average precipitation in Hamidieh (0.13 mm; 48.516364°, 31.435083°) was almost 280 times lower than Mirash. The mean annual temperature in Hamidieh (35.82°C) was two-fold more than the coldest location in SagzAbad (18.1°C; 49.934878°, 35.766543°). Fasa (53.450804°, 29.137368°) received the greatest amount of annual solar radiation (2211 kJ/m2), while Mirash received nearly half of Fasa solar radiation (1692 kJ/m2). The bioactive compounds accumulation was significantly different among the locations. 4-hydroxyisoleucine concentration was from 99 mg/gr dry seed in Shushtar (48.832836°, 32.121287°) to 3.36 mg/gr dry seed in Mirash. Trigonelline concentration was from 0.36 mg/gr dry seed in Borujerd (48.813873̊, 33.805145°) to 0.11 mg/gr dry seed in Hamadan (48.527574°, 34.835419°). Furthermore, the maximum value of seed weight was 17.54gr for Kazerun (51.944223°, 29.568908°) in the warm zone at the west-northwest of Iran. In contrast, the minimum (9.54gr) rate of seed weight belonged to Khomein (50.317967°, 33.663119°) in the centre of Iran (Table 1).

The heritability and genetic advance were used to capture the variations explained by the influence of genetics, including the addition, dominance, and epistasis against the environment18. The Phenotypic Coefficient Variation (PCV) range was between 4.4 and 76% for phenol and aspartic acid content (Table 2). Furthermore, isoleucine content had the maximum Genotypic Coefficient Variation (GCV) (72%), while phenol showed the minimum rate (3.4%). The inheritance of the total phenol, flavonoid, antioxidant, and trigonelline was medium and equal 60%, 50%, 53%, and 58%, respectively. Furthermore, 4-hydroxyisoleucine had the minimum inheritance (12%). The Genetic Advance (GA) varied from 3.9% for 4-hydroxyisoleucine to 97% for aspartic acid (Table 2).

Hierarchical Clustering on the Principal Components (HCPC) is a convenient method for multidimensional datasets. This method reduces the data dimensionality into less and most crucial information and stable clustering. The results of the Principal Component Analysis (PCA) showed that the two first dimensions explained up 33.6% of the variation among the geographical locations (Figure S1). The PCA loading plot exhibited the two regions of Minab (27.323446°, 56.542827°) and Shushtar (36.182436°, 50.731482°) contributed mostly to dimensions 1 and 2, respectively (Fig. 1). Furthermore, seed weight and seed width had the highest contribution to dimensions 1 (Fig. 1). In contrast, valine and aspartic acid mainly contributed to dimensions 2.

The Bar graph of percentages of contribution yield related triats and metabolite variables a: Minab with 19% and Sabzevar with 0.5% were the highest and lowest contribution variables to dimensions 1 (a and b), respectively. In dimensions 2 (c and d), Shushtar with 12% and Malayer with 3% were the highest and the lowest contribution variables, respectively.: seed weight (17%) and flavonoids (4%) showed the highest and the lowest percentage of contribution in dimensions 1 (a), respectively. In dimension 2 (b), valine (16%) and seed per pod (4%) had the highest and lowest percentages of contribution, respectively.

In the second step, hierarchical cluster analysis divided the regions into the three distinct clusters (v. test ≥ 2) (Table 3, Figure S2). The cluster I included 20 members, of which four important variables, including the seed per pod, seed length, aspartic acid, and pod length, passed the significant test (Table 3, Fig. 2, Excel S2, S3). Cluster II had 29 members. The genotypes of clusters I and II were mainly located in the south and the center of Iran and received almost the same average of solar radiation (1954.053 kJ/m2), temperature (25.074°C), and precipitation (9.91 mm) (Excel S2). Cluster III had 11 members and was located in the southwest and northeast of Iran, where the fenugreek’s genotypes receive more light (2047 kJ/m2) and warm (30.23 ̊C) form of the sun. The average content of metabolites including 4-hydroxy isoleucine (87.86 mg/ gr dry seed), trigonelline (0.25 mg/gr dry seed), isoleucine (1.49 mg/gr dry seed), leucine (4.72 mg/gr dry seed), and flavonoid (230. 86 mgQE/100gr) were higher genotypes located in cluster III than cluster 1 and II.

The association among measured variables and two first axis of PCA. The first and second dims explained up 21.4% and 12.9% of the total variation, respectively. The colours of vectors show the contribution rate of each variable to Dim1 and Dim2.

We then investigated to what extent the environmental variations exert their impacts on the relationships between traits when the clusters were considered as the three different populations. The correlation between traits of the clusters was investigated by the Pearson correlation coefficient method (Tables S3, S4, and S5). The bioactive compounds of trigonelline and 4-hydroxyisoleucine differed in the direction of correction between the clusters. 4-hydroxyisoleucine was positively correlated with the trigonelline in cluster II. For the traits, 4-hydroxyisoleucine vs isoleucine and lysine vs trigonelline the correlations differed in the direction in cluster I. Seed weight and trigonelline showed the same direction of strong positive correlation in cluster I and cluster III. This correlation is useful for indirect breeding and selection for improved seed yield. The pods' length and seeds number were negatively correlated with isoleucine and 4-hydroxyisoleucine. The paired of traits like phenol, antioxidant, and flavonoid showed the same direction of correlation in all three clusters.

The Canonical correspondence analysis (CCA) was carried out to identify the link between the metabolites and yield related traits variables with various environmental variables. The result displayed that CCA1 and CCA2 explained 94% of phenotypic variation (Figure S3, S4, and Tables S6, S7). Furthermore, the analysis of CCA also displayed environmental variables can explain the distribution of fenugreeks accessions by affecting metabolites and yield related trait. In CCA1, rainfall was the most significant contribution to environmental variation and positively correlated with the regions with higher average rainfall (17.81 mm) (Excel S3). Furthermore, the most coordinated traits with CCA1 were leucine, seed per pod, and pod length. These traits were positively correlated. However, the correlation between 4-hydroxyisoleucine and trigonelline in the CCA1 axis was significantly negative. In the CCA2, the correlation among temperature, solar radiation, isoleucine, 4-hydroxy isoleucine, trigonelline, phenol, flavonoid, and antioxidant were significantly positive. (Figs. S3, S4, Table S4). Likewise, the average content of isoleucine, 4-hydroxy isoleucine, trigonelline, phenol, flavonoid, and antioxidant in the CCA2 axis was higher than the total mean value (Table S8, Excel S3). Furthermore, the regions which showed a positive correlation in CCA2 had a higher average temperature and solar radiation (27.04̊ and 1980.40 kJ/m2) (Figs. S3, S4, Excel S3) compared with all regions (25.58̊ and 1957.22 kJ/m2).

Partial least squares regression (PLSR) analysis is a practical statistical analysis for linking the predictors and dependent variable19. Since trigonelline and 4-hydroxyisoleucine are the most effective pharmaceutical metabolites in fenugreek1, they were chosen as the response variables (Y value). At the same time, the yield related traits, metabolic traits, and environmental variables were considered the latent variables (X values) to find the best predictors for the trigonelline and 4-hydroxyisoleucine yield. In the first step, PLSR was performed by all the variables to determine and eliminate the useless predictors (Fig. 3). Then, the Variable Importance in the Projection (VIP) was assessed for selecting the relevant latent variables19. The threshold value for VIP was chosen greater than one (VIP > 1) (Table 4). The results revealed that only seven traits were the most important variables for predicting the 4-hydroxy isoleucine and trigonelline in the model. The seed weight showed the most significant VIP score for 4-hydroxyisoleucine and a lower value for the trigonelline. Environmental variables including the temperature, solar radiation, and precipitation also showed the maximum VIP score value for the 4-hydroxyisoleucine than the trigonelline, confirming that the environmental variables had the greatest effect on 4-hydroxyisoleucine content in the model. However, two yield related traits variables of the seed width and weight showed the highest effect on the trigonelline.

Observed and cross-validated predicted values of 4-hydroxyisoleucine (a) and trigonelline (b). Each point shows one region. The root mean square error of prediction (RMSEP) indicates the quality of the model in predicting 4-hydroxyisoleucine and trigonelline content. The regression coefficient (R2) indicates the fit between the predicted and observed 4-hydroxyisoleucine and trigonelline content.

Therefore, the structural equation models are as following for 4-hydroxy isoleucine and Trigonelline:

In the present study, the effects of the environmental variables were investigated on the yield related traits and bioactive compound accumulation of the Persian fenugreek’s seeds at 50 different geographical habitats. The correlations between the bioactive compounds and yield related traits were also identified. This association might be helpful to exploit the natural diversity and bioengineer the valuable bioactive compound and metabolic pathways in the fenugreek20.

Fenugreeks grown in higher temperatures and solar radiation regions had slightly bigger and heavier seeds (Table S8, Figs. S1 & S3) based on CCA and HCPC analyses. This may indicate that the favorable environment for the growth of the fenugreek is the dry and cold temperature at the vegetative stage. In contrast, dry and practically high temperature is favorable for seed production21. Plants that grow under high temperatures and solar radiation had higher seed yields22,23. The content of phenol and flavonoid in the rainy regions of SagzAbad and Mirash was lower than in drier regions, which is in contrast with the study by Larson result24. Warmer regions with higher solar radiation, including Haji-Abad, Kazerun, Kalateh, Fasa, and Bandarabbas, had a higher 4-hydroxyisoleucine, isoleucine, antioxidant, and trigonelline content in seeds. It is proven that phenol content increased under stress conditions such as high temperature and solar radiation to survive plant's life25,26. The accumulation of the trigonelline grown under high temperature and sunlight was greater compared to their shade-grown counterparts in the coffee trees27. Trigonelline, as an alkaloid1, accumulates by the increase in the solar radiation to protect the plant cells against Reactive oxygen species (ROS), leading to the enhanced antioxidant activity1,28,29,30. However, the occurrence of an interrelation between active metabolites and environmental variables does not imply causality. In some cases, the levels of metabolites are dramatically influenced by the stresses or different nutrient regimes31.

Heritability estimation helps to separate the effect of genetics from nature on the variation in the plants̕ traits18. Furthermore, heritability indicates the potential of transmission of the traits from one generation to the next, helping the breeders choose the appropriate selection methods to improve the desired traits. However, broad-sense heritability may not be reliable without estimating the genetic advance32. PCV and GCV give us a more distinct picture regarding the influence of the surrounding environment on the heritage of the variables. The traits with high GCV and low PCV are less influenced by the environmental condition32. Phenol and antioxidant presented moderately high broad-sense heritability as previously reported33,34,35. In general, phenolic content is mainly responsible for the response to the variation in the surrounding condition of the plants. Although the phenolic compound's inheritance is relatively high, their GA and GCV are too low because of their sensitivity to the environmental condition (Table 2). These metabolic traits, which play an important role in the plant protection and against the oxidative stress in human disease36, were related to in the correlation matrix (Tables S3, S4 , S5). Low heritability of the amino acids (Table 2) showed that they are not stable in variations of the environmental condition, as previously reported in the literature37.

The environmental variables could predict the quantity of metabolic compounds in the fenugreek. Here, metabolites were measured on samples collected from the field to assess the phenotypic plasticity of fenugreek varieties that showed different magnitudes regarding the correlation between environmental variables and the accumulation of bioactive compounds. It was found that three different patterns of relationships between the yield related traits and metabolic traits in the three clustered populations under different environmental conditions (Fig. 2, Tables S3, S4, S5). Clusters I and II, which have a close geographical location, showed similar correlation patterns (Fig. 2, Tables S3, S5). It was found that the environmental condition influences the correlations between the yield related traits and metabolic traits38. The similar correlations among compounds like phenol, flavonoid, and antioxidant in all three clusters may indicate that these compounds are more structurally and functionally related. It has been previously shown that the metabolites with a similar structure were correlated at a global level in the correlation matrix36,39,40. There were also significant associations between metabolites in two different sectors of central metabolism, including amino acids and bioactive compounds. Such correlations might be acceptable as the processes related to the biosynthesis of the bioactive compounds rely mainly on the primary metabolite41. There was also a relationship between isoleucine as the precursor and 4-hydroxyisoleucine as the end product42. A positive correlation between the 4-hydroxyisoleucine and antioxidant activity was seen, suggesting that the antioxidant activity increases upon enhancement of the 4-hydroxyisoleucine content43. There were few correlations between any of the yield related traits and metabolic traits. The lower number of such a relationship indicates that although the metabolites are the mediator between the plant's genome and the environment, they are not necessarily the outcome of this interaction. So, it is maybe the reason for the absence of some correlations between metabolites and yield related traits29. The heritability of the yield related traits was higher than the metabolic traits (Table 2), which is consistent with the results of several previous reports41,44,45,46. Multivariate data analysis using the PLSR, which is preferable for large numbers and diverse variables, revealed that 4-hydroxyisoleucine was predicted by overlapping sets of the metabolites and yield related traits, indicating their close link to the accumulation of 4-hydroxyisoleucine content in the fenugreek (Fig. 3).

The current research was examined the effects of the different environmental traits on yield related traits and bioactive compounds on a small panel of 50 different Persian fenugreeks genotypes at different geographical locations. A simple and multivariate statistical analysis showed that metabolic traits had negligible heritability than the yield related traits in the Persian fenugreek accessions. Further, it was shown that the proposed model, predicted 4-hydroxyisoleucine bioactive compound accumulation, is closely related to a set of metabolites and changes in the environmental variables, which in turn will provide helpful information for breeding strategies for fenugreek. However, it is not possible to recommend plant breeding without removing the ecological effect by cultivation under the same environmental conditions. Thus, further information is also needed to confirm that current results are either a plastic response to the environment or genetically controlled.

The Fenugreek’s seeds were collected from 50 different geographical locations in Iran (Table S1, Fig. 4). The sampling was done based on standard instruction. Briefly, the fenugreek distribution areas were identified based on the existing flora, such as Iranica, the national herbarium of the Forest and Rangeland Research Institute, and the Ferdowsi Mashhad University. This was chosen based on the difference in the altitude and latitude and the topography of the area, which ultimately causes climate change. Sampling was done randomly with constant distances from each other using the walking distance. The sample unit was landrace. 50 to 100 plants were collected at each site. The collection route was determined using maps at a scale of 1: 500,000. The collection was done at the maximum stage of seeds maturity. In the case of endemic and native crops, the required seeds were obtained directly from the farmers. The data were then entered into a collection form. 100–150 g of each landrace with the completed collection information and forms were sent to the Agricultural Biotechnology Research Institute in Karaj for storage and further evaluation and stored at −20 °C until further analysis. Yield related traits of each landrace and metabolite analysis were carried out on collected seeds at the field.

The geographical locations of collected Trigonella foenum-graecum L. seeds in Iran. The purple points represent the collected sample’s locations. The map was provided by courtesy of the Earth Science and Remote Sensing Unit, NASA Johnson Space Center (https://eol.jsc.nasa.gov) and was modified using Adobe Photoshop CC v15.2 and Microsoft PowerPoint 2016.

The climatic data, including average annual precipitation, average temperature, and average solar radiation, were obtained by the Islamic Republic of Iran Meteorological Organization (IRIMO). The average Solar radiation (kJ/m2) data were obtained from the Global Solar Atlas (http://globalsolaratlas.info).

Seed features associated parameters with the yield, including the pod number, seed per pod, pod width (mm), pod length (mm), 100-seed weight (mg), seed length (mm), and seed width (mm), were recorded numerically in three replications (Excel S1).

100 mg of the powdered seeds were used to extract the free amino acids and 4-hydroxyisoleucine. Briefly, 1000 ul of 80% ethanol (CAS #: 64-17-5) was added to each tube. Then, the tubes were incubated at 80 °C for 60 min and centrifuged at 10,000 rpm for 5 min. Then, the supernatant was collected and dried in a concentrator47. O-phthaldialdehyde (OPA) (CAS #: 643-79-8) was used for the derivatization of the free amino acids. Separation and detection of the amino acids were performed fluorometrically by the HPLC at 340 nm excitation and 450 nm emission (Knauer, Berlin, Germany). OPA- derivatives of the amino acid were separated by the Eurospher C18 column (250 mm × 4.6 mm, with five µm of particle size) and incubated at 28 °C. The OPA reagent was a mixture containing 0.05 g of OPA powder, 4500 ul of methanol, and 500 ul of Borate buffer (2.85 g of Borax dissolved in 100 ml of water at 100 °C, pH 9.5) as well as 2 ul of 2-Mercaptoethanol. Solvent A consisted of 80 ml of 0.05 M acetate sodium and 20 ml of methanol. A mixture containing 30 ml of 0.05 M acetate sodium and 70 ml of methanol was used as Solvent B. Furthermore, 100 ul of the sample and 50 ul of OPA reagent were mixed for 30 min before injection. Then, 20 ul of the mixture was used for injection. Chromatographic separations were achieved at a column temperature of 28 °C and a flow rate of 1 mL/min. The ChromGate software was used for peak integration and the data were manually reviewed for quality of integration and compared against a known standard to confirm identities. 4-hydroxyisoleucine was measured fluorometrically at excitation of 330 nm and emission of 440 nm according to the literature48.

Trigonelline was extracted from 100 mg of the powdered seeds. First, one ml of methanol was added, and the mixture was vortexed for 10 min. Then, the mixture was sonicated (BANDELIN, SONOREX DIGITEC, Germany) for 30 min at room temperature. The supernatants were transferred to the new tubes, and the previous step was repeated as previously described49. The quantification of trigonelline was measured using Agilent 1260 Infinity series HPLC (Agilent Technologies, Santa Clara, CA) consisting of a No. 616 pump, No. 996 diode array detector as well as No. 717 auto‐sampler, and Aminex column. The mobile phase was water (pH 3.5): methanol (70.0:30.0 v/v) with a flow rate of 0.5 ml/min with an elution time of 6 min. The injection volume was 10.0 µl of the methanolic extract. The absorption wavelength was 263 nm. The OpenLab software was applied for data acquisition.

100 mg of the powdered seeds were sonicated in 80% methanol (CAS #: 67-56-1) for 1 h at room temperature and then centrifuged at 5000 rpm for 15 min. The supernatants were collected and transferred to the new tubes, and the volume reached 100 ml by 80% methanol. The phenol content was measured according to the method described by Agbor and Vinson49. Briefly, 100 ul of methanolic extract was diluted with 9 ml of water and mixed with 1 ml of Folin-Ciocalteu reagent shaken for 10 min. Thereafter, 10 ml of w/v sodium carbonate aqueous solution was added to the tubes and vortexed for 30 min. The absorbance was measured at 760 nm using the UV–Visible spectrophotometer (CARY 330 UV–Vis system) against a blank containing Methanol without extract. The results were expressed as mgQE/100 gr of dry seed.

The flavonoid contents were measured as previously described by Kim et al.50. An aliquot of 1 mL of extract solution, 4 ml of distilled water, and 300 ul of 0.3% NaNO2 were vortexed for 5 min. After that, 300 ul of 10% AlCl3 and 200 ul of 1 M NaOH were added to the mixture. In the final step, 2.4 ml of distilled water was added to the mixture and thoroughly shaken at room temperature. The flavonoid content was measured at 510 nm against the blank by the UV/VIS Spectrometer (CARY 330 UV–VIS system). Quercetin compound was used to quantify the samples and the results were reported as mgQE/100 gr of dry seed.

The Radical Scavenging Activity (RSA) of the crude extract solution of the fenugreek’s seed was used to measure the antioxidant activity using the DPHH (2,2-diphenyl-1-picrylhydrazyl free radical) solution using a method previously described by Molyneux51. Briefly, 200 μl of the crude extract was added to 300 μl of methanol and 2400 μl of DPPH solution (0.025 g/l), and vigorously shaken and then incubated in a dark room for 30 min. The absorbance of the mixture was measured at 517 nm using the UV/VIS Spectrometer (CARY 330 UV–VIS system). The control had all reagents except seed extract. The percentage of antioxidant activity was estimated using the equation:

The following formula was used to report the quantification of all primary and secondary metabolites in mg per gr dry seed (Table S2): X (mg/gr dry seed) = (VC)/W, where X = amount of amino acids compounds in the extract, V = final content of the extract solution in ml, W = weight of the extract and C = concentration of standard compounds obtained from the calibration curve.

Pearson correlation was done by corrplot52. The HCPC method was employed for the PCA and HCA analyses53. The GCV and PCV were estimated using the method described by Hanson et al.54. Genetic Advance (GA) was calculated by a method described by Herbert et al.55. Canonical correspondence analysis (CCA) was carried out by the vegan56 package of R57. Partial least square regression (PLSR) analysis was performed using XLSTAT software. The method of leave–one–out cross-validation was utilized to determine the significant component numbers and Root Mean Square Error of Prediction (RMSEP) for the cross-validation process was used in the PLSR analysis to prevent the overfitting.

All data generated or analysed during this study are included in this published article and its supplementary information files.

Megias, C. et al. Free amino acids, Including Canavanine, in the Seeds from 24 Wild Mediterranean Legumes. J. Food Chem. Nanotechnol. 2, 178–183 (2016).

Legume Phylogeny Working Group. Legume phylogeny and classification in the 21st century: Progress , prospects and lessons for other species-rich clades. Taxon. 62, 217–248 (2013).

Grela, E. R., Kiczorowska, B., Samolińska, W. & Matras, J. Chemical composition of leguminous seeds : part I — content of basic nutrients, amino acids, phytochemical compounds, and antioxidant activity. Eur. Food Res. Technol. 243, 1385–1395 (2017).

Bradshaw, A. D. Producing patterns in plants. New Phytol. 170, 639–641 (2006).

Brunetti, C., George, R. M., Tattini, M., Field, K. & Davey, M. P. Metabolomics in plant environmental physiology. J. Exp. Bot. 64, 4011–4020 (2013).

Allevato, D. M., Kiyota, E., Mazzafera, P. & Nixon, K. C. Ecometabolomic analysis of wild populations of Pilocarpus pennatifolius (Rutaceae ) Using Unimodal Analyses. Front. Plant Sci. 10 (2019).

Chen, W., Hou, L., Zhang, Z., Pang, X. & Li, Y. Genetic diversity, population structure, and linkage disequilibrium of a core collection of Ziziphus jujuba assessed with genome-wide SNPs developed by genotyping-by-sequencing and SSR. Markers. 8, 1–14 (2017).

Aljuhaimi, F., Şimşek, Ş., Özcan, M. M., Ghafoor, K. & Babiker, E. E. Effect of location on chemical properties, amino acid and fatty acid compositions of fenugreek (Trigonella foenum-graecum L.) seed and oils. J. Food Process. Preserv. 42, e13569 (2018).

Kapoor, N. & Pande, V. Antioxidative defense to salt stress in Trigonella foenum-graecum L. Curr. discov. 2, 123–127 (2015).

Kyani, A. & Niknam, V. Comparative responses of two Trigonella species to salinity and drought stresses in vitro. Prog. Biol. Sci. 5, 233–248 (2015).

Saberali, S. F. & Moradi, M. Effect of salinity on germination and seedling growth of Trigonella foenum-graecum, Dracocephalum moldavica, Satureja hortensis and Anethum graveolens. J. Saudi Soc. Agric. Sci. 18, 316–323 (2019).

Ahari, D. S., Kashi, A. K., Hassandokht, M. R., Amri, A. & Alizadeh, K. Assessment of drought tolerance in Iranian fenugreek landraces. J. Food Agric. Environ. 7, 414–419 (2009).

Meena, S. et al. Water stress induced biochemical changes in fenugreek (Trigonella foenum graecum L .) genotypes. International J. Seed Spices. 6, 61–70 (2016).

Nour, A. A. M. & Magboul, B. I. Chemical and amino acid composition of fenugreek seeds grown in Sudan. Food Chem. 22, 1–5 (1986).

Hassanzadeh, E., Chaichi, M. R., Mazaheri, D., Rezazadeh, S. & Badi, H. A. N. Physical and chemical variabilities among domestic Iranian Fenugreek (Trigonella foenum-graceum) seeds. Asian J. Plant Sci. 10, 323–330 (2011).

Nagulapalli Venkata, K. C., Swaroop, A., Bagchi, D. & Bishayee, A. A small plant with big benefits: Fenugreek (Trigonella foenum-graecum Linn.) for disease prevention and health promotion. Mol. Nutr. Food Res. 61 (2017).

Robinson, A. R., Ukrainetz, N. K., Kang, K., Mansfield, S. D. & Mansfield, S. D. Metabolite profiling of Douglas-fir (Pseudotsuga menziesii ) field trials reveals strong environmental and weak genetic variation. New Phytol . 174, 762–773 (2003).

Wang, C. et al. Extraction of sensitive bands for monitoring the winter wheat (Triticum aestivum) growth status and yields based on the spectral reflectance. PLoS ONE. 12, 1–16 (2017).

Mehrafarin, A. et al. Bioengineering of important secondary metabolites and metabolic pathways in fenugreek (Trigonella foenum-graecum L.). J. Med. Plants 9, 1–18 (2010).

Bhutia, P. H. & Sharangi, A. B. Influence of dates of sowing and irrigation scheduling on phenology, growth and yield dynamics of fenugreek (Trigonella foenum greacum L .). Legume Res. 41, 275–280 (2018).

Guillermo A. A. Dosio, Luis A. N. Aguirreza´bal,* Fernando H. Andrade & ABSTRACT, V. R. P. Solar Radiation Intercepted during Seed Filling and Oil Production in Two Sunflower Hybrids. Crop Sci. 40, 1637–1644 (2000).

Ishimaru, T. et al. High temperature and low solar radiation during ripening differentially affect the composition of milky-white grains in rice (Oryza sativa L.). Plant Prod. Sci. 21, 370–379 (2018).

Larsson, S., Wirén, A., Lundgren, L. & Ericsson, T. Effects of light and nutrient stress on leaf phenolic chemistry in salix dasyclados and susceptibility to galerucella lineola (Coleoptera). Oikos. 47, 205–210 (1986).

Chua, I. Y. P., King, P. J. H., Ong, K. H., Sarbini, S. R. & Yiu, P. H. Influence of light intensity and temperature on antioxidant activity in Premna serratifolia L. J. Soil Sci. Plant Nutr. 15, 605–614 (2015).

Sharma, A., Shahzad, B., Rehman, A., Bhardwaj, R., Landi, M. Response of phenylpropanoid pathway and the role of polyphenols in plants under abiotic stress. Molecules. 4, 2452. https://www.mdpi.com/1420-3049/24/13/2452

Vaast, P. et al. Shade: A key factor for coffee sustainability and quality. 20th Int. Conf. Coffee Sci. 4, 887–896 (2005).

Borges, C. V., Minatel, I. O., Gomez-Gomez, H. A. & Lima, G. P. P. Medicinal plants: Influence of environmental factors on the content of secondary metabolites. Med. Plants Environ. Challenges 259–277 (2017). https://doi.org/10.1007/978-3-319-68717-9_15

Hanifah, A., Maharijaya, A., Putri, S. P., Laviña, W. A. & Sobir. Untargeted metabolomics analysis of eggplant (Solanum melongena L.) fruit and its correlation to fruit morphologies. Metabolites 8, (2018).

Szakiel, A. & Henry, M. Influence of environmental biotic factors on the content of saponins in plants Influence of environmental abiotic factors on the content of saponins in plants. Phytochem Rev. 10, (2011). https://doi.org/10.1007/s11101-010-9177-x

Obata, T. & Fernie, A. R. The use of metabolomics to dissect plant responses to abiotic stresses. Cell. Mol. Life Sci. 69, 3225–3243 (2012).

Hamidou, M. et al. Genetic Variability and its implications on early generation sorghum lines selection for yield, Yield contributing traits, and resistance to sorghum midge. 2018, (2018).

Tierno, R. & Galarreta, J. I. R. De. Heritability of target bioactive compounds and hydrophilic antioxidant capacity in purple- and red- fl eshed tetraploid potatoes. Crop Pasture Sci. 67, 1309–1317 (2016).

Culley, D. Variation of anthocyanin and carotenoid contents and associated antioxidant values in potato breeding Lines. J. Am. Soc. Hortic. Sci. 130 (2005). https://doi.org/10.21273/JASHS.130.2.174

Stephens, M. J., Hall, H. K. & Alspach, P. A. Variation and heritabilities of antioxidant activity and total phenolic content estimated from a red raspberry factorial experiment. J. Am. Soc. Hortic. Sci. 130, 403–411 (2015). https://doi.org/10.21273/JASHS.130.3.403

Antoine, M., Nicolas, Y., Noubissié, T. J., Marcel, R. & Martin, J. Genetics of seed flavonoid content and antioxidant activity in cowpea (Vigna unguiculata L . Walp .). Crop J. 4, 391–397 (2016).

Matros, A. et al. Genome-metabolite associations revealed low heritability, high genetic complexity, and causal relations for leaf metabolites in winter wheat (Triticum aestivum L.). J. Exp. Bot . 68, 415–428 (2017).

Labarrere, B., Prinzing, A., Dorey, T., Chesneau, E., Hennion, F. Variations of secondary metabolites among natural suggest functional redundancy and versatility. Plants. 19, 234 (2019).

Ghaffari, M. R., Shahinnia, F. & Schreiber, F. The metabolic signature of biomass formation in barley. Plant Cell Physiol. 57, 1943–1960 (2016).

Aryal S, Baniya MK, Danekhu K, Kunwar P, Gurung R, Koirala N. Total phenolic content, flavonoid content and antioxidant potential of wild vegetables from Western Nepal. Plants. 11, 96 (2019).

Praksh, R., Singh, D., Meena, B. L., Kumari, R. & Meena, S. K. Assessment of genetic variability, heritability and genetic advance for quantitative traits in Fenugreek (Trigonella foenum-graecum L.). Int. J. Curr. Microbiol. App.Sci. 6, 2389–2399 (2017).

Haefelé, C., Bonfils, C. & Sauvaire, Y. Characterization of a dioxygenase from Trigonella foenum-graecum involved in 4-hydroxyisoleucine biosynthesis. Phytochemistry 44, 563–566 (1997).

Zafar, M. I. & Gao, F. 4-Hydroxyisoleucine: A Potential New Treatment for Type 2 Diabetes Mellitus. BioDrugs 30, 255–262 (2016).

Hosamath, J. V., Hegde, R. V & Venugopal, C. K. Studies on genetic variability, heritability and genetic advance in Fenugreek (Trigonella foenum-graecum L .). Int. J. Curr. Microbiol. App.Sci. 6, 4029–4036 (2017).

Al-Naggar AM, El-Salam R, Badran AE, Boulos ST, El-Moghazi M. Heritability and genetic advance from selection for morphological, biochemical and anatomical traits of Chenopodium quinoa under water stress. Bionature. 38, 66–85 (2018).

Yadav, T. C., Meena, R. S. & Dhakar, L. Genetic variability analysis in Fenugreek (Trigonella foenum-graecum L.) Genotypes. Int. J. Curr. Microbiol. App. Sci. 7, 2998–3003 (2018).

Di Martino, C., Delfine, S., Pizzuto, R., Loreto, F. & Fuggi, A. Free amino acids and glycine betaine in leaf osmoregulation of spinach responding to increasing salt stress. New Phytol. 158, 455–463 (2003).

Haji, M. H., Sadat, E. S., Amanzadeh, Y., Izaddoust, M., Givi, E. Identification and quantitative determination of 4-hydroxyisoleucine in Trigonella foenum-graecum L. from Iran. J. Med. Plants 9, 29–34 (2010).

Zhuo, R., Wang, L., Wang, L., Xiao, H. & Cai, S. Determination of trigonelline in Trigonella foenum-graecum L. by hydrophilic interaction chromatography. Se pu. 28, 379—382 (2010).

Kim, D. O., Jeong, S. W. & Lee, C. Y. Antioxidant capacity of phenolic phytochemicals from various cultivars of plums. Food Chem. 81, 321–326 (2003).

Molyneux, P. The use of the stable free radical diphenylpicryl- hydrazyl (DPPH) for estimating antioxidant activity. Songklanakarin J. Sci. Technol. 26, 211–216 (2004)

Wei, T. Title visualization of a correlation matrix. R Packag. 56, 1–17 (2017).

Husson, F., Josse, J. and Pages, J., 2010. Principal component methods-hierarchical clustering-partitional clustering: why would we need to choose for visualizing data. Applied Mathematics Department, 17 (2010)

Hanson, C. H., Robinson, H. F. & Comstock, R. E. Biometrical studies of yield in segregating populations of Korean Lespedeza1. Agron. J. 48, 268–272 (1956).

Herbert, W., Robinson, H. F. & Comstock, R. E. Estimates of Genetic and Environmental Variability in Soybeans. Agron J. 46, 314–318 (1955).

Oksanen, J. et al. Package ‘vegan’ title community ecology package. Community Ecol. Packag. 2, 1–297 (2019).

R Foundation for Statistical Computing. R: A language and environment for statistical computing. Vienna, Austria 2, (2008).

This project was supported by the Agricultural Biotechnology Research Institute of Iran (ABRII). We also acknowledge Earth Science and Remote Sensing Unit, NASA Johnson Space Center, to provide the map source in our manuscript.

This project was funded by the Agricultural Biotechnology Research Institute of Iran (ABRII).

Department of Agronomy and Plant Breeding, Faculty of Agriculture and Natural Resources, University of Mohaghegh Ardabili, Ardabil, Iran

Masoumeh Rajabihashjin & Ali Asghari

Systems and Synthetic Biology Department, Agricultural Biotechnology Research Institute of Iran (ABRII), Agricultural Research, Education and Extension Organization (AREEO), Karaj, Iran

Masoumeh Rajabihashjin, Mehrshad Zeinalabedini, Mohammad Reza Ghaffari & Ghasem Hosseini Salekdeh

Department of Molecular Sciences, Macquarie University, North Ryde, NSW, Australia

You can also search for this author in PubMed  Google Scholar

You can also search for this author in PubMed  Google Scholar

You can also search for this author in PubMed  Google Scholar

You can also search for this author in PubMed  Google Scholar

You can also search for this author in PubMed  Google Scholar

M.Z. and A.A. conceived and designed the experiments. M.R. and M.R.G. analyzed the data. M.R., M.R.G., G.H.S. and M.Z. wrote the manuscript.

Correspondence to Mehrshad Zeinalabedini or Ali Asghari.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Rajabihashjin, M., Zeinalabedini, M., Asghari, A. et al. Impact of environmental variables on yield related traits and bioactive compounds of the Persian fenugreek (Trigonella foenum-graecum L.) populations. Sci Rep 12, 7359 (2022). https://doi.org/10.1038/s41598-022-10940-3

DOI: https://doi.org/10.1038/s41598-022-10940-3

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Scientific Reports (Sci Rep) ISSN 2045-2322 (online)

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.