Integrating electrical parameters with agronomic, yield, and quality traits for screening superior Pseudostellaria heterophylla genotypes Authors: Jiaqi Liu, Zelan Wang, Shiyi Pan, Bing Tian, Chenglin Tang, Qiuping Wang, Cheng Zhang, Qinghai Zhang Published: 2026 ABSTRACT ABSTRACT Pseudostellaria heterophylla is increasingly utilized as a functional food ingredient in East Asian cuisines, valued for its bioactive polysaccharides and saponins. Concurrently, the scarcity of genetically diverse and regionally adapted cultivars poses a significant challenge to the sustainable production of this crop. This study integrates the evaluation of agronomic, yield, and nutritional traits with electrical parameters to evaluate 13 P. heterophylla varieties from diverse geographical origins. Significant varietal differences were observed in intracellular water metabolism, nutrient transport, and metabolic activity, as reflected by intrinsic electrical parameters (p < 0.05). Among the tested varieties, “Zheshen I” exhibited the best results, with a fresh weight of 675.74 kg/667 m² and a dry weight of 267.41 kg/667 m². “Zheshen II” contained the highest polysaccharide content (49.27%), while “Zheshen IV” showed the highest saponin content (2.83%). Grey relational analysis identified six varieties—“Guishen I,” “Zheshen IV,” “Xuanshen I,” “Shitai I,” “Anguo,” and “Space breed”—as having superior comprehensive performance, making them suitable for promotion and cultivation in the Guizhou region. Notably, “Guishen I” (the current main cultivated variety) and “Shitai I” (a local variety) ranked among the top performers, suggesting the potential of our experimental evaluation to reflect practical cultivation outcomes. Overall, this study provides quantitative data on the nutritional components of P. heterophylla to support the standardized procurement of raw materials for functional foods, while simultaneously identifying potentially superior genotypes based on phenotypic vigor for regionally adapted breeding. Introduction Pseudostellaria heterophylla (Miq.) Pax ex Pax et Hoffm., a medicinal and edible herb widely cultivated in East Asia, has gained increasing prominence as a functional food ingredient. 1 In culinary traditions, its dried tuberous roots are extensively used in nutritious soups, porridges, and health beverages, valued for their mild flavor and health-promoting properties. 2 The nutritional value of P. heterophylla is primarily attributed to its high concentrations of bioactive polysaccharides and saponins, which serve as key functional components with immunomodulatory and antioxidant activities. 3 As consumer demand for functional foods continues to grow, ensuring the compositional consistency and quality stability of such raw materials has become a critical challenge for the food industry. In China, the major cultivation areas for P. heterophylla include Shibing in Guizhou Province, Zherong in Fujian Province, and Xuanzhou in Anhui Province. 4 Guizhou Province, with its unique ecological conditions, currently leads the nation in terms of cultivation area. However, the sustainable development of the local P. heterophylla industry faces significant challenges. Prolonged dependence on vegetative propagation has caused varietal confusion, genetic degeneration, and a narrowing of the genetic base, thereby increasing susceptibility to diseases and the risk of genetic vulnerability. 5 , 6 Moreover, non-standardized cultivation practices have led to unstable yield and inconsistent product quality. 7 Several cultivars of P. heterophylla have been developed in different regions, including the Zheshen series from Fujian, Kangdu I from Shandong, Xuanshen I from Anhui, as well as Guishen I, Guishen II, and Shitai I from Guizhou. 8 However, the regional adaptability of these varieties remains limited, and continuous cropping obstacles, along with pest and disease pressures, further constrain their performance. 9 P. heterophylla exhibits strong ecological adaptability, and the same variety often shows significant differences across different ecological regions. Therefore, evaluating the adaptability of germplasm from diverse sources is a fundamental prerequisite for germplasm innovation and precision breeding. Consequently, there is a pressing need to establish systematic compositional datasets for the standardization of raw material sourcing in the functional food industry while identifying superior varieties with balanced quality traits to support breeding programs tailored for regional adaptation. In the process of screening new varieties, traditional evaluation methods can obtain key trait data such as yield and quality. However, they often fail to capture the real-time physiological status and adaptive strategies of the varieties throughout the growth cycle. Plant electrical characteristics, manifested through the electrical properties of tissues, represent a highly sensitive indicator of a plant's physiological response to abiotic or biotic stresses, such as salinity, drought, low temperature, pests, and diseases. 10 A seminal theoretical advancement has established a specific relationship between leaf electrical parameters—including resistance (R), impedance (Z), inductive reactance (XL), and capacitive reactance (Xc)—and the clamping force (F) applied during measurement, expressed as R,Z,XL,or Xc=y+ke−bF . 11 , 12 Importantly, when F = 0 (simulating the natural, undisturbed state of the plant), this model enables the nondestructive, real-time, and accurate determination of intrinsic electrical parameters: intrinsic resistance (IR), intrinsic impedance (IZ), intrinsic inductive reactance (IXL), and intrinsic capacitive reactance (IXc). 13 Previous studies have confirmed the biological relevance of these parameters. R, Z, XL, and Xc reflect a plant's capacity for nutrient transport and uptake, 14 while Z correlates with intracellular water transport rates, indicative of overall physiological vigor. 15 These findings suggest that electrical profiling could serve as a rapid, non-destructive tool for both quality assessment of functional food raw materials and early-stage screening of elite varieties. This perspective aligns with recent advances that advocate for the efficient aggregation and utilization of plant electrical signal data to rapidly assess plant physiological states. 16 The primary objective of this study was to develop a robust, multi-dimensional screening system for P. heterophylla germplasm by integrating bioelectrical properties with comprehensive assessments of agronomic traits, yield, and nutritional components. The passive electrical properties of leaves, which reflect cellular membrane integrity and ionic dynamics, serve as physical proxies for the plant's overall physiological vigor. We hypothesized that these biophysical traits are intrinsically coupled with the plant's capacity for biomass accumulation and secondary metabolite synthesis. Consequently, an integrated evaluation using Grey Relational Analysis (GRA) to bridge these different biological scales would provide a more reliable and efficient method for identifying superior genotypes compared to single-trait evaluations. To test this, 13 varieties were evaluated, aiming to provide a theoretical and practical basis for the standardized production of high-quality functional food resources. Materials and methods Experimental site and experimental materials This study was conducted in 2024 at an experimental site in Niu Dachang Village (Niu Dachang Town, Shibing County), within the Qiandongnan Miao and Dong Autonomous Prefecture of Guizhou Province, China. The geographical coordinates of the region range from 107°52′37″ to 108°28′47″ E longitude and 26°46′46″ to 27°20′16″ N latitude. The site is characterized by a warm and humid subtropical monsoon climate, with an annual average temperature ranging from 16.8 °C to 27.8 °C and an average annual precipitation of 1205 mm. Thirteen varieties of P. heterophylla were selected to screen for high-performing varieties suitable for Guizhou. The collection comprises dominant cultivars from major Chinese production areas, representative local varieties, and a space-mutagenesis-derived variety. This comparative framework facilitates the identification of optimal varieties that balance high yield and quality under Guizhou's specific environmental conditions. The information for the tested varieties is detailed in Table 1. Table 1.Key information of the studied varieties.NumberVarietySourceDry weight (kg/667 m2)Growth period (d)CharacteristicsReferencesN1Xuanshen ⅠXuanzhou, Anhui200–280.5130Strong resistance to leaf spot, but susceptible to severe viral disease outbreaks.[17]N2Kangdu ⅠLinshu County Agricultural Bureau, Shandong Province200–250120Highly resistant to mosaic virus; cold-tolerant but sensitive to high temperatures.[18]N3Zheshen ⅡZherong County Agricultural Technology Extension Center, Fujian Province100130Moderate resistance to leaf spot and strong cold hardiness; poor tolerance to heat and waterlogging.[19]N4Zheshen ⅣFujian Zheshen Biotechnology Research Co., Ltd.366.67 (fresh weight)130High yield, large tuberous roots, and virus-resistant.[20]N5Guishen ⅡInstitute of Chinese Materia Medica and Ethnomedicine Resources, Guizhou University of Traditional Chinese Medicine150–180150High extract and cyclic peptide content; excellent resistance to extreme weather and diseases.aN6Qiantai ⅠGuizhou Changhao Traditional Chinese Medicine Development Co., Ltd.208.6150Strong resistance to mosaic disease.[21]N7Space breedFujian Tianren Pharmaceutical Co., Ltd.––––N8Shitai ⅠGuiyang College of Traditional Chinese Medicine530.3 (fresh weight)145–155Lodging-resistant with stable disease resistance.[8]N9Fengkang ⅠShibing, Guizhou180–220130Resistant to leaf spot and viral diseases; wide adaptability.aN10Zheshen ⅢNingde Normal University200125High-yielding and excellent quality, though with a higher incidence of mosaic disease and leaf spot.[22]N11Zheshen ⅠZherong County Agricultural Technology Extension Center, Fujian Province135118Strong cold hardiness but moderately susceptible to leaf spot; poor heat and waterlogging tolerance.[23]N12AnguoAnguo, Hebei210150Tolerant to high-density planting; high yield.aN13Guishen ⅠInstitute of Chinese Materia Medica and Ethnomedicine Resources, Guizhou University of Traditional Chinese Medicine120–150150Extremely strong resistance to viral diseases.[24]Note: “–“ denotes no test data available; “a“ denotes data derived from original field surveys; no formal published literature is currently available. Key information of the studied varieties. Note: “–“ denotes no test data available; “a“ denotes data derived from original field surveys; no formal published literature is currently available. Experimental design A field experiment was conducted from April to November 2024, using a randomized complete block design with three replications (Figure 1). Each variety was randomly assigned to a single-row plot. The planting density was maintained at 90,000 plants per 667 m², with 8 cm spacing between plants and 6 cm between rows. The experimental plot area was 4 m² (0.4 × 10 m), and the soil was loess. Plants were grown in a greenhouse under natural light conditions (natural photoperiod). Temperature control was achieved through programmed side-wall ventilation to ensure the internal environment reflected ambient seasonal fluctuations while preventing extreme heat stress. Overview of the experimental layout illustrating the randomized complete block design used for cultivating different P. heterophylla varieties. Measurement of electrical information (1)Primary data collection (1)Primary data collection Primary data collection Fresh, fully expanded leaves of P. heterophylla were collected at the peak growth stage (June 21) for electrical measurements. For each variety, nine independent biological replicates were performed (n = 9), with three representative plants sampled from each of the three randomized plots. To minimize the influence of leaf age on electrical parameters, the 5th fully expanded leaf, counting from the base upward, was consistently selected across all varieties. Immediately after sampling, the leaves were immersed in water for 30 minutes. 6 The capacitance (C), resistance (R), capacitive reactance (Xc), inductive reactance (XL), and impedance (Z) of the leaves were measured using a high-precision LCR tester (Model LCR-6300; Good Will Instrument Co., Ltd., Taiwan, China). The measurements were conducted at a test voltage of 1.5 V and a frequency of 3.0 kHz, under a series of clamping forces (1.139, 2.149, 3.178, 4.212, and 5.245 N). The configuration of the measurement device, as illustrated in Figure 2, follows the established model described by Zhang et al. 11 The selection of clamping forces was based on preliminary experiments and previous optimizations. 11 This specific range (1.139–5.245 N) ensures a reliable exponential fit between the electrical parameters and force while remaining below the “crushing point” of P. heterophylla leaves, thus preventing irreversible tissue damage during measurement. Diagram of the LCR meter test setup and schematic of a parallel-plate capacitor. Note: In the figure: 1—support bracket, 2—foam board, 3—electrode plate, 4—conductor wire, 5—iron block, 6—plastic rod, 7—mounting bracket. (2)Intrinsic parameter modeling (2)Intrinsic parameter modeling Intrinsic parameter modeling Calculations were conducted according to following the method described by Zhang et al. 13 The intrinsic electrical parameters—intrinsic resistance (IR), intrinsic impedance (IZ), intrinsic inductive reactance (IXL), and intrinsic capacitive reactance (IXc)—were derived by extrapolating the fitted relationship between each measured parameter and the clamping force (F) to F = 0, representing the natural, undisturbed state of the leaf, where e⁰ = 1. The calculations were performed as follows: (1) IR,IZ,IXL or IXc=y+k. The intrinsic capacitance (IC) was subsequently calculated from IXc using the following formula: (2) IC=12πfIXc. In this formula, π = 3.1416 and f represents the measurement frequency. (3)Derived physiological indices (3)Derived physiological indices Derived physiological indices Based on these intrinsic electrical parameters, a series of derived physiological indices was calculated to evaluate plant water status, nutrient transport efficiency, and metabolic activity. The specific calculation formula is shown in Table 2. Table 2.Calculation formulas for physiological indices.CategoryPhysiological indexFormulaPhysiological significanceWater statusIntracellular water holding capacity (IWHC) IWHC=(IC)3 (3)Water retention capacity indicates growth potential.Intracellular water utilization efficiency (IWUE) IWUE=dIWHC (4)Efficiency of metabolic water use; stress adaptation.Intracellular water holding time (IWHT) IWHT=IC×IZ (5)Water retention duration; drought adaptation strategy.Nutrient transportWater or nutrient transfer rate (WTR or NTR) WTR or NTR=IWHCIWHT (6)Transfer speed: supports vigorous metabolism.Nutrient unit relative flux (NUF) NUF=IRIXc+IRIXL (7)Nutrient flux per unit area; poor-environment compensation.Nutrient active transport flow (UAF) UAF=IXcIXL (8)Active nutrient transport capacity per unit membrane area.Nutrient transport capacity (NTC) NTC=UNF×NTR (9)Overall nutrient transport, absorption, and allocation capacity.Nutrient active transfer capacity (NAC) NAC=UAF×NTR (10)Overall active nutrient transport capability and life vitality.Metabolic activityMetabolic flow (MF) MF=1IR+IZ+IXc+IXL (11)Smoothness of metabolic processes based on intrinsic electrical properties.Metabolic rate (MR) MR=NTR×NAC (12)Rate of metabolic processes driven by nutrient transport capacity.Metabolic activity (MA) MA=MF×MR6 (13)Comprehensive indicator of overall plant metabolic vigor and health status.Note: The sixth root in the MA formula is utilized as a geometric mean to integrate six sub-indicators of metabolic vigor, ensuring the resulting index is dimensionally normalized and reflects the synergistic effect of water and nutrient transport efficiency. Calculation formulas for physiological indices. Note: The sixth root in the MA formula is utilized as a geometric mean to integrate six sub-indicators of metabolic vigor, ensuring the resulting index is dimensionally normalized and reflects the synergistic effect of water and nutrient transport efficiency. It should be noted that these derived physiological indices (such as IWHC, NUF, and MA) represent relative capabilities, efficiencies, and fluxes rather than direct physical measurements. Through the mathematical operations of ratios and geometric means, the original physical units of the electrical parameters cancel out or are dimensionally normalized. Consequently, these derived indices are dimensionless quantities and are expressed as pure numbers without unit designations in the subsequent text and figures. The exponential model and the derived physiological indices used in this study follow the theoretical framework established by our research group. The applicability of these models and indices specifically for P. heterophylla has been rigorously verified in our previously published work. 6 It has been confirmed that higher values of IC, intracellular water metabolism, nutrient transport, and plant metabolic activity, coupled with lower values of IR, IZ, IXL, and IXc, are significantly positively correlated with the biomass, agronomic traits, and root development and quality of P. heterophylla. This provides a robust physiological and mathematical foundation for utilizing these electrical indicators as non-destructive predictors for the comprehensive evaluation and screening of the 13 varieties in this study. Measurement of agronomic traits and yield indicators A total of three replicate rows were assigned per variety. From each replicate, three individual plants of P. heterophylla were randomly sampled for the measurement of agronomic traits. The measured parameters included the number of leaves and branches, leaf dimensions (length and width), aboveground and below-ground characteristics (stem length, number of nodes, and biomass), as well as root and cotyledon attributes (weight, number, and diameter). For each variety, all tuberous roots within three separate 50 × 50 cm areas were harvested. The fresh and dry weights of the tuberous roots (the primary economic and edible part) were recorded for yield evaluation. After harvesting, the roots were cleaned, and surface water was removed to determine the fresh weight. For dry weight determination, the samples were dried in a forced-air oven at 60 °C for 8 hours until a constant weight was achieved. These weights were subsequently converted to yield per 667 m2. A comprehensive list of all evaluated traits, along with their corresponding measurement methods, is provided in Table 3. Table 3.Agronomic traits of P. heterophylla and their corresponding measurement methods.Agronomic traitsMeasurement methodsNumber of main stem branchesCount of branches originating from the base of the root crown.Number of primary branchesCount of first-order branches on the main stem.Number of leavesTotal number of fully expanded leaves per plant.Number of underground stem nodesCount of nodes from the stem base to the soil surface.Number of aboveground stem nodesCount of nodes from the soil surface to the stem apex.Number of daughter rootsTotal number of indeterminate roots, including those from underground stems and the root crown.Number of expanded daughter rootsNumber of tuberous roots, i.e., adventitious roots swollen into tubers.Total plant lengthDistance from the root tip to the stem apex (cm).Underground stem lengthLength from the stem base to the soil surface, including underdeveloped stems (cm).Aboveground stem lengthLength of the stem above the soil surface, including underdeveloped stems (cm).Leaf lengthDistance from base to tip of one of the four largest leaves in the upper “+” arrangement (cm).Leaf widthWidth at the broadest point of the same leaf measured for leaf length (cm).Diameter of expanded daughter rootsMean of the maximum diameters of three randomly selected tuberous roots (cm).Stem diameterDiameter of the first node on the most robust aboveground stem emerging from the root crown (cm).Weight of expanded daughter rootsTotal fresh weight of all tuberous roots (g).Total biomassTotal fresh weight of the whole plant (aboveground and belowground parts) (g).Underground biomassFresh weight of all belowground parts, including stems, roots, and tubers (g).Aboveground biomassFresh weight of all aboveground parts, including stems, leaves, and flowers (g). Agronomic traits of P. heterophylla and their corresponding measurement methods. Analysis of nutritional components The nutritional components of P. heterophylla tubers were analyzed following standardized protocols. The moisture content was determined using the oven-drying method, 25 while the ash content was measured with a muffle furnace. 26 The extract content was quantified through cold-water extraction. Following the method of ref. 27 , polysaccharides were isolated via hot-water extraction and ethanol precipitation, and their content was determined using the phenol-sulfuric acid method. Based on the method of ref. 28 , the saponin content was measured by vanillin-sulfuric acid colorimetry. All analyses were performed using nine independent biological replicates (n = 9). Grey relational analysis To evaluate the comprehensive performance of the 13 P. heterophylla varieties, GRA was employed following these steps:(1)Data normalization: raw data for yield, nutritional components, and electrical parameters were normalized using the “initial-value” or “optimal-value” processing method to eliminate dimensional effects.(2)Determination of reference sequence: an ideal reference sequence X 0 = {x 0(1), x 0(2),..., x 0(n)} was constructed using the maximum values of positive indicators (e.g., yield, polysaccharide content) and minimum values of negative indicators (e.g., intrinsic resistance).(3)Calculation of grey relational coefficient (ξ): the coefficient for the i-th variety at the k-th indicator was calculated as: ξi(k)=Δmin+ρΔmaxΔik+ρΔmax, where Δ ik = |x 0(k) − xi (k)|, and the resolution coefficient ρ was set to 0.5. (1)Data normalization: raw data for yield, nutritional components, and electrical parameters were normalized using the “initial-value” or “optimal-value” processing method to eliminate dimensional effects. Data normalization: raw data for yield, nutritional components, and electrical parameters were normalized using the “initial-value” or “optimal-value” processing method to eliminate dimensional effects. (2)Determination of reference sequence: an ideal reference sequence X 0 = {x 0(1), x 0(2),..., x 0(n)} was constructed using the maximum values of positive indicators (e.g., yield, polysaccharide content) and minimum values of negative indicators (e.g., intrinsic resistance). Determination of reference sequence: an ideal reference sequence X 0 = {x 0(1), x 0(2),..., x 0(n)} was constructed using the maximum values of positive indicators (e.g., yield, polysaccharide content) and minimum values of negative indicators (e.g., intrinsic resistance). (3)Calculation of grey relational coefficient (ξ): the coefficient for the i-th variety at the k-th indicator was calculated as: ξi(k)=Δmin+ρΔmaxΔik+ρΔmax, Calculation of grey relational coefficient (ξ): the coefficient for the i-th variety at the k-th indicator was calculated as: ξi(k)=Δmin+ρΔmaxΔik+ρΔmax, (4)Calculation of grey relational grade (G): The final grade was calculated by averaging the coefficients, where a higher G value indicates a variety closer to the ideal ideotype. (4)Calculation of grey relational grade (G): The final grade was calculated by averaging the coefficients, where a higher G value indicates a variety closer to the ideal ideotype. Calculation of grey relational grade (G): The final grade was calculated by averaging the coefficients, where a higher G value indicates a variety closer to the ideal ideotype. Statistical analysis For all statistical evaluations, a consistent significance threshold was pre-set at α = 0.05. Statistical comparisons among the 13 P. heterophylla varieties (regarded as the experimental treatments) were performed using one-way analysis of variance (ANOVA) via IBM SPSS Statistics 25.0. The analysis was based on three independent biological replicates per variety, corresponding to the three blocks in the randomized complete block design. Duncan's multiple range test was subsequently applied for post hoc comparisons to determine significant differences among the varieties (p < 0.05). Results and discussion Varietal differences in electrical parameters Significant varietal differences in the intrinsic electrical parameters among the 13 P. heterophylla varieties were observed during the peak growth stage (Table 4, p < 0.001). Specifically, variety N1 exhibited the lowest values of IR, IZ, IXL, and IXc, whereas variety N9 displayed the highest values for these parameters. In contrast, IC reached its maximum in variety N13 and its minimum in variety N9. Table 4.IR, IZ, IXL, IXc, and IC of different P. heterophylla varieties.NumberIR (MΩ)IZ (MΩ)IXL (MΩ)IXc (MΩ)IC (pF)N10.45 ± 0.12 g0.25 ± 0.06 c0.62 ± 0.17 e0.29 ± 0.08 c191.34 ± 44.57 aN21.61 ± 0.08 de0.62 ± 0.20 bc2.10 ± 0.78 cd0.72 ± 0.14 b75.84 ± 15.47 bcN33.52 ± 0.43 b0.59 ± 0.26 bc3.67 ± 1.03 b0.68 ± 0.20 b83.14 ± 28.61 bcN40.77 ± 0.45 fg0.36 ± 0.12 bc1.02 ± 0.26 de0.41 ± 0.13 bc139.22 ± 46.85 abN52.78 ± 0.55 c0.74 ± 0.10 b3.06 ± 0.38 bc0.77 ± 0.11 b70.31 ± 10.62 bcN62.21 ± 0.33 cd0.59 ± 0.10 bc2.06 ± 0.73 cd0.63 ± 0.14 bc86.15 ± 17.33 bcN71.09 ± 0.60 efg0.61 ± 0.22 bc1.57 ± 0.67 de0.75 ± 0.22 b75.33 ± 25.45 bcN81.50 ± 0.10 def0.43 ± 0.23 bc1.81 ± 0.08 d0.49 ± 0.18 bc119.46 ± 50.46 bN95.36 ± 0.53 a1.37 ± 0.36 a5.88 ± 0.67 a1.49 ± 0.34 a37.12 ± 9.18 cN102.47 ± 0.24 c0.61 ± 0.24 bc2.22 ± 1.22 cd0.68 ± 0.17 b81.26 ± 20.07 bcN111.69 ± 0.27 de0.63 ± 0.17 bc2.05 ± 0.57 cd0.70 ± 0.18 b79.13 ± 19.72 bcN121.66 ± 0.20 de0.55 ± 0.33 bc1.69 ± 0.10 de0.60 ± 0.26 bc98.34 ± 35.50 bcN130.81 ± 0.75 fg0.25 ± 0.11 c1.00 ± 0.16 de0.29 ± 0.11 c207.62 ± 102.12 a F value31.3135.26114.1657.7764.342 p value<0.001<0.001<0.001<0.001<0.001IR, IZ, IXL, IXc, and IC stand for intrinsic resistance, intrinsic impedance, intrinsic inductive reactance, intrinsic capacitive reactance, and intrinsic capacitance, respectively. Data show the mean ± standard deviation of three biological replicates (n = 3), and different lowercase letters indicate statistically significant differences (p <0.05). IR, IZ, IXL, IXc, and IC of different P. heterophylla varieties. IR, IZ, IXL, IXc, and IC stand for intrinsic resistance, intrinsic impedance, intrinsic inductive reactance, intrinsic capacitive reactance, and intrinsic capacitance, respectively. Data show the mean ± standard deviation of three biological replicates (n = 3), and different lowercase letters indicate statistically significant differences (p <0.05). Plant electrical parameters serve as robust indicators of physiological activity and metabolic status. 29 It must be recognized that the intrinsic parameters obtained when F = 0 can reflect the overall physiological state and active metabolic activity of the plant in its natural environment. These intrinsic biophysical traits are inherently coupled with the plant's capacity for biomass accumulation and secondary metabolite synthesis. Generally, lower resistance, impedance, and capacitive reactance, coupled with higher capacitance, are associated with more vigorous metabolic activity. Conversely, plants subjected to stress conditions—such as successive cropping or nutritional imbalance—typically exhibit elevated IR, IZ, IXL, and Ixc values, along with reduced IC. 30 The higher IC indicates a greater capacity for charge storage and structural integrity, which serves as a physiological buffer against environmental fluctuations. The elevated IC suggests enhanced cell membrane stability and higher vacuolar water storage, providing a material basis for efficient intracellular transport. The highest IC value of N13 aligns closely with its status as the dominant cultivar in Guizhou Province, providing additional evidence for the potential association between electrical characteristics and environmental adaptability. It is essential to acknowledge that the electrical parameters measured in this study, such as IR and IZ, function as indirect physical proxies rather than direct measurements of internal biochemical processes. These parameters are derived from the physical response of leaf tissues to an external electric field, which is modulated by the plant's cellular structure and ionic environment. Specifically, IC is inherently coupled with cell membrane integrity and vacuolar storage, while IR reflects the concentration and mobility of intracellular electrolytes. Therefore, while these indicators do not directly track specific metabolic pathways, they serve as reliable surrogate indicators that manifest the integrated physiological status and metabolic vigor of the plant in its natural state. The significant varietal differences observed in this study confirm that electrical profiles are strongly influenced by the specific genotypic background of each variety. While this study did not directly analyze the genome, the distinct intrinsic electrical parameters likely reflect the physiological manifestation of genotypic variation. This has a robust physiological basis; For instance, the heterogeneity of H+-ATPase—a master protein responsible for generating transmembrane ion gradients—could lead to varietal-specific differences in leaf resistance and impedance. 31 , 32 Similar genotypic influences on physiological phenotypes have been documented in other crops, supporting the use of EP indicators as proxies for evaluating varietal fitness. These findings highlight the importance of varietal selection for optimizing cultivation in Guizhou, where intrinsic electrical parameters provide a reliable, nondestructive measure of plant vitality and adaptability. Significant varietal differences in intracellular water metabolism, nutrient transport, and metabolic activity were observed among the 13 P. heterophylla varieties during the vigorous growth stage (Figure 3). Variety N9 exhibited the lowest IWHC and WTR, whereas varieties N13 and N1 displayed significantly higher IWHC and WTR than N9 (p < 0.05). In contrast, the trend for IWUE was the opposite. As shown in Figure 4, the UAF in the leaf cells of N7 was significantly higher (p < 0.05) than that of N3, while the UNF showed an inverse trend. The NAC and NTC of N9 were significantly lower (p < 0.05) than those of N13. Furthermore, the leaf MR and MA of N13 and N1 were both significantly greater than those of N9 (p < 0.05; Figure 5). The intracellular water utilization parameters of leaves in different varieties of P. heterophylla. (A) IWHC (intracellular water holding capacity), (B) IWUE (intracellular water utilization efficiency), (C) IWHT (intracellular water holding time), and (D) WTR or NTR (water or nutrient transfer rate). The nutrient transport parameters of leaves in different varieties of P. heterophylla. (A) The active transport flow of nutrients (UAF), (B) nutrient unit relative flux (UNF), (C) nutrient active transfer capacity (NAC), and (D) nutrient transport capacity (NTC). The metabolic activity of leaves in different varieties of P. heterophylla. (A) Metabolic activity (MA), (B) metabolic rate (MR). Previous studies have demonstrated that IWHC and IWUE accurately reflect the life strategies and diversity of intracellular water metabolism in plant leaves. A higher IWHC indicates more efficient plant water utilization, whereas under low IWHC conditions, plants often enhance IWUE to maintain physiological activity and growth. For instance, Zhang et al. 33 reported that potatoes with higher tissue water content exhibit greater IWHC, WTR, and overall physiological vigor, whereas peppers with lower water content sustain growth by improving IWUE. Moreover, the UNF and NAC serve as reliable indicators of nutrient transport strategies, providing an effective means for real-time monitoring of plant nutrient status. Concurrently, the high MA value underscores a high-frequency ion exchange and enzymatic turnover, which is significantly associated with the synthesis of bioactive compounds like polysaccharides and saponins. We hypothesize that the vigorous ionic environment reflected by MA may provide an optimized physiological basis for the enzymatic pathways involved in biosynthesis. Since electrical parameters are driven by ion fluxes across membranes, the high MA value underscores a high-frequency ion exchange and enzymatic turnover, which directly fuels the synthesis of bioactive compounds like polysaccharides and saponins. Our results confirm that varietal differences significantly influence intracellular water metabolism, nutrient transport, and metabolic activity in P. heterophylla. This finding highlights the pivotal role of varietal selection in ensuring physiological adaptability and long-term cultivation success under the unique ecological conditions of Guizhou Province. Histograms and error bars represent the mean and standard deviation of the nine biological replicates (n = 9), respectively. Different lowercase letters indicate statistically significant differences according to Duncan's test (p < 0.05). Histograms and error bars represent the mean and standard deviation of the nine biological replicates (n = 9), respectively. Different lowercase letters indicate statistically significant differences according to Duncan's test (p < 0.05). Histograms and error bars represent the mean and standard deviation of the nine biological replicates (n = 9), respectively. Different lowercase letters indicate statistically significant differences according to Duncan's test (p < 0.05). Comparison of agronomic traits among varieties Significant differences (p < 0.05) were observed in the agronomic traits of both above-ground (Figure 6) and below-ground (Figure 7) parts among the 13 varieties of P. heterophylla. Above-ground traits varied considerably. Varieties N11 and N13 exhibited significantly greater leaf length, whereas N1 and N10 had the shortest leaves. N13 exhibited the maximum leaf width, in contrast to N1, which had the smallest. The leaf number was highest in N12 and lowest in N9. The number of main stems ranged from 1 to 4, with N12 possessing a single stem, which was significantly lower than that of other varieties. The number of primary branches varied from 1 to 5, with N6 possessing significantly more branches than the others. Above-ground stem length ranged from 5.50 cm (N5) to 11.50 cm (N2). The average stem diameter was 0.23 cm, with N5 having a notably larger diameter of 0.42 cm. The number of above-ground stem nodes ranged from 2 to 6, with N5 having relatively fewer nodes. Notably, N4 achieved the highest above-ground biomass of 13.72 ± 1.70 g. Significance analysis of agronomic traits in the aboveground parts of different varieties of P. heterophylla. (A) Leaf length, (B) leaf width, (C) number of blades, (D) number of main stems, (E) number of primary branches, (F) ground stem length, (G) stem diameter, (H) number of stem nodes above ground, and (I) above-ground biomass. Significance analysis of agronomic traits in the underground parts of different varieties of P. heterophylla. (A) Whole plant length, (B) number of underground stem nodes, (C) rhizome length, (D) number of roots, (E) number of enlarged roots, (F) enlarged root diameter, (G) enlarged root weight, (H) underground biomass, and (I) total biomass. Data are expressed as the mean, with error bars denoting the standard deviation from nine independent biological replicates (n = 9). Different lowercase letters indicate significant differences (Duncan's test, p < 0.05). Below-ground traits also showed pronounced variation. Total plant length ranged from 11.60 cm to 26.10 cm, with N13 being the tallest. The number of below-ground stem nodes varied between 1 and 3, with N2 having fewer nodes. Below-ground stem length ranged from 2.00 cm (N2) to 12.90 cm (N6). N8 produced the highest number of lateral roots, while N9 had the fewest. For tuberous roots, N11 had the highest number, and N9 again had the fewest. The diameter of these roots ranged from 0.50 cm to 0.95 cm, with N5 being significantly larger than the other varieties. The fresh weight of these roots varied from 0.41 g to 1.50 g, with N12 being significantly lighter. Below-ground biomass and total biomass differed significantly among varieties; N4 exhibited the highest values (9.87 ± 0.24 g and 23.59 ± 1.64 g, respectively), significantly exceeding others. Bar graphs and error bars indicate the mean and standard deviation of the nine biological replicates (n = 9), respectively. Bars not sharing a common lowercase letter are significantly different according to Duncan's test (p < 0.05). The agronomic traits of P. heterophylla are determined by both genetic and environmental factors. Under stable environmental conditions, genetic factors predominantly govern variety-specific developmental characteristics, such as leaf size and branching pattern, which ultimately influence yield and quality. 34 Larger leaves and more branches typically indicate more vigorous plants, which are often associated with enhanced photosynthetic efficiency and nutrient uptake capacity. 35 The yield of P. heterophylla is directly governed by the developmental status of its tuberous roots. The swelling of roots is a complex process driven by the accumulation of storage compounds, such as polysaccharides. 36 The combination of a high number of swollen lateral roots and a large root diameter is a key determinant of tuber yield. In conclusion, the interplay between above-ground architecture (source strength) and below-ground tuber development (sink strength) dictates the final yield and quality. Yield variation among varieties The yield performance of the 13 varieties, measured in both fresh and dry weight per 667 m², revealed significant differences (Figure 8). Regarding fresh weight yield, variety N11 demonstrated superior performance, achieving 675.74 ± 6.51 kg/667 m², which was significantly higher than all other varieties (p < 0.05). In contrast, varieties N10 and N1 yielded significantly lower (p < 0.05), at 221.44 ± 18.96 kg/667 m² and 236.30 ± 12.36 kg/667 m², respectively. A similar trend was observed for dry weight yield. Varieties N11, N5, and N2 constituted the high-yielding group, with yields of 267.41 ± 9.32, 239.71 ± 34.14, and 237.58 ± 35.57 kg/667 m², respectively, showing no significant difference among themselves while remaining significantly higher than the remaining varieties (p < 0.05). Consistent with the fresh weight results, N10 and N1 again formed the low-yielding group, with dry weights of 73.09 ± 21.79 kg/667 m² and 86.79 ± 5.09 kg/667 m², significantly lower than the others (p < 0.05). It is important to note that while the sampling protocols ensured statistical reliability, the yield figures calculated here represent estimated production potentials under the specific ecological conditions of the experiment. These values provide a comparative framework for screening superior genotypes rather than absolute commercial yield forecasts. Differences in yield among different varieties of P. heterophylla. Data show the mean ± standard deviation of nine biological replicates (n = 9), and different lowercase letters indicate statistically significant differences according to the Duncan test (p < 0.05). Grey relational analysis between yield and agronomic traits GRA was conducted to quantify the relationship between agronomic traits and the dry weight yield of P. heterophylla. As shown in Table 5, the correlation degrees were ranked, elucidating the five traits that exert the most substantial influence on yield: expanded cotyledon root diameter > leaf length > stem diameter > leaf width > aboveground stem length. Table 5.Correlation analysis of agronomic traits and dry weight yield.ItemGrey relational gradeRankItemGrey relational gradeRankDiameter of expanded daughter roots0.80441Number of underground stem nodes0.716510Leaf length0.79882Total biomass0.715011Stem diameter0.79533Expanded root weight0.714412Leaf width0.78534Number of aboveground stem nodes0.705413Aboveground stem length0.74585Aboveground biomass0.703014Whole plant length0.73716Underground stem length0.677715Number of daughter roots0.72897Number of primary branches0.676616Number of expanded daughter roots0.72598Number of leaf blades0.656417Belowground biomass0.71679Number of main stem branches0.630218 Correlation analysis of agronomic traits and dry weight yield. In grey system theory, the magnitude of the grey relational coefficient directly reflects the degree of association between a comparative sequence (agronomic trait) and a reference sequence (dry weight yield). 37 A higher correlation weight for an indicator signifies its stronger intuitive and effective influence on the final yield. Our results echo the strong trait-yield association reported in quinoa, 38 revealing a parallel relationship in P. heterophylla. For instance, a larger diameter of the tuberous roots indicates enhanced accumulation of storage compounds such as starch and polysaccharides, directly contributing to single root weight and overall yield. Similarly, adequate leaf dimensions (length and width) are crucial for sustaining photosynthetic capacity, which in turn fuels tuber expansion. The integrated interpretation of these traits reveals an optimal plant ideotype for high-yielding P. heterophylla: a cultivar with large, efficient leaves, a strong stem, and a high potential for root swelling. From a breeding perspective, these traits should be prioritized as key selection criteria. Nutritional composition of different varieties Significant variations in key nutritional quality indicators were observed among the 13 varieties of P. heterophylla (Table 6). The average moisture content across all varieties was 62.23%, with variety N5 exhibiting the lowest value (p <0.05). The mean ash content was 2.97%, and N13 showed a significantly lower ash content than the others (p <0.05). The extract content averaged 38.55% with N2 being significantly higher than the other varieties (p <0.05). Table 6.Nutritional composition of different P. heterophylla varieties.NumberMoisture content (%)Ash content (%)Extract content (%)Polysaccharide content (%)Saponin content (%)N163.25 ± 1.90 abc3.30 ± 0.02 a39.98 ± 0.69 b47.75 ± 2.13 ab2.76 ± 0.40 abN262.09 ± 1.34 bcd2.80 ± 0.02 gh41.86 ± 0.62 a45.24 ± 0.26 bcd2.68 ± 0.21 abN362.57 ± 1.97 ab3.02 ± 0.03 cd37.52 ± 0.46 d49.27 ± 1.65 a2.56 ± 0.32 abcN464.50 ± 2.43 d3.28 ± 0.05 a39.04 ± 0.14 bc42.77 ± 1.41 d2.83 ± 0.23 aN559.45 ± 0.27 cd2.87 ± 0.02 fg38.59 ± 0.65 c48.67 ± 1.21 ab2.10 ± 0.28 cdN660.79 ± 1.39 a3.04 ± 0.01 c35.93 ± 0.50 e48.29 ± 1.94 ab1.80 ± 0.11 dN766.22 ± 1.52 cd3.14 ± 0.03 b36.62 ± 0.48 de46.39 ± 2.40 abc2.35 ± 0.25 abcN860.88 ± 1.39 bcd2.96 ± 0.05 de37.52 ± 0.27 d45.61 ± 1.06 bcd2.55 ± 0.21 abcN962.17 ± 1.68 ab2.92 ± 0.02 ef39.97 ± 0.20 b43.11 ± 1.48 cd2.66 ± 0.12 abN1065.44 ± 3.12 cd2.98 ± 0.03 cde39.42 ± 0.19 bc49.24 ± 2.13 a2.23 ± 0.56 bcdN1160.42 ± 1.51 cd2.95 ± 0.05 de36.34 ± 0.49 e45.89 ± 0.99 abcd2.69 ± 0.19 abN1260.36 ± 2.06 cd2.74 ± 0.04 hi39.04 ± 0.34 bc47.14 ± 0.22 ab2.26 ± 0.16 bcdN1360.88 ± 1.57 cd2.67 ± 0.06 i39.38 ± 0.41 bc45.79 ± 0.79 abcd2.46 ± 0.15 abc F value3.98553.48228.1173.9043.514 p value0.002 < 0.001 < 0.0010.0020.004Values represent the mean ± SD of three biological replicates (n = 3), and different lowercase letters denote significant differences (Duncan's test, p <0.05). Nutritional composition of different P. heterophylla varieties. Values represent the mean ± SD of three biological replicates (n = 3), and different lowercase letters denote significant differences (Duncan's test, p <0.05). For the primary bioactive compounds, polysaccharides and saponins exhibited particularly notable variation. Polysaccharide content ranged from 42.77% to 49.27%, with N3 achieving the highest concentration (49.27 ± 1.65%), followed by N10 (49.24 ± 2.13%) and N5 (48.67 ± 1.21%). Saponin content varied from 1.80% to 2.83%, with N4 showing the highest level (2.83 ± 0.23%), substantially exceeding the mean value of 2.46%. Extract content, indicative of the total water-soluble components. Polysaccharides and saponins are the main bioactive constituents responsible, and their concentrations directly determine its quality. 39 These quantitative data provide a critical reference for both the functional food industry and breeding programs. For food manufacturers seeking standardized raw material sourcing, N3 represents an optimal choice for high-polysaccharide applications such as dietary fiber supplements or prebiotic formulations, while N4 offers superior saponin content for products targeting antioxidant functionality. For breeding programs, these compositional profiles serve as essential phenotypic data for selecting parental materials with desirable quality traits. The pronounced variation observed across varieties underscores the importance of variety selection as a determinant of final product quality and as a foundational resource for genetic improvement. Integrated quality assessment and relationship with electrical parameters In this study, the integrated data represent a continuum from micro-scale biophysical signals to macro-scale phenotypic outcomes. At the biophysical level, intrinsic electrical parameters such as IC and MA serve as the functional foundation, reflecting cellular membrane integrity, ionic mobility, and the real-time “metabolic engine” of the plant. These biophysical states drive the biochemical level, where nutrient transport efficiency and enzymatic activity—manifested through MA—directly fuel the synthesis and accumulation of bioactive polysaccharides and saponins. Ultimately, these internal processes culminate at the phenotypic level, manifesting as morphological development (e.g., leaf dimensions and stem diameter) and economic yield (tuber biomass). By employing GRA, we successfully bridged these distinct biological scales into a unified relational grade. The derivation of the final rankings involved the systematic integration of 36 performance indicators, including intrinsic electrical parameters, agronomic traits, yield data, and nutritional components. Each variety was compared against an artificially constructed “ideal ideotype,” which consisted of the best-performing values for each indicator across the entire test group. A higher relational grade signifies a closer approximation to this ideal reference and indicates superior comprehensive performance. 40 The results showed that the grey relational grade ranged from 0.9282 to 0.9821 (Figure 9). N13 achieved the highest grey relational grade (0.9821), followed by N4 (0.9746), N1 (0.9688), N8 (0.9627), N12 (0.9591), and N7 (0.9550). These six varieties demonstrated superior and well-balanced profiles combined with robust physiological vigor, making them the most suitable candidates for both functional food applications and breeding programs. Comprehensive evaluation analysis of different varieties of P. heterophylla. The ecological adaptability of P. heterophylla is encoded in its cellular response to environmental stressors. Guizhou's typical “low sunlight and high humidity” climate requires plants to have high nutrient transfer efficiency to compensate for limited photosynthetic rates. The superior NAC and NTC values observed in local varieties like N13 and N8 represent a specialized physiological strategy to optimize resource acquisition in these specific agro-ecological conditions. In contrast, introduced varieties (e.g., N9) exhibited significantly higher IR and IZ, indicating higher internal resistance and metabolic energy loss when grown outside their optimal range. Notably, N13, which is currently the dominant cultivated variety in Guizhou Province, exhibited the highest intrinsic capacitance (IC) and metabolic activity (MA) values among all varieties (Table 4, Figure 5). This alignment between electrical parameters and comprehensive nutritional quality is particularly significant. It suggests that parameters measurable non-destructively using leaf-based electrical profiling may serve as effective surrogate indicators of overall nutritional quality. N13 achieved the highest comprehensive evaluation score in this study. This variety is currently one of the most widely cultivated varieties in the region. Its high ranking further confirms its reliable performance and superior adaptability to local growing conditions. Likewise, N8—a locally bred cultivar—also ranked among the top performers, demonstrating the success of regional breeding efforts in developing varieties suited to specific agro-ecological conditions. The strong alignment between experimental outcomes and actual cultivation practices validates the robustness of the evaluation system employed in this study. Furthermore, the remaining elite varieties—N4, N1, N12, and N7—exhibited well-balanced traits in terms of electrical activity, agronomic performance, yield, and quality, indicating their high potential for widespread promotion and adoption in similar ecological regions. This study serves as a foundational germplasm evaluation and screening for regional adaptation. The high-performing varieties identified (N4, N1, N12, and N7) provide validated parental materials for future breeding. A limitation of the current study is the absence of direct genetic analysis, which precludes definitive conclusions regarding the broad genetic diversity of the collection. Future research will build upon the elite varieties identified here (e.g., N13 and N4) to conduct multi-omics analyses (genomics and transcriptomics). Future research will build upon the elite varieties identified here (e.g., N13 and N4) to conduct multi-omics analyses (genomics and transcriptomics), aiming to decipher the genetic mechanisms underlying high yield and superior polysaccharide accumulation. Conclusions This study integrated a comprehensive evaluation of agronomic, yield, and nutritional traits with plant electrical technology to comprehensively evaluate thirteen P. heterophylla varieties. Among all measured nutritional components, with polysaccharide content ranging from 42.77% to 49.27% and saponin content from 1.80% to 2.83%. These quantitative data provide a fundamental reference for the standardized procurement of raw materials in the functional food industry. Regarding electrical parameters, specifically IC and MA, a strong consistency was demonstrated with the overall quality rankings. This observed consistency suggests that leaf-based electrical measurement could potentially serve as a rapid and non-destructive tool, suitable for both real-time quality assessment during food processing and early germplasm screening in breeding programs. By integrating all parameters through GRA, six superior cultivars were identified: N13, N4, N1, N8, N12, and N7. In summary, this research provided preliminary quantitative data to support the standardized sourcing of raw materials for functional foods. The identified superior cultivars, possessing diverse quality traits, hold potential for application in industrial production or breeding initiatives tailored for the Guizhou region. Furthermore, the potential utility of electrical technology as an innovative non-destructive method for quality evaluation and germplasm prioritization has been demonstrated under specific experimental conditions. These findings may offer helpful indicators for the real-time quality control of functional food resources, potentially contributing to the sustainable development of the P. heterophylla industry and regional economic growth.