Phenotyping elucidates the association between traits and bioactive components in Notopterygium incisum Authors: Ping Chen, Yili Zhang, Xiaoyan Zheng, Jingyao Tang, Qian Tian, Jia He, Lilin Zhang, Shengyou Li, Renlin Zheng, Liangchun Li, Zhikang Tang, Dabin Hou Published: 2026 ABSTRACT BackgroundNatural products enriched in medicinal plants have broad prospects in addressing global public health challenges and meeting diverse human health needs. The traditional Chinese herbal medicine Notopterygii Rhizoma et Radix (NRR) is the dried rhizome and root of Notopterygium incisum Ting ex H. T. Chang, which is used to treat wind-cold-caused headaches, rheumatic disease, osteoarthritis, and anti-thrombotic treatment. Commonly, NRR collected from wild collection faces the issues of resource depletion and unstable quality.ResultsIn this study, we investigated the phenotypical traits and bioactive components in the wild collected individuals of N. incisum to find and conserve the excellent germplasm. Results showed that the genetic diversity of N. incisum is rich and led to the high variation of the phenotypic traits, e.g., aboveground traits with coefficient of variation (CV) ranged from 21.3% ~ 55.5% and reproductive growth traits ranged from 34.7% to 90.5%, and bioactive pharmaceutical components, e.g., notopterol ranging from 0.02% to 1.27% and isoimperatorin ranging from 0.16% to 3.32%. According to the variation in bioactive pharmaceutical components, N. incisum was divided into three groups: high-quality, medium-quality, and low-quality group. The bioactive pharmaceutical components isoimperatorin and notopterol peaked at 2.38%, 0.83% in the high-quality group; in contrast, the lowest isoimperatorin and notopterol at 0.69% and 0.33% were obtained in the low-quality group. The Mantel test showed that basal leaf petiole length (r = 0.187, p < 0.05) and basal leaf width (r = 0.107, p < 0.05) were significantly affected by genetic diversity, which contributed to germplasm identification.ConclusionsThis study integrates phenotypic characterization with chemical analysis to investigate the correlations between key traits and bioactive compounds in N. incisum. The high-quality group was characterized by robust vegetative growth and weak reproductive growth, while the low-quality group showed the opposite. The current research provides a potential practical reference method for screening high-quality germplasm resources of N. incisum. That is, the quality of N. incisum germplasm resources can be distinguished based on the strength of reproductive growth ability.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12870-026-08871-6. Introduction As traditional Chinese medicine, the use history of Notopterygii Rhizoma et Radix (NRR) is more than 1,800 years, which was first documented in Shengnong’ Classic of Materia Medica [1]. NRR is the dried rhizome and root of Notopterygium incisum Ting ex H. T. Chang or Notopterygium franchetii H. de Boiss [2], and it serves to relieve exterior syndrome and dispel cold, as well as expel wind [3]. Therefore, NRR is used to treat wind-cold-caused headaches, shoulder pain, or rheumatic disease [4]. According to the Pharmacopoeia of the people’s Republic of China, the critical indicators of NRR are alcohol soluble extract (≥ 15.5%), volatile oil (≥ 1.4%), and the total content of notopterol and isoimperatorin (≥ 0.4%) [3]. However, N. incisum has better efficacy than N. franchetii due to the greater content of coumarins, e.g., notopterol and isoimperatorin [5]. Recent studies have reported that coumarins exhibit anti-cancer [6], anti-inflammatory and antioxidant [7], block calcium channels, anti-coagulant, act as anti-convulsant, and vasodilator [8]. Commonly, NRR comes from wild collection, which leads to serious issues. On the one hand, the rapidly rising market demand potentially threatens the wild germplasm resources of N. incisum and N. franchetii; on the other hand, the uncertain quality of NRR limits its efficacy [9]. Germplasm resources are fundamental for breeding and improvement, and screening and utilization of high-quality wild germplasm of N. incisum contribute to the conservation of wild germplasm and ensure the quality of NRR. A highly variable genetic marker is crucial for diversity research, which can generate multi-locus information from the genome [10]. Generally, molecular markers contribute to revealing genetic variation at the DNA level, for example, simple sequence repeat (SSR) [11–15] and inter-simple sequence repeat (ISSR) methods [16]. However, the ISSR method is suitable for evaluating species with limited genomic information. Besides, ISSR has the advantage of stable amplification and high resolution, which is widely employed to analyze genetic diversity and genetic relationship of plant germplasm resources, including medicinal plants like Rhizoma corydalis [17], Rhodiola rosea L [18], and Paris polyphylla var. yunnanensis [19], Zingiber officinale Roscoe [20], and crops like barley [21], Sesamum indicum L [22]., and Vigna unguiculata (L.) Walp [23], as well as woody plants like Camellia oleifolia [24], Myristica fragrans Houtt [25], and Sapindus L [26]. Despite ISSR having advantages for genetic diversity and genetic relationship analysis, plant phenotypic traits are easily collected in the field. The plants’ phenotypic traits are shaped by the interactions of genotype and environmental factors [27]. Genotype determines plants’ morphological characteristics, e.g., plant height [28], leaf shape [29], leaf color [30], and root traits [31]. In addition, differences in genotypes are multifaceted characteristics of plants, including growth, development, physiological processes, and metabolism [28, 31, 32]. Deciphering the variation and driving mechanisms resulting from genotype differences can help optimize plant growth and physiological processes. Especially for medicinal plants, the medicinal bioactive components are mainly secondary metabolites. Breeding high-quality medicinal cultivars contributed to the conservation and utilization of germplasm resources [32]. Phenotypes of medicinal plants are associated with the active pharmaceutical components. For instance, the total saponin content of Paris polyphylla var. yunnanensis is significantly positively correlated with leaf length and the leaf length-to-width ratio [33], and the leaf blade width and stem diameter of Bletilla are negatively correlated with polysaccharide and total phenol contents [34]. The roots of N. incisum with dark green basal leaves and purple petioles have high notopinol isoimperatorin contents [35]. Regarding the trait of belowground, the total alkaloid content of Fritillaria cirrhosa is notably positively correlated with fruit length and bulb fresh weight [36]. The total saponin cumulative quantity of Panax notoginseng is significantly positively correlated with root diameter and total root volume [37]. Revealing the relationship between the morphological traits and the bioactive components contributed to the conservation and breeding of medicinal plants. However, few studies have investigated the correlation between phenotypes and bioactive components of N. incisum germplasm resources. In this study, we collected 51 germplasms of N. incisum from the genuine producing areas in Sichuan province, China. The objectives of this study were (1) to investigate characteristics of genetic diversity, phenotypes, and active pharmaceutical ingredients of collected N. incisum germplasm resources, (2) to evaluate the associations among genetic diversity, phenotypes, and active pharmaceutical ingredients. The ISSR was used to investigate the genetic diversity and genetic relationship. By comprehensively analyzing the relationships among genetic, phenotypes, and active pharmaceutical ingredients, this work can provides a practical basis for marker-assisted breeding and the conservation of high-quality N. incisum germplasm. Besides, further studies should validate the identified phenotype–bioactive correlations in larger populations and explore their underlying molecular mechanisms. Then, developing rapid, non-destructive quality evaluation tools based on critical phenotype markers for field-level selection and sustainable utilization. Materials and methods Samples collections In this study, a total of 51 wild-collected individuals were collected at different locations in Xiaojin County, and then transplanted to the resource garden in Xiaojin County (102.38˚E, 30.85˚N), Sichuan province, China, at an altitude of 3207 m. The plant traits of each individual were investigated at the flowering and sprout tumble stages. The samples of Notopterygium incisum Ting ex H. T. Chang were identified by Prof. Dabin Hou (Southwest University of Science and Technology), and the representative voucher specimens have been deposited in the Herbarium of the School of Life Sciences and Agri-forestry, Southwest University of Science and Technology (QH2024001-QH2024051). The average annual temperatures in summer and winter are 18℃ and 3℃. The young and tender leaves of N. incisum that without pests and diseases damaged were sampled and stored in liquid nitrogen. Then, transport back to the laboratory and store at -80℃. At the sprout tumble stages, the root of the N. incisum was dug out and shade-dried at a final water content of 13%. Determine of phenotype traits At the flowering stage, the morphological traits of N. incisum individuals were evaluated as listed in Table 1. A total of 17 quantitative traits were quantified, and the corresponding method was described in Table 1. Table 1Quantitative parameters used in the characterization of Notopterygium incisumNoParameterCharacter description and measurement method1RL-Root length (cm)Measured from the base of the stem to the main root tip.2UD-Upper root diameter (mm)Diameter of the upper end of the root.3UD-Middle root diameter (mm)Diameter of the middle of the root.4LD-Lower root diameter (mm)Diameter of the lower end of the root.5DW-Root dry weight (g)Weighing after oven-drying to constant weight.6LL-Basal leaf length (cm)After the first basal leaf has fully deployed, measure the length of the first pair of pinna petiolules along the central axis of the leaf to the tip of the leaf.7LW-Basal leaf width (cm)After the first basal leaf has fully deployed, measure the maximum width of the first pair of pinna petiolules.8PL-Basal leaf petiole length (cm)After the first basal leaf has fully deployed, measure the length of the first pair of pinna petiolules to the end of the basal leaf.9PD-Basal leaf petiole diameter (mm)After the first basal leaf has fully deployed, measure the diameter of the basal leaf petiole.10PH-Plant hight (cm)The height of the main flower inflorescence at full flowering stage.11FSD-Flower stem diameter (mm)Diameter of the lower end of the main flower inflorescence at the full flowering stage.12FSN-Flower stem number (No.)Flower stem number per plant.13IN-Inflorescences number (No.)Inflorescences number per plant.14NN-Node number (No.)Node number of flower inflorescence.15SLN-Number of leaves on stem (No.)Number of cauline leaves per plant.16UF-Umbel width (cm)The maximum width of the umbel17UH-Umbel height (cm)The hight from the base of the umbel to the top floret. Table 1Quantitative parameters used in the characterization of Notopterygium incisumNoParameterCharacter description and measurement method1RL-Root length (cm)Measured from the base of the stem to the main root tip.2UD-Upper root diameter (mm)Diameter of the upper end of the root.3UD-Middle root diameter (mm)Diameter of the middle of the root.4LD-Lower root diameter (mm)Diameter of the lower end of the root.5DW-Root dry weight (g)Weighing after oven-drying to constant weight.6LL-Basal leaf length (cm)After the first basal leaf has fully deployed, measure the length of the first pair of pinna petiolules along the central axis of the leaf to the tip of the leaf.7LW-Basal leaf width (cm)After the first basal leaf has fully deployed, measure the maximum width of the first pair of pinna petiolules.8PL-Basal leaf petiole length (cm)After the first basal leaf has fully deployed, measure the length of the first pair of pinna petiolules to the end of the basal leaf.9PD-Basal leaf petiole diameter (mm)After the first basal leaf has fully deployed, measure the diameter of the basal leaf petiole.10PH-Plant hight (cm)The height of the main flower inflorescence at full flowering stage.11FSD-Flower stem diameter (mm)Diameter of the lower end of the main flower inflorescence at the full flowering stage.12FSN-Flower stem number (No.)Flower stem number per plant.13IN-Inflorescences number (No.)Inflorescences number per plant.14NN-Node number (No.)Node number of flower inflorescence.15SLN-Number of leaves on stem (No.)Number of cauline leaves per plant.16UF-Umbel width (cm)The maximum width of the umbel17UH-Umbel height (cm)The hight from the base of the umbel to the top floret. Quantitative parameters used in the characterization of Notopterygium incisum DNA extract and detection The genomic DNA of individual samples was extracted according to the manual of the plant genomic DNA extraction kit (#DP305, Tiangen Biotech, Beijing, China). The quality of the extracted DNA was tested by 1% (w/v) agarose gel electrophoresis in 1× Tris-acetate-EDTA (TAE) buffer with 2kD DNA marker at 140 V for 25 min. Then, the gel was stained with ethidium bromide and checked under UV light (Bio-Rad Gel Doc XR Imaging Systems, USA) to ensure that the genomic DNA was not degraded. The concentration of DNA was determined with NanoDrop One (Thermo Fisher, USA), and diluted to 50 ng µL− 1 and stored at -80 °C for ISSR amplification analysis. ISSR Polymerase Chain Reaction (PCR) primer screening and amplification A total of 15 ISSR primers were randomly selected from the University of British Columbia primer set #9, and the ISSR-PCR amplification was performed (PCR instrument, BIO-RAD T100 PCR, USA). The ISSR-PCR reaction system was established as 20 µL: 1 µL of DNA template, 1 µL of primer (10 µmol L− 1), 10 µL of 2× Taq master mix (Sangon Biotech, Ltd.: B639295-0005), and 8 µL of dd H2O. The PCR reaction program: pre-denaturation at 94℃ for 3 min, denaturation at 94℃ for 30s, annealing for 30s (Table 2), extension at 72℃ for 1 min with 38 cycles, extended at 72℃ for 5 min, and storage at 4℃. The ISSR amplification reaction was conducted as 1% (w/v) agarose gel electrophoresis in 1× TAE buffer with 2kD DNA marker at 140 V for 25 min. Then, the gel was stained with ethidium bromide and scanned with a gel imaging system (Bio-Rad Gel Doc XR Imaging Systems, USA). A total of 4 random samples were tested (Fig. S1), and screened out 9 ISSR primers with high polymorphism, good stability, and good clarity were screened (Table 2). Table 2Primer sequences of ISSRPrimer IDSequence (5’-3’)Annealing temperature (℃)UBC808(AG)8C52UBC809(AG)8G50UBC811(GA)8C52UBC827(AC)8G52UBC834(CT)8G50UBC835(AG)8YC52UBC836(AG)8YA50UBC841(GA)8YC52UBC842(GA)8YG52 Table 2Primer sequences of ISSRPrimer IDSequence (5’-3’)Annealing temperature (℃)UBC808(AG)8C52UBC809(AG)8G50UBC811(GA)8C52UBC827(AC)8G52UBC834(CT)8G50UBC835(AG)8YC52UBC836(AG)8YA50UBC841(GA)8YC52UBC842(GA)8YG52 Primer sequences of ISSR Genetic parameter estimation and cluster analysis The results of ISSR-PCR amplification were extracted by constructing the 0/1 matrix for each band denoted without or with (Fig. S2). The percent of polymorphic bands (PPB) was calculated as follows:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:PPB=\frac{NPL}{NAL}\times\:100\%$$\end{document} Where, NAL is the numbers of amplification band, NPL is the number of polymorphic band, respectively. The POPGENE v.1.32 was used to calculated the number of alleles (Na), number of effective alleles (Ne), Nei’s diversity index (H), Shannon’s index (I), and Nei’s genetic distance [38]. Determination of quality indicators According to the method of Pharmacopoeia of the People’s Republic of China [3], the notopterol and isoimperatorin concentrations were determined using a Shimadzu (LC-2030 Plus, Japan) high-performance liquid chromatography (HPLC) with a Kromasil C18 HPLC column (Kromasil 100-5C18, 250 mm × 4.6 mm). The mobile phase was acetonitrile-water (44:56) with the flow rate at 1 mL per min, and the column temperature was 30 °C with a detection wavelength at 310 nm. The reference solutions were prepared with the standard notopterol (CAS: 88206-46-6) and isoimperatorin (CAS: 482-45-1), and then determined by HPLC (Fig. S3). For each sample preparation, take 0.4000 g of sample powder (passed through a No. 3 sieve) and mix with 50 mL of methanol, sonicate for 30 min, and finally, the filtrate passes through a 0.22 μm filter membrane, and then was determined by HPLC (Fig. S4). Based on the dry weight of samples, the content of notopterol and isoimperatorin was calculated according to the external standard method (general clauses 0512); the threshold of the sum of notopterol and isoimperatorin was 0.40%. According to the method for determination of volatile oil (general clauses 2204), the volatile oil was determined, and the volatile oil threshold is 1.4% [3]. Screening of core germplasm resources According to the method of Wang et al. [39], the core collections of N. incisum was constructed. Two parameters, e.g., mean difference percentage (MD) and coincidence rate of range (CR) were used to evaluate the screening results, which were calculated as follows:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:MD=\frac{{S}_{t}}{n}\times\:100\%$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:CR=\frac{1}{n}\sum\limits_{i=1}^{n}\frac{{R}_{C\left(i\right)}}{{R}_{I\left(i\right)}}\times\:100\%$$\end{document} Where the St is the number of traits with significant differences (t-test, α = 0.05) of mean values between the initial collection and the core collection, and n is the total number of quantitative traits. The RC(i) and RI(i) is the range of the i-th trait in the core collection and the initial collection. The core germplasm resources were screened out on the basis of mean difference percentage (MD) ≤ 20% and coincidence rate of range (CR) ≥ 80% [39]. Statistical and analysis The open source software R version 4.4.1 was used for data analysis [40]. Correlation analysis was conducted to assess the relationships among the indicators, and Pearson correlations were calculated using the R package corrplot [41]. The principal component analysis (PCA) was performed with the base R package stats, and the raw data were scaled with the scale() function before conducting the PCA analysis, then the kmeans() function was applied, and the optimal cluster number was identified as three in this study. To conduct the cluster analysis of ISSR, the raw data were transformed to the Jaccard distance matrix with the vegdist() function and conducted with the UPGMA method with the R package vegan [42], and then visualized with the R package factoextra [43]. The principal coordinate analysis (PCoA) was performed with the R package ape by converting the raw data of ISSR to the Jaccard distance matrix using the vegdist() function [44]. Besides, the relationships between genetic diversity and phenotypes were evaluated by the Mantel test. Firstly, the raw data of ISSR and phenotypes, respectively, were converted to the Jaccard distance matrix and scaled as above, and then converted to the Euclidean distance matrix with the dist() function in the vegan package [42]. Finally, the Mantel test analysis was performed with the method of Pearson and a permutation of 9999 using the R package vegan [42]. The cluster heatmap analysis was performed to identify the characteristics among groups. All the data were scaled with the scale() function, and then visualized with the function pheatmap() in the R package pheatmap [45]. The network analysis was conducted to identify the hub traits and mediator traits and reveal the relationships among those traits. The hub traits were defined as those traits with higher connectivity to other traits, and the mediators were defined as those traits with higher centrality (i.e., betweenness) [46]. In the network analysis, only paths with significant coefficients were visualized (Pearson, p < 0.05). The network analysis was performed with the R package igraph [47]. The other figures were drawn with the R package ggplot2 [48]. Results Functional traits of Notopterygium incisum Results showed that the root dry weight ranged from 256.9 g to 874.1 g, root length ranged from 10.4 cm to 34.3 cm, and root diameter ranged from 3.9 cm to 4.7 cm (Table 3). Although the N. incisum germplasm resources exhibited robust vegetative growth, a subset of individuals had already initiated reproductive development (e.g., flowering stem elongation or inflorescence formation). The CV of root traits, e.g., RL, UD, MD, LD, and DW, ranging from 27.6% to 34.8%, was lower than that of aboveground traits, e.g., LL, LW, PL, PD, and PH, ranging from 21.3% to 55.5%. Besides, the CV of reproductive growth traits, e.g., (FSD, FSN, IN, UW, and UH), peaked and ranged from 34.7% to 90.5%. Table 3Characters of Notopterygium incisum germplasm resourcesParametersMean ± S.D.MaxMinCV (%)RL-Root length (cm)20.12 ± 5.5634.3010.4027.61UD-Upper root diameter (mm)10.52 ± 3.1320.314.6829.76UD-Middle root diameter (mm)9.8 ± 3.2717.713.8933.38LD-Lower root diameter (mm)8.05 ± 2.817.673.8734.79DW-Root dry weight (g)500.09 ± 165.65874.08256.8733.12LL-Basal leaf length (cm)18.52 ± 3.9426.309.1021.26LW-Basal leaf width (cm)22.9 ± 5.6134.5011.4024.50PL-Basal leaf petiole length (cm)13.24 ± 7.3550.602.5055.53PD-Basal leaf petiole diameter (mm)4.05 ± 0.956.531.4123.38PH-Plant hight (cm)79.74 ± 20.26113.5024.8025.41FSD-Flower stem diameter (mm)4.03 ± 1.47.470.0034.73FSN-Flower stem number (No.)8.22 ± 6.3627.000.0077.38IN-Inflorescences number (No.)18.31 ± 16.5869.000.0090.51NN-Node number (No.)3.94 ± 1.146.000.0028.89SLN-Number of cauline leaves (No.)30.88 ± 26.33108.000.0085.25UW-Umbel width (cm)6.6 ± 3.5611.700.0053.96UH-Umbel height (cm)3.63 ± 2.046.900.0056.18 Table 3Characters of Notopterygium incisum germplasm resourcesParametersMean ± S.D.MaxMinCV (%)RL-Root length (cm)20.12 ± 5.5634.3010.4027.61UD-Upper root diameter (mm)10.52 ± 3.1320.314.6829.76UD-Middle root diameter (mm)9.8 ± 3.2717.713.8933.38LD-Lower root diameter (mm)8.05 ± 2.817.673.8734.79DW-Root dry weight (g)500.09 ± 165.65874.08256.8733.12LL-Basal leaf length (cm)18.52 ± 3.9426.309.1021.26LW-Basal leaf width (cm)22.9 ± 5.6134.5011.4024.50PL-Basal leaf petiole length (cm)13.24 ± 7.3550.602.5055.53PD-Basal leaf petiole diameter (mm)4.05 ± 0.956.531.4123.38PH-Plant hight (cm)79.74 ± 20.26113.5024.8025.41FSD-Flower stem diameter (mm)4.03 ± 1.47.470.0034.73FSN-Flower stem number (No.)8.22 ± 6.3627.000.0077.38IN-Inflorescences number (No.)18.31 ± 16.5869.000.0090.51NN-Node number (No.)3.94 ± 1.146.000.0028.89SLN-Number of cauline leaves (No.)30.88 ± 26.33108.000.0085.25UW-Umbel width (cm)6.6 ± 3.5611.700.0053.96UH-Umbel height (cm)3.63 ± 2.046.900.0056.18 Characters of Notopterygium incisum germplasm resources Characteristics of quality indicators The volatile oil content (VO) of 51 samples ranged from 0.29% to 2.54%, and that of 19 samples higher than 1.4% (Fig. 1). The notopterol and isoimperatorin contents, respectively, ranging from 0.02% to 1.27% and 0.16% to 3.32%. The total effective content of isoimperatorin and notopterol (TE) ranged from 0.32% to 4.29%, and the TE of 49 samples greater than 0.4%. Although the TE was mainly contributed by isoimperatorin rather than notopterol, the VO was independent of TE, isoimperatorin, and notopterol (Figs. 1 and 2). Fig. 1The quality indicators of Notopterygium incisum germplasm resources. Panel A, volatile oil content (%). The total effective content (%) in panel B was the sum of notopterol (%) in panel C and isoimperatorin content (%) in panel D The quality indicators of Notopterygium incisum germplasm resources. Panel A, volatile oil content (%). The total effective content (%) in panel B was the sum of notopterol (%) in panel C and isoimperatorin content (%) in panel D Fig. 2Correlation relationships between quality indicators. Isoimp, isoimperatorin; Notop, notopterol; TE, the total effective components of isoimperatorin and notopterol; VO, volatile oil. The *, **, and *** denote significant differences at the 0.05, 0.01, and 0.001 levels, respectively Correlation relationships between quality indicators. Isoimp, isoimperatorin; Notop, notopterol; TE, the total effective components of isoimperatorin and notopterol; VO, volatile oil. The *, **, and *** denote significant differences at the 0.05, 0.01, and 0.001 levels, respectively Quality group classification and corresponding features Based on the content of volatile oil, notopterol, isoimperatorin, and the total components, the current dataset can be classified into three quality groups (QG) using the k-means clustering method. Results of principal component analysis (PCA) showed that the variance explanation rates of PC1 and PC2 were 63.7% and 23.7%, and led to a total variance explanation rate at 87.4% (Fig. 3). The PC1 axis mainly explained the change in quality, and the QG1 and QG3 were almost completely separated, while QG1 and QG3, respectively, were partially mixed with QG2.The PC2 axis mainly explained the variation between QG2 and the other groups. Fig. 3Classification of quality group based on the quality indicators. The k-means method was used to find the optimal number of clusters, which was three in this study. Hence, the PCA was conducted with three groups. QG1, quality group 1; QG2, quality group 2; QG3, quality group 3. The dots represent samples, and the number was samples ID Classification of quality group based on the quality indicators. The k-means method was used to find the optimal number of clusters, which was three in this study. Hence, the PCA was conducted with three groups. QG1, quality group 1; QG2, quality group 2; QG3, quality group 3. The dots represent samples, and the number was samples ID The quality indicators, e.g., isoimperatorin, notopterol, the total components, and volatile oil contents of the three groups were almost completely separated (Fig. 4). Isoimperatorin, notopterol, and the total component contents peaked at 2.38%, 0.83%, and 3.2% in quality group 3 (QG3, high quality), while the volatile oil content was moderate. In contrast, the content of isoimperatorin, notopterol, and the total component in quality group 2 (QG2) were at medium level, while the content of volatile oil peaked. Quality group 1 (QG1, low quality) had the lowest isoimperatorin, notopterol, the total component, and volatile oil at 0.69%, 0.33%, 1.02%, and 0.91%, respectively. Fig. 4The indicators of quality groups. QG1, quality group 1; QG2, quality group 2; QG3, quality group 3. Isoimp, isoimperatorin; Notop, notopterol; TE, the total effective components of isoimperatorin and notopterol; VO, volatile oil. Different lower-case letters denoted significant differences (Tukey, p < 0.05) The indicators of quality groups. QG1, quality group 1; QG2, quality group 2; QG3, quality group 3. Isoimp, isoimperatorin; Notop, notopterol; TE, the total effective components of isoimperatorin and notopterol; VO, volatile oil. Different lower-case letters denoted significant differences (Tukey, p < 0.05) Phenotypic traits of quality groups Results of cluster on the basis of quality group showed that most of the functional traits of QG3 were higher than the average of the overall dataset, while those of QG1 were lower than the overall dataset (Figs. 5 and S1). The performance of QG2 was between QG3 and QG1. QG3 showed robust vegetative growth, e.g., relative high value of plant height, basal leaf length, basal leaf width, basal leaf petiole length, basal leaf petiole diameter, root diameter, and root dry weight (Fig. S2). Besides, QG3 had weak reproductive characteristics, e.g., relative low value of flower stem number, inflorescences number, and umbel width. Fortunately, the well vegetative growth and sufficient nutrient supply potentially contributed to high umbrella height and flower stem diameter. On the contrary, the functional traits of QG1 showed the opposite trend, and those of QG2 were between QG3 and QG1. Fig. 5Functional traits of quality groups. QG1: quality group 1; QG2: quality group 2; QG3: quality group 3. Trait abbreviations: RL, root length; UD, upper root diameter; MD, middle root diameter; LD, lower root diameter; DW, root dry weight; LL, basal leaf length; LW, basal leaf width; PL, basal leaf petiole length; PD, basal leaf petiole diameter; PH, plant height; FSD, flower stem diameter; FSN, flower stem number; IN, inflorescence number; NN, node number; SLN, number of cauline leaves; UW, umbel width; UH, umbel height. All data are presented as z-scored values Functional traits of quality groups. QG1: quality group 1; QG2: quality group 2; QG3: quality group 3. Trait abbreviations: RL, root length; UD, upper root diameter; MD, middle root diameter; LD, lower root diameter; DW, root dry weight; LL, basal leaf length; LW, basal leaf width; PL, basal leaf petiole length; PD, basal leaf petiole diameter; PH, plant height; FSD, flower stem diameter; FSN, flower stem number; IN, inflorescence number; NN, node number; SLN, number of cauline leaves; UW, umbel width; UH, umbel height. All data are presented as z-scored values Genetic diversity and core germplasm resources Results of ISSR showed that 9 primers were screened out for PCR amplification, and a total of 68 brands were obtained (Table 4). A total of 63 polymorphic brands with the average of 7 polymorphic brands and 92.9% of percentages polymorphic bands (PPB). Except for UBC8009, UBC8011, and UBC8042, the PPB for the other primers was ranging from 85.71% to 91.67%. Besides, the Na, Ne, H, I, and PPB were 1.9412, 1.457, 0.273, and 0.4203, and 94.12% (Table 5). The Na, Ne, H, I, and PPB of QG3 were lower than those of QG1, while the genetic distance the lowest (0.0139). Results of PCoA indicated that high overlaps of genetic diversity among the three quality groups (Fig. 6). QG1 was part of QG2 while partially overlaps with QG3, indicating that the genetic characteristics of QG3 differed from those of QG1 (Figs. 6 and S5). Meanwhile, results of the hierarchical clustering also indicated the high overlaps among quality groups (Figs. 6 and S6). Table 4Primer information and amplification results of ISSRPrimerBandsPolymorphic bandsPercentage ofpolymorphic bands (%)UBC8008121191.67UBC8009UBC80116767100100UBC8027UBC8034786885.71100UBC8035UBC8036877687.585.71UBC8041UBC8042Average76866785.7110092.92 Table 4Primer information and amplification results of ISSRPrimerBandsPolymorphic bandsPercentage ofpolymorphic bands (%)UBC8008121191.67UBC8009UBC80116767100100UBC8027UBC8034786885.71100UBC8035UBC8036877687.585.71UBC8041UBC8042Average76866785.7110092.92 Primer information and amplification results of ISSR UBC8009 UBC8011 6 7 6 7 100 100 UBC8027 UBC8034 7 8 6 8 85.71 100 UBC8035 UBC8036 8 7 7 6 87.5 85.71 UBC8041 UBC8042 Average 7 6 8 6 6 7 85.71 100 92.92 Table 5Genetic diversity parameters of Notopterygium incisumGroup N a N e HIPPB (%)Nei’s genetic distanceQG1QG2QG3QG11.89711.47220.28050.426489.71%0.98120.9862QG21.83821.42330.25330.387683.82%0.01900.9777QG31.79411.42940.25390.384879.41%0.01390.0226Overall1.94121.4570.2730.420394.12%Na the number of alleles, Ne number of effective alleles, H Nei’s diversity index, I Shannon’s index. QG1, quality group 1; QG2, quality group 2; and QG3, quality group 3 Table 5Genetic diversity parameters of Notopterygium incisumGroup N a N e HIPPB (%)Nei’s genetic distanceQG1QG2QG3QG11.89711.47220.28050.426489.71%0.98120.9862QG21.83821.42330.25330.387683.82%0.01900.9777QG31.79411.42940.25390.384879.41%0.01390.0226Overall1.94121.4570.2730.420394.12%Na the number of alleles, Ne number of effective alleles, H Nei’s diversity index, I Shannon’s index. QG1, quality group 1; QG2, quality group 2; and QG3, quality group 3 Genetic diversity parameters of Notopterygium incisum Na the number of alleles, Ne number of effective alleles, H Nei’s diversity index, I Shannon’s index. QG1, quality group 1; QG2, quality group 2; and QG3, quality group 3 Fig. 6Principal coordinate analysis (PCoA) of ISSR. QG1, quality group 1; QG2, quality group 2; QG3, quality group 3. The dots represent samples, and the number was samples ID Principal coordinate analysis (PCoA) of ISSR. QG1, quality group 1; QG2, quality group 2; QG3, quality group 3. The dots represent samples, and the number was samples ID Mantel test showed that the basal leaf width and basal leaf petiole length were significantly positively correlated with ISSR (Fig. 7), while the other were independent of ISSR (Table S1). All the individuals were raised in the same condition. Namely, indicating that basal leaf width and basal leaf petiole of N. incisum were genetically driven, the other functional traits were genetically independent. With the increases in genetic distance, the basal leaf width and petiole length increased (Fig. 7). Based on quality group classification, a total of 16 core germplasm resources were screened out, e.g., 6 samples in QG3, 3 samples in QG2, and 7 samples in QG1 (Fig. 8). The PC1 axis explained the variation of quality indicators, e.g., isoimperatorin, notopterol, and the total component contents. Namely, QG3 was better than QG2, while QG1 was the worst. Besides, the quality indicators were positively correlated with most vegetative traits, e.g., NN, LL, LW, LD, and DW, while they were independent of reproductive traits, e.g., FSN and FSD. The PC2 axis explained the variation of volatile oil, and that of QG3 was higher than the other groups. Fig. 7Mantel test analysis of functional traits and ISSR genetic distances. Panel A Mantel test analysis of basal leaf petiole length and ISSR genetic distance; Panel B Mantel test analysis of basal leaf width and ISSR genetic distance. All the functional traits were performed with mantel test and statistically significant models were presented (p < 0.05) Mantel test analysis of functional traits and ISSR genetic distances. Panel A Mantel test analysis of basal leaf petiole length and ISSR genetic distance; Panel B Mantel test analysis of basal leaf width and ISSR genetic distance. All the functional traits were performed with mantel test and statistically significant models were presented (p < 0.05) Fig. 8Screening of core germplasm resources and relationship between functional traits. PC1, principal component 1; PC2, principal component 2. The core germplasm resources were screened out on the basis of mean difference percentage (MD) ≤ 20% and coincidence rate of range (CR) ≥ 80%. QG1, quality group 1; QG2, quality group 2; QG3, quality group 3. Trait abbreviations: RL, root length; UD, upper root diameter; MD, middle root diameter; LD, lower root diameter; DW, root dry weight; LL, basal leaf length; LW, basal leaf width; PL, basal leaf petiole length; PD, basal leaf petiole diameter; PH, plant height; FSD, flower stem diameter; FSN, flower stem number; IN, inflorescences number; NN, node number; SLN, number of cauline leaves; UW, umbel width; UH, umbel height. Isoimp, isoimperatorin; Notop, notopterol; TE, the total effective components of isoimperatorin and notopterol; VO, volatile oil. The dots represent samples, and the number was samples ID Screening of core germplasm resources and relationship between functional traits. PC1, principal component 1; PC2, principal component 2. The core germplasm resources were screened out on the basis of mean difference percentage (MD) ≤ 20% and coincidence rate of range (CR) ≥ 80%. QG1, quality group 1; QG2, quality group 2; QG3, quality group 3. Trait abbreviations: RL, root length; UD, upper root diameter; MD, middle root diameter; LD, lower root diameter; DW, root dry weight; LL, basal leaf length; LW, basal leaf width; PL, basal leaf petiole length; PD, basal leaf petiole diameter; PH, plant height; FSD, flower stem diameter; FSN, flower stem number; IN, inflorescences number; NN, node number; SLN, number of cauline leaves; UW, umbel width; UH, umbel height. Isoimp, isoimperatorin; Notop, notopterol; TE, the total effective components of isoimperatorin and notopterol; VO, volatile oil. The dots represent samples, and the number was samples ID However, the relationships between functional traits and quality indicators varied between quality groups (Figs. S8-S10). In the low quality group (QG1), the TE was positively correlated with NN, Isoimp, and Notop, while LW was negatively correlated with Isoimp (Fig. S8). The NN was positively correlated with Isoimp, PH, FSN, and IN, while negatively correlated with UD and MD in QG1. Regarding QG2, the TE was positively correlated with FSN, DW, and Isoimp. The FSN was positively correlated with Isoimp, while Isoimp was positively correlated with LD and UD (Fig. S9). On the contrary, the TE was positively correlated with DW and Isoimp in the high-quality group (QG3) (Fig. S10). The DW was positively correlated with Isoimp and UD, while negatively correlated with FSD in QG3. Moreover, the functional traits network analysis showed that trait relationships among the three groups were different (Fig. 9). In QG1, the hub traits NN and PH influence aboveground growth, and the mediator trait FSD influences Isoimp via leaf traits, e.g., LL and LW (Fig. 9A). Regarding QG2, the hub traits NN, PH, and LL affect aboveground growth, and the mediator trait MD, UD, and RL influences Isoimp (Fig. 9B). The hub traits FSD, DW, PD, and TE determine the bioactive components of N. incisum (Fig. 9C). Generally, aboveground vegetative traits, e.g., PH and SLN, and reproductive traits, e.g., FSD, IN, and UW, were critical traits shaping the quality of N. incisum (Fig. 9). Fig. 9Networks analysis based on functional traits and quality indicators. Panels A, B, and C were the networks analysis of quality group 1 (low quality), quality group 2 (moderate quality), and quality group 3 (high quality). Traits were set as nodes, and the lines among traits were Pearson correlation coefficients. Blue and red lines denoted positive and negative correlation (p < 0.05). The orange color denoted hub traits, cyan color denoted mediator traits, and gray color denoted common traits. Hub traits with higher betweenness centrality (eigenvector > 0.75), while mediator traits with higher betweenness centrality (betweenness > 0.75). RL, root length; UD, upper root diameter; MD, middle root diameter; LD, lower root diameter; DW, root dry weight; LL, basal leaf length; LW, basal leaf width; PL, basal leaf petiole length; PD, basal leaf petiole diameter; PH, plant hE height; FSD, flower stem diameter; FSN, flower stem number; IN, inflorescences number; NN, node number; SLN, number of cauline leaves; UW, umbel width; UH, umbel height. Isoimp, isoimperatorin; Notop, notopterol; TE, the total effective components; VO, volatile oil Networks analysis based on functional traits and quality indicators. Panels A, B, and C were the networks analysis of quality group 1 (low quality), quality group 2 (moderate quality), and quality group 3 (high quality). Traits were set as nodes, and the lines among traits were Pearson correlation coefficients. Blue and red lines denoted positive and negative correlation (p < 0.05). The orange color denoted hub traits, cyan color denoted mediator traits, and gray color denoted common traits. Hub traits with higher betweenness centrality (eigenvector > 0.75), while mediator traits with higher betweenness centrality (betweenness > 0.75). RL, root length; UD, upper root diameter; MD, middle root diameter; LD, lower root diameter; DW, root dry weight; LL, basal leaf length; LW, basal leaf width; PL, basal leaf petiole length; PD, basal leaf petiole diameter; PH, plant hE height; FSD, flower stem diameter; FSN, flower stem number; IN, inflorescences number; NN, node number; SLN, number of cauline leaves; UW, umbel width; UH, umbel height. Isoimp, isoimperatorin; Notop, notopterol; TE, the total effective components; VO, volatile oil Discussion Diversity of functional traits and quality indicators in Notopterygium incisum Results showed a large variation in the functional traits of N. incisum germplasm resources (Table 3; Fig. 1). The CV of reproductive traits was the highest, and the CV of root traits was lower than that of aboveground traits. It is indicated a high plasticity of reproductive traits and aboveground traits, and a high stability of the root traits. Plant traits including the morphological, anatomical, physiological, and biochemical characteristics, which are co-regulated by the growth environment and genes [49]. Therefore, the plant trait differences reflect the diversity of species genetic resources. Some medicinal materials are collected from wild sources, and the high genetic richness of medicinal plant germplasm resources in natural populations leads to large variation in phenotype and unstable quality [50–52]. On the contrary, it provides rich potential germplasm resources for the breeding of new varieties of medicinal plants [32, 53]. In this study, the N. incisum germplasm resources were classified into three groups, e.g., QG1, QG2, and QG3, according to the PCA analysis based on quality indicators (Fig. 3). Substantially, the quality of medicinal plants can be distinguished by the content of medicinal bioactive components, which is defined as the quality marker (Q-marker). Recent works on Bupleuri chinense DC demonstrated saikosaponins, flavonoids, carboxylic acids, and lignan as the real bio-active components responsible for the established clinical effects of Bupleuri Radix for thousands of years [54, 55]. However, Bupleurum chinense has superior quality than Bupleurum smithii and Bupleurum marginatum var. stenophyllum due to the high content of saikosaponin [56]. Regarding N. incisum, a recent study reported that the coumarins, e.g., nodakenin and isoimperatorin, are potential Q-markers for the treatment of cardiovascular diseases [57]. Besides, the critical bioactive components in N. incisum, e.g., notopterol and isoimperatorin [58], which can be used to treat osteoarthritis and anti-thrombotic [59, 60]. In this study, QG3 characteristiced as high Q-makers contents, while the Q-makers of QG1 were the lowest (Fig. 4). Namely, the QG3 was high quality group and the QG1 was the low quality group. Although the quality of medicinal plants can be accurately determined through equipment, such as high-performance liquid chromatography, or gas chromatography, or mass spectrometer, these methods are complex and expensive. The functional traits of quality groups were easily distinguished, which indicated a convenient method for quality evaluation of N. incisum germplasm resources (Figs. 5 and S2). Namely, revealing the relationship between traits and bioactive components can help to reduce qualification costs of medicinal plant and shorten the breeding period. Numerous studies documented that phenotypes traits of medicinal plants are associated with the active pharmaceutical components [34, 61, 62]. For example, the Q-makers of medicinal plants’ roots are associated with aboveground traits, e.g., the leaf length of Paris polyphylla var. yunnanensis [33], the petiole length of Gymnema sylvestre [53], the leaf and inflorescenece traits of Bletilla Rchb. f [34]., and the stem diameter of Bletilla [34], or associated with belowground traits, e.g., the bulb fresh weight of Fritillaria cirrhosa [36], the root fresh weight of Astragalus membranaceus var. mongholicus [62], the root diameter and total root volume of Panax notoginseng [37], the total root surface of Salvia miltiorrhiza [61]. In this study, the high-quality group (QG3) is characterized as robust vegetative growth, e.g., relative high value of plant height, basal leaf length, basal leaf petiole length, root diameter, and root dry weight, and weak reproductive growth, e.g., relative low value of flower stem number and inflorescences number (Figs. 5 and S2). The Q-maker associated plant traits, such as inflorescences appear or not, provide a rapid screening method for genetic germplasm resources identification. Results of functional traits network and correlation analysis showed the intrinsic connections between vegetative traits, reproductive traits, and quality indicators (Figs. 9 and S8-S10). Indeed, the reproductive growth affect the Q-maker content of medicinal plants. Nutrients are primarily retained in source organs (leaves and roots) during vegetative growth, but are redirected to storage or reproductive sinks (e.g., rhizomes or seeds) during the reproductive stage [63–65]. For instance, the rhizome and total saponin content of Polygonatum cyrtonema are improved by topping [66]. Probably inhibition of reproductive traits (e.g., FSD and FSN) promotes vegetative growth (e.g., LL and DW), leading to enhanced root yield and Q-marker concentration (Fig. 9). The influence of topping, e.g., topping time and topping height, on the Q-maker of N. incisum is worth further studying. Core germplasm resources of in Notopterygium incisum Generally, the rare and endangered plants with limited geographical distribution have low genetic diversity, such as Pulsatilla patens in Europe [67] and Henckelia longisepala in Southeast Asia [68]. On the contrary, the rare and endangered plants can have high genetic diversity, such as Calycanthus chinensis W. C. Cheng & S. Y. Chang [69] and Parashorea chinensis [70]. In the current study, the overall Na, Ne, H, I, and PPB were 1.9412, 1.4570, 0.273, 0.4203, and 94.12, respectively, suggesting relatively high genetic diversity of N. incisum (Table 5). Indeed, results of chloroplast DNA makers indicated the high genetic diversity of N. incisum germplasm [71]. Although the phenotypes are genetically and environmentally co-regulated. In this study, phenotypic distance, e.g., basal leaf petiole and basal leaf width, was notably correlated with ISSR genetic distance (Fig. 7). This indicated a genetically driven variation rather than environmental factors, because all the germplasm were grown at the same location. While results of Nei’s genetic distance and ISSR indicate close association among the three quality groups (Table 5; Figs. 6, and S7), factors contributing to low genetic diversity include small population sizes, habitat fragmentation, human disturbances, and limited gene flow [72, 73]. Maintaining or enhancing gene flow and protect populations to preserve existing genetic variations are critical for species resources [74, 75]. Although N. incisum germplasm resources had a high genetic diversity, the wild collection, destruction of wild habitats, and extensive collection potentially threaten its genetic diversity. A recent study reported that wild germplasm has remarkably higher bioactive components than cultivated N. incisum [5]. The protection and utilization of N. incisum have a long way to go, but the fundamental is the germplasm resources. Generally, germplasm resource conservation is based on the collection of numerous individuals, but the cost of germplasm cultivation needs to be considered, e.g., land and labour. How to preserve as many germplasm resources as possible without increasing cost is a question. Previous study have proposed to build ‘core germplasm resources’, which can achieve the resources diversity and reduce costs [39]. In this study, the core germplasm resources include 16 of 51 samples, and 6 samples in QG3 (Fig. 8). The high-quality samples characterized with robust vegetative growth traits, e.g., LW, LL, and NN, rather than reproductive traits, e.g., FSD, FSN, and IN. Those traits can be considered as the potential characteristics for the high-quality resource screen. Deciphering the variation and driving mechanisms associated with functional traits contributed to cultivar breeding. Especially for the perennial medicinal plants with long growth time, distinguish good quality individuals based on phenotypic traits will help accelerate the breeding process. In return, breeding high-quality medicinal cultivars contributed to meet the market demand and the conservation of rare medicinal resources. Conclusions The genetic diversity of Notopterygium incisum germplasm resources is rich, which leads to variation in phenotypic traits and bioactive pharmaceutical components. According to the bioactive pharmaceutical components of Notopterygium incisum, the samples are divided into three groups, i.e., high-quality, medium-quality, and low-quality. The high-quality group showed robust vegetative growth and weak reproductive growth, while the low-quality group showed the opposite. These findings suggest that reproductive intensity may serve as a rapid, non-destructive field indicator for preliminary quality assessment, contributing to efficient germplasm screening. Future studies integrating genomic, metabolomic, and phenotypic data will enable the development of molecular markers linked to high-quality chemotypes, accelerating marker-assisted selection and sustainable cultivation of N. incisum. Supplementary Information Supplementary Material 1. Supplementary Material 1.