Maternal and paternal contributions to seed and seedling traits in Ginkgo Biloba hybrids

Abstract

Background

Ginkgo biloba L. is a dioecious plant with a life history of billions of years, and hybridization is one of its main forms of generation replacement in nature. The traditional hybridization research of ginkgo is relatively weak, and previous work has ignored the genetic distance between parental selection, relying solely on subjective human selection, making it difficult to maximize the utilization of genetic variation in ginkgo hybridization.

Result

The results indicate that maternal genotype has a significant impact on seed morphogenesis and accumulation of toxic components; The influence of the paternal parent is stronger in the inheritance of resistance related substances; The F₁ generation showed a super parental advantage in photosynthetic parameters, and growth and development were significantly positively correlated with photosynthetic efficiency.

Conclusions

This study reveals the differentiation patterns in maternal vs. paternal inheritance in ginkgo hybridization, providing physiological bases for the selection of medicinal varieties, development of stress resistant strains, and optimization of photosynthetic traits. It has important practical significance for improving the industrial utilization efficiency of ginkgo resources.

Supplementary information

The online version contains supplementary material available at 10.1186/s12870-025-07278-z.

Introduction

Natural hybridization in plants represents a widespread biological phenomenon, initially conceptualized by Linnaeus and Kölreuter during the 1760 s [1], with systematic investigation of plant hybridization commencing in the mid-18th century [2]. Subsequent research primarily focused on hybrid discovery and phenotypic characterization [3, 4]. The earliest documented hybrid verification dates to 1716 involving maize and Cucurbita species [5]. Forest tree hybridization refers to controlled mating between distinct species or provenances, achievable through natural processes or artificial intervention. As a pivotal strategy for woody plant genetic improvement, this approach facilitates trait optimization via parental genetic recombination. However, the prolonged generational cycles inherent to tree species impose significant challenges in refining parental selection strategies and advancing the analysis of hereditary effects.

As a dioecious Mesozoic relict species, ginkgo has persisted through eons of natural hybridization in the wild. This protracted hybridization process, inherently involving chromosomal recombination and segregation, generates substantial genetic variation. But fossil evidence demonstrates remarkable morphological stasis in this species spanning over 200 million years [6]. The genetic principles governing hybridization balance species stability through inheritance while driving biodiversity via variation. As a single species in a single family and single genus, the genetic variation of intraspecific hybridization in ginkgo is particularly valuable. Therefore, it is particularly important to investigate the macroscopic genetic differences and hybridization effects in the hybridization process of ginkgo, based on the fact that the hybridization process has been going on for billions of years without any obvious morphological changes. China currently conserves over 90% of global ginkgo resources [7], yet artificial hybridization research remains underdeveloped compared to model tree species like poplars and pines. Natural hybridization in ginkgo predominantly occurs between closely related individuals, with interspecific crosses being evolutionarily constrained. Although producing abundant pollen, the species’ brief flowering period and frequent phenological asynchrony necessitate artificial pollination to ensure mating precision and achieve controlled hybridization. In past research on ginkgo hybridization, the issue of genetic diversity of germplasm was generally neglected, and there was a lack of purpose and theoretical basis for the selection of parents. The so-called ‘’good varieties’’ were artificially defined and did not have good adaptability and genetic characteristics. The detection of differences in hybrid progeny is limited to the traditional phenotypic growth and physiological and biochemical levels, and fails to penetrate into the genetic level, ignoring the influence of hybrid parents on the progeny and the effect of parental hybridization. Due to the long growth period and the lack of modern biotechnology, the traditional hybrid research has not been able to achieve a breakthrough.

In the Chinese national standard GB/T 20 397–2006, the seeds are divided into five variety groups based on their shape and ratio of nut length to width (RLW): the ‘Changzi’ variety group (RLW ≥ 1.7), the ‘Fozhi’ variety group (1.5 ≤ RLW < 1.7), the ‘Maling’ variety group (1.2 ≤ RLW < 1.5), the ‘Meihe’ variety group (1.2 ≤ RLW < 1.5), and the ‘Yuanzi’ variety group (RLW < 1.2). Subsequently, the ‘Maling’ and ‘Meihe’ variety group were merged and collectively referred to as ‘Zhongzi’. Ginkgo seeds possess significant nutritional and medicinal value, containing diverse bioactive compounds with demonstrated antioxidant properties, antitumor activity, free radical scavenging capacity, and therapeutic effects on cardiovascular and cerebrovascular diseases [8, 9]. Despite these pharmacological benefits, their utilization is constrained by inherent toxic components. Studies indicate that ginkgo toxins can induce adverse reactions in humans, including intermittent convulsions, emesis, generalized seizures, and even fatal intoxications [10]. The neurotoxic 4’-O-methylpyridoxine (MPN) was first isolated from ginkgo seeds by Japanese researchers in 1985 and identified as the primary toxic agent [11]. Subsequent investigations revealed additional toxic derivatives: Scott detected 4’-O-methylpyridoxine-5’-glucoside (MPNG) [12], while Kobayashi confirmed its presence in ginkgo kernels and demonstrated its reduced toxicity compared to neurotoxic 4’-O-methylpyridoxine (MPN) [13]. The collective term “ginkgotoxin” now encompasses both MPN and MPNG, reflecting their synergistic toxicological profile.

In this study, based on the genome-wide differences and genetic distances of the main cultivated germplasm of ginkgo [14], we constructed a 3 × 3 hybrid system of ginkgo, bagged, artificially controlled pollination, and collected and sown seed fruits to cultivate the F₁ generation population. We also took the seed harvested by artificial controlled hybridization as the object to explore the genetic differences of the seed and F₁ offspring populations under different hybrid combinations, and further analysed their hybrid effects and parental contributions to provide a certain theoretical basis for the hybrid selection of ginkgo for fruit use. This study makes up for the shortcomings of earlier studies in parental selection, and expects to provide a basis for parental selection to explore the hybrid effects and genetic variation of ginkgo. It also emphasizes that ginkgo seed traits and quality are crucial for the selection and breeding of ginkgo varieties for fruit use. This study provides a macroscopic understanding of the genetic variation of ginkgo hybridization, and lays the foundation for subsequent studies of genetic variation across the genome.

Materials and methods

Construction of hybridization system

Based on our previous analysis of the genetic structure and genetic diversity of ginkgo varieties for fruit use [14], we selected three copies of varieties (‘Changzi’, ‘Fozhi’ and ‘Zhongzi’) with the same average genetic distances and belonging to different species groups as hybrid females, and based on the growth characteristics of ginkgo males, we chose three male trees to collect pollen as hybrid fathers, and the hybrid combinations are shown in Table 1.

Table 1.

Hybrid combination

♂ ♀	1 (NLXY)	2 (GFSX)	3 (DHS)
A (Fozhi)	A1	A2	A3
B (Zhongzi)	B1	B2	B3
C (Changzi)	C2 *	C2	C3

* C1 and C2 species were inadvertently mixed during collection. The C1 and C2 populations are collectively recorded as C2.

From approximately 20 March 2018, daily dawn and dusk monitoring of male floral development was conducted. Pollen collection initiated when catkin-like staminate inflorescences exhibited yellowish pigmentation and released particulate matter upon gentle manipulation. On clear mornings before 09:00, inflorescences were excised and dehydrated on sulfuric acid-treated paper under ambient conditions, followed by sieve purification (150 μm mesh) and storage at 4 °C. Three clonal replicates per maternal genotype were established. The bagging time is determined by observing under a stereomicroscope when the opening angle of the female flower’s microplate scales is stable at more than 30 ° and pollination droplets have not yet been secreted, while ensuring that it is at least 48 h earlier than the target male plant’s pollen dispersal period. Controlled pollination was uniformly implemented following paternal pollen collection, with immediate rebagging maintained for five days post-pollination. Controlled hybrid seeds were harvested in mid-October 2018, with 50 seeds per cross reserved for experimental analyses. Remaining seeds underwent cold stratification (4 °C) for after-ripening and germination induction. Open-pollinated seeds from a control tree (CK) were processed concurrently under identical protocols. As parents, various ginkgo varieties are grafted onto rootstocks with relatively consistent growth conditions, with 3–5 rootstocks per germplasm. All of them are stored in the China Ginkgo Germplasm Base and identified by Professor Fuliang Cao (email: fuliangcaonjfu@163.com). The relevant experiments and sample collection were conducted with the permission of the local relevant departments.

All hybrid progeny and control population were divided into three plots, and each plot was sown with three replicated units. All external environmental factors were kept the same in daily management, including planting method, water and fertilizer, weeding, density, etc., to ensure that the seedling growth conditions were suitable. At the end of July 2021, all mature leaves of hybrid parents and 798 F₁ population individuals were collected separately, placed in sampling boxes, and then transferred to the laboratory for storage in a −20 °C freezer for subsequent measurement of experimental indicators.

General information about the trial site

The hybridization site and the F₁ population seeding site were located within the China Ginkgo Germplasm Base, as summarized in Table 2.

Table 2.

Overview of the test site

Item	General Information
Geographic location	34°39’ N, 118°03’ E
Climate type	Semi-humid warm temperate monsoon climate
Average annual temperature	13.9 °C
Extremely high temperature	42.6 °C
Extremely low temperature	−13.1 °C
Annual rainfall	893 mm
Soil type	damp soil, brown loam

Methods

Determination of morphological and physiological and biochemical indices of seeds

After removing fruit stalks from hybrid seeds, the fresh weight of each hybrid combination was determined using an analytical balance with 50 replicates per measurement. Vernier calipers were employed to measure the longitudinal and transverse diameters of seed kernels, with the seed shape index calculated as the ratio of longitudinal to transverse diameters. Ash content was quantified following the food testing methodology [15]. The dual-wavelength method [16] was applied to determine both amylose and amylopectin contents in seed kernels, with total starch content calculated by summation. The glutinous characteristics were evaluated based on the amylose-to-amylopectin ratio. Ginkgotoxin levels in ginkgo fruit were analyzed using High Performance Liquid Chromatography (HPLC) [10, 17].

Growth and photosynthesis measurement of hybrid F₁ population

​The photosynthetic parameters of parental hybrids and two-year-old F₁ progeny were measured using a CIRAS-3 photosynthesis system (PP Systems, USA). During late June mornings under clear, windless conditions, ten median-vigor plants were randomly selected from each F₁ population. Four mature leaves per plant were analyzed to determine photosynthetic parameters including: net photosynthetic rate (P_n), transpiration rate (T_r), stomatal conductance (G_s), and intercellular CO₂ concentration (C_i). Parental hybrids underwent identical measurements for comparative analysis. Light response characteristics were quantified between 08:30 − 11:30 AM under stable meteorological conditions using the CIRAS-3 leaf chamber with integrated LED light source (spectral composition: 90% red, 5% blue, 5% white). Three randomly selected plants per F₁ population were analyzed under controlled leaf chamber temperature (25 °C) with photosynthetically active radiation (PAR) gradients descending sequentially from 2000 to 0 µmol·m^-2·s^-1 (14-step protocol: 2000, 1700, 1400, 1200, 1000, 800, 600, 400, 300, 200, 150, 100, 50, 0). During the determination process, the atmospheric carbon dioxide concentration was controlled at 400 ± 20 µmol∙mol^-1, the relative Humidity was stabilized at 60± 4%, and the temperature of the leaf chamber was kept consistent with the surrounding ambient temperature. For each photosynthetically active radiation value, the duration of a single measurement was set to 3 min, and the relevant photosynthetic parameters were automatically recorded by the instrument. For each specific light response curve (LRC), the study used the modified rectangular hyperbolic model [18] for fitting, and its expression is as follows: P_n = α × (1 − β × I)/(1 + γ × I) × I – R_d. Here, P_n represents the net photosynthetic rate, α is the apparent quantum yield, β and γ are modified coefficients, R_d refers to the dark respiration rate, and I is the photosynthetically active radiation. The four related photosynthetic parameters, including photosynthetic capacity (P_nmax), dark respiration rate (R_d), were obtained by accessing http://photosynthetic.sinaapp.com.

Physiological and biochemical measurements of hybrid parents and F₁ population

The soluble sugar content in ginkgo leaves was determined using the anthrone colorimetric method [19, 20]. Chlorophyll extraction was performed using 95% (v/v) ethanol (C₂H₅OH), and its concentration was quantified through spectrophotometric analysis using a UVmini-1240 spectrophotometer (Shimadzu, Japan). Soluble protein content was measured via the Coomassie Brilliant Blue G-250 binding method [19]. Total flavonol glycosides were calculated according to the Chinese Pharmacopoeia protocol: Total flavonol glycoside content = (Quercetin content + Kaempferol content + Isorhamnetin content) × 2.51. Use HPLC to determine the content of ginkgo terpenoid lactones.

Data processing

All indicators were subjected to 3 technical replicates and organized using Microsoft Excel 2021, followed by statistical analysis using SPSS 24.0. Before conducting any analysis of variance, the normality of distribution and homogeneity of variance of the data were tested. The drawing was completed by Origin 2021 and R (R.4.2.1).

Result

Matrilineal genetic dominance in the seedling stage

Maternal specificity of RLW and fresh weight

Comparative analysis of ginkgo seeds phenotypic traits between controlled pollination and open pollinated populations revealed maternally predominant inheritance patterns for both RLW and fresh weight, with paternal contributions lacking statistical significance (Fig. 1a). In controlled crosses with maternal parent A, the mean RLW of A1, A2, and A3 populations were 1.45 ± 0.06, 1.49 ± 0.08, and 1.44 ± 0.06 respectively, marginally higher than open pollinated counterparts (1.43 ± 0.07). Maternal B exhibited identical mean RLW between controlled and open-pollinated progenies. Notably, open-pollinated fruits from maternal parent C showed elevated mean RLW (1.81 ± 0.08) compared to controlled crosses C2 (1.73 ± 0.10) and C3 (1.68 ± 0.08), though statistically insignificant (P = 0.23). Significant inter-maternal variation was detected in RLW among controlled crosses (P = 0.036). Hybrids significantly surpassed both A and hybrids B, while intra-maternal comparisons across different paternal lines showed no significant differences (Table S1). Maternal-specific coefficients of variation (CV) were calculated as 4.38% (A), 4.41% (B), and 5.08% (C), indicating high maternal genetic stability in seed morphology transmission.

Fig. 1.

a RLW of different hybrid populations. b The fresh weight of seeds of different hybrid populations. A, B, and C represent seeds produced by free pollination in the same hybrid maternal parent

The overall trend of changes in seed fresh weight and RLW is consistent. Compared to free pollination, the average fresh weight of hybrid seeds with controlled hybridization did not show significant changes (Fig. 1b). The average fresh weight coefficient of A1(6.77 ± 0.75 g), A2(6.75 ± 0.79 g), and A3(6.39 ± 1.10 g) is slightly higher than the average fresh weight of freely pollinated seeds on the same mother plant (A: 6.29 ± 0.47 g). The average fresh weight of seeds from C as the mother plant (7.79 ± 0.78 g) is greater than that of hybrid seeds from A (6.55 ± 0.77 g) and B (6.32 ± 0.68 g).

There is no significant difference in the average fresh weight of hybrid seeds with the same maternal parent but different paternal parents. There was a significant difference (P = 0.021, Table S1) in the average fresh weight of the hybrid seeds of A (CV: 11.87%), B (CV: 10.77%), and C (CV: 9.96%) as maternal parents. These findings collectively demonstrate that phenotypic differentiation in both RLW and fresh weights is predominantly governed by maternal genotypes in ginkgo, with neither paternal contributions nor pollination methods significantly altering inheritance patterns.

Ash content and starch composition were negatively correlated with fruit type

Plant ash, primarily composed of mineral oxides, sulfates, and phosphates, critically determines adsorption capacities for ionic organic pollutants and heavy metals [21, 22]. The ginkgo seed coat exhibits exceptional ash characteristics due to its porous architecture, demonstrating potential as premium activated carbon feedstock. Our findings revealed no significant paternal effects on ash content across hybrid combinations (Fig. 2a), with maternal genotype emerging as the predominant factor governing ash variation. Hybrid seeds from maternal A and B showed comparable ash contents (A: 9.28%; B: 9.26%), while hybrids C exhibited significantly reduced ash levels (6.64%; P_C&A = 0.026, P_C&B = 0.035). Ash content displayed no correlation with fruit shape indices. A maternal-specific “strain effect” was observed, showing the highest seed shape index corresponded with the lowest ash content. Intraspecific ash variation remained smaller than interspecific differences across maternal genotypes (Fig. 2b).

Fig. 2.

a The ash content of seeds of different hybrid populations. b Relationship between ash and RLW in seeds of different hybrid populations. c The starch content of seeds of different hybrid populations. d The glutinousness of seeds of different hybrid populations. Different lowercase letters indicate significant differences at the P < 0.05 level

As the primary component of ginkgo endosperm, starch abundance paralleled staple crops like potato and maize [23]. Maternal genotypes significantly influenced total starch distribution: hybrids B as the maternal parent exhibited the highest mean content (68.96%), followed by C (65.82%), and A (64.13%). Notably, the B3 hybrid achieved 71.71% total starch, representing an 11.34% increase over the lowest-performing A1 hybrid (Fig. 2c). While total starch content showed no significant inter-group variation, amylopectin proportion displayed maternal specificity. Hybrids B contained the highest amylopectin percentage (79.67%), significantly exceeding those with A (68.30%) and C (66.26%) as maternal parents (P = 0.029). Glutinous texture is one of the important factors affecting the taste of ginkgo fruit endosperm as a food, and the glutinous texture of endosperm is mainly influenced by the composition ratio of different starch types [24]. The higher the proportion of amylopectin content, the stronger the glutinous nature exhibited by the endosperm [25]. Further analysis shows that the RLW is negatively correlated with endosperm viscosity: the smaller the RLW, the redder the viscosity indicator color (high proportion of amylopectin) of the white fruit, while the larger the RLW, the bluer the color (high proportion of amylose) (Fig. 2d). The smaller the fruit type, the more round the fruit, and the higher the viscosity. This discovery provides a phenotype marker basis for screening high glutinous food varieties through fruit type.

Strong association of ginkgo toxicity content with maternal

Ginkgotoxin was predominantly detected in ginkgo endosperms. No significant differences in total ginkgotoxin content were observed between open pollinated and controlled pollination hybrids derived from the same maternal plants (Fig. 3a). Maternal genotype significantly influenced toxin accumulation, with B hybrids exhibiting the highest mean content (419.36 µg·g^-1), significantly surpassing other maternal lines (P = 0.019). Open pollinated fruits from B reached peak ginkgotoxin levels (425.54 µg·g^-1). There was no significant difference in the average content of total ginkgo toxin in the hybrid seeds of A (378.95 µg·g^-1) and C (384.77 µg·g^-1).

Fig. 3.

The a ginkgotoxin content and b ratio of MPN and MPNG in seeds of different hybrid populations. Different lowercase letters indicate significant differences at the P < 0.05 level

MNP and MPNG exhibit significant differences in toxicity, with MNP being markedly more toxic than MPNG. However, MNP can be converted to MPNG under heating conditions, thereby reducing toxicity [26]. In hybrid fruit, ginkgotoxin primarily exists as MNP, accounting for an average of 91.51% of the total content (Fig. 3b). Among maternal lines, hybrids with C showed the highest MNP proportion (94.69%), while those with A (90.25%) and B (90.38%) as maternal parents maintained consistent MNP ratios. No statistically significant differences in the MNP/MPNG ratio were observed across hybrid combinations (P = 0.133).

Patterns of inheritance of growth and photosynthesis in the F₁ generation

Within-population height variation in seedling height diameter

Through the measurement and analysis of growth indicators in three-year-old F₁ progeny seedlings, consistent patterns of genetic variation were observed across different traits (Table S2). The population mean height was 79.78 ± 18.21 cm, with an inter-population range of 29.32 cm. The CK (93.75 ± 15.39 cm) was significantly taller than all other progeny (P < 0.05), while no significant differences were detected between the A2 population and the B1 population (Fig. 4a). The coefficient of CV among hybrid combinations was 11.49%, but the average CV within progeny populations increased to 25.02% (total variation: 27.86%), indicating that height differences primarily originated from intra-family individual variation. Similar patterns emerged for ground diameter growth.

Fig. 4.

The growth of F₁ populations: a Seedling heigh b Ground diameter c Shoot growth. d Coefficient of variation of seedling height, ground diameter, and shoot growth

The population mean diameter was 10.66 ± 1.30 mm, with an inter-population range of 4.47 mm. The CK exhibited the largest diameter (13.16 ± 2.35 mm), significantly exceeding other progeny, while the A2 population showed the smallest diameter (8.70 ± 1.26 mm) (Fig. 4b). The CV among hybrid combinations was 12.22%, whereas the average CV within progeny populations reached 24.83% (total variation: 26.10%).

The third-year shoot growth averaged 23.40 ± 8.01 cm, with an inter-population range of 9.19 cm. The B2 population displayed the longest shoots (26.51 ± 5.81 cm), while the A1 (17.88 ± 4.26 cm) and A2 (17.32 ± 3.94 cm) populations were significantly shorter than others progeny (Fig. 4c). The CV among hybrid combinations was 13.80%, but the average CV within progeny populations rose to 31.80% (total variation: 34.25%, Fig. 4d). These results collectively demonstrate that genetic variation in seedling height, ground diameter, and shoot growth of ginkgo hybrids is predominantly governed by intra-family genetic differences.

Further analysis of parental genetic effects revealed no significant differences in seedling height, ground diameter, or shoot growth among progeny derived from the same maternal or paternal parent through ANOVA (Fig. 5). Overlapping coefficients of variation across progeny populations and strong concordance of normal distribution curves demonstrated that parental genetic backgrounds exerted no significant influence on progeny phenotypic divergence in growth traits.

Fig. 5.

Growth differences in offspring of crosses from the same maternal/paternal parent. a, c, and e are the offspring statistics of the same female parent/male parent for seedling height, ground diameter, and shoot growth; b, d, and f are the corresponding normal distribution diagrams. ♀A, ♀B, ♀C, ♂1, ♂2, ♂3 represent all offspring of the same maternal parent or male parent, respectively

Superparental dominance and maternal bias in photosynthetic parameters

As the primary energy source for plant growth, photosynthetic intensity reflects developmental capacity. Analysis of F₁ progeny photosynthetic parameters revealed significant transgressive heterosis across all measured indices. For net photosynthetic rate (P_n), progeny populations exhibited a mean of 1.13 ± 0.21 µmol CO₂·m^-2·s^-1, significantly surpassing the parental averages (maternal: 1.33 ± 0.11; paternal: 0.88 ± 0.12 µmol CO₂·m^-2·s^-1). The B2 population achieved peak P_n (4.32 ± 0.21 µmol CO₂·m^-2·s^-1), while A2 (2.51 ± 0.14) and C2 (2.53 ± 0.19) showed significantly lower values than other progeny (range: 1.81 µmol CO₂·m^-2·s^-1; P < 0.05). Positive correlations emerged between P_n and growth vigor: the B2 population demonstrated maximum shoot growth, contrasting with the minimal growth in A2. No significant P_n differences were observed between controlled and open-pollinated progeny (Fig. 6a).

Fig. 6.

Differences in photosynthetic parameters in F₁ populations. The dotted lines in the figure represent the light sum parameters of the parents. a Net photosynthetic rate; b Transpiration rate; c Stomatal conductance; d Intercellular CO₂ concentration. Different lowercase letters indicate significant differences at the P < 0.05 level

Transpiration facilitates water potential gradient formation in plants, enabling efficient transport/absorption of water, minerals, and organic compounds while maintaining thermal homeostasis [27]. Progeny populations exhibited a mean T_r of 10.68 ± 1.57 mmol H₂O·m^-2·s^-1, significantly exceeding parental values (maternal: 11.50 ± 1.69; paternal: 9.86 ± 0.91 mmol H₂O·m^-2·s^-1). The B2 population demonstrated peak T_r (43.33 ± 3.52 mmol H₂O·m^-2·s^-1), significantly surpassing other progeny (P = 0.032), while A1 showed the lowest values (14.53 ± 1.58 mmol H₂O·m^-2·s^-1), yielding an inter-population range of 28.80 mmol H₂O·m^-2·s^-1 (Fig. 6b).

Under normal conditions, C_i and net photosynthetic rate show a positive correlation, where higher C_i corresponds to increased photosynthetic rates [28]. Analysis of G_s and C_i further revealed genetic differences: progeny G_s means (B2: 1.20 ± 0.19 mmol·m^-2·s^-1; A2: 1.11 ± 0.12 mmol·m^-2·s^-1; A1: 0.93 ± 0.14 mmol·m^-2·s^-1) were significantly higher than parental values, with the A3 population showing the lowest G_s (0.43 ± 0.08 mmol·m^-2·s^-1, Fig. 6c). Progeny C_i means (110.07 ± 19.21 µmol·mol^-1) exhibited notable transgressive heterosis compared to parents (maternal: 125.01 ± 12.94 µmol·mol^-1; paternal: 95.12 ± 25.26 µmol·mol^-1). The CK displayed the highest C_i (303.56 ± 21.29 µmol·mol^-1, P = 0.043), while C2 had the lowest progeny C_i (157.64 ± 11.69 µmol·mol^-1), yet still exceeded that of the maternal parent A (Fig. 6d).

To elucidate parent-progeny inheritance patterns of photosynthetic capacity across hybrid combinations, comparative analysis of parental and filial generation means was conducted, with genetic deviation degrees calculated for each parameter (Table S3). Genetic transmission analysis revealed that progeny photosynthetic parameters consistently exceeded mid-parental values, with all parent-progeny deviation indices exhibiting positive values. G_s showed the highest deviation. Transgressive heterosis rates reached 100% (excluding minor progeny groups) for P_n, T_r, G_s, and C_i, with corresponding Genetic Transmission Ability (GTA) averages of 310.73%, 264.27%, 242.92%, and 193.62%.

It should be noted that parental photosynthetic parameters were measured during mature, stable growth stages, whereas progeny data were collected from juvenile, vigorous growth phases. This developmental stage discrepancy may lead to overestimation of transgressive heterosis, thus requiring cautious interpretation of direct parental genetic contributions.

Differential genetic pathways for chlorophyll a/b

Chlorophyll, the primary photosynthetic pigment complex reflecting green light in photoautotrophic plants, comprises chlorophyll a, chlorophyll b, and carotenoids. As solar energy harvesters in chloroplast-containing leaves, chlorophyll content critically determines photosynthetic energy conversion efficiency [29]. Total chlorophyll content in parental hybrids averaged 2.04 ± 0.21 mg·g^-1 FW (maternal: 2.06 ± 0.19; paternal: 2.02 ± 0.13 mg·g^-1 FW), while progeny populations showed significant increases (2.09 ± 0.21 mg·g^-1 FW, CV: 12.74%). The B2 population exhibited peak values (2.29 ± 0.19 mg·g^-1 FW), with B1-B3 progeny significantly surpassing others (P = 0.031, Fig. 7a).

Fig. 7.

The difference of chlorophyll content in F₁ population. The dotted lines in the figure represent the light sum parameters of the parents. a Total chlorophyll; b Chlorophyll a; c Chlorophyll b; d Carotenoids. Different lowercase letters indicate significant differences at the P < 0.05 level

Chlorophyll a followed analogous trends: parental mean = 0.97 ± 0.11 mg·g^-1 FW (maternal: 1.02 ± 0.21; paternal: 0.92 ± 0.13 mg·g^-1 FW), increasing to 1.03 ± 0.19 mg·g^-1 FW (CV: 17.53%) in progeny. B1 progeny peaked at 1.14 ± 0.14 mg·g^-1 FW, while A1 showed minima (0.87 ± 0.10 mg·g^-1 FW, P < 0.05, Fig. 7b). Chlorophyll b displayed distinct inheritance: parental mean = 0.47 ± 0.09 mg·g^-1 FW (paternal parents [0.49 ± 0.10] exceeding maternal [0.44 ± 0.07] mg·g^-1 FW), with progeny marginally increasing to 0.50 ± 0.03 mg·g^-1 FW (CV: 49.82%). C3 progeny achieved maxima (0.52 ± 0.04 mg·g^-1 FW), significantly outperforming C2 and A1 (Fig. 7c).

Carotenoid analysis revealed parental mean = 0.295 ± 0.02 mg·g^-1 FW (maternal: 0.30 ± 0.03; paternal: 0.295 ± 0.13 mg·g^-1 FW), rising to 0.30 ± 0.05 mg·g^-1 FW (CV: 11.63%) in progeny. B1 progeny peaked (0.325 ± 0.03 mg·g^-1 FW), while C2 (0.28 ± 0.01 mg·g^-1 FW) and A2 showed significant reductions (Fig. 7d).

Genetic analysis revealed that progeny total chlorophyll and chlorophyll a exhibited mid-parent heterosis rates of 60.72% and 57.55%, respectively, with maternal values exceeding paternal levels—indicating maternal inheritance dominance. Carotenoids demonstrated pronounced heterosis (77.77%). Conversely, chlorophyll b showed paternal transmission tendencies (negative parent-progeny deviation indices). GTA values were ordered as follows: total chlorophyll (98.83%), chlorophyll a (114.15%), carotenoids (105.90%), and chlorophyll b (81.20%), confirming ​​component-specific regulatory mechanisms​​ in chloroplast-related trait inheritance (Table S4).

Genetic transmission of key traits

Soluble sugar

Soluble sugars play critical roles in plant development, serving as both energy sources and signaling molecules. As mobile carbon reservoirs derived from photosynthesis, they maintain osmotic balance and regulate processes including seed germination, early seedling development, and hormonal modulation [30, 31].

Parental hybrids exhibited lower average soluble sugar content compared to progeny populations (Fig. 8a). With the exception of paternal line 1, whose soluble sugar levels exceeded those of progeny groups A1 and CK, all parental values were lower than progeny levels. Parental hybrids showed a mean soluble sugar content of ​​6.36 ± 0.79 mg·g^-1​​, with maternal parents averaging ​​6.03 ± 0.58 mg·g^-1​​—lower than paternal parents (​​6.70 ± 0.84 mg·g^-1​​). Progeny populations demonstrated a significantly higher mean soluble sugar content of ​​8.35 ± 0.98 mg·g^-1​​. The B2 progeny group exhibited the highest levels (​​9.41 ± 0.74 mg·g^-1​​), significantly surpassing the lowest groups A1 (​​6.97 ± 0.39 mg·g^-1​​) and CK (​​7.01 ± 0.41 mg·g^-1​​) (P = 0.044). The average CV for progeny soluble sugar content was ​​19.26%​​.

Fig. 8.

The a soluble sugar content, b soluble protein content, c total flavonoid content, d terpene lactones content of F₁ population. Different lowercase letters indicate significant differences at the P < 0.05 level

Through genetic analysis of the soluble sugar content of the hybrid offspring population, it was found that 90.60% of the hybrid offspring had a soluble sugar content exceeding the mid parent value, demonstrating significant hybrid vigor (Table S5). Moreover, the parent-child deviation of soluble sugar content in the offspring is positive, with an average of 10.81, indicating that the inheritance of soluble sugar content in ginkgo has a significant bias, and the paternal parent has a greater contribution in hybrid inheritance. The average genetic transmission capacity of soluble sugar content in ginkgo is 135.59%.

Soluble protein

Soluble proteins, critical for osmotic regulation and fruitrient storage in plants, enhance stress resistance by stabilizing intracellular biomembrane structures [32, 33]. This study revealed that paternal soluble protein content (0.83 ± 0.05 mg·g^-1) systematically exceeded maternal levels (0.69 ± 0.04 mg·g^-1), yielding a differential of 0.14 mg·g^-1. Progeny groups B3 (1.01 ± 0.12 mg·g^-1), CK (0.98 ± 0.07 mg·g^-1), B1 (0.96 ± 0.08 mg·g^-1), and B2 (0.89 ± 0.06 mg·g^-1) exhibited significantly higher content than other groups (P = 0.038). The B3 group demonstrated peak values alongside the highest CV (26.57%), while progeny-wide CV averaged 21.82% (Fig. 8b). Genetic analysis indicated that ​​55.88%​​ of progeny exceeded mid-parental values, though no significant heterosis was observed. Parent-progeny deviation indices averaged 1.59​​, suggesting weaker paternal inheritance dominance compared to soluble sugar traits. Negative deviations in A3 and C3 groups (lowest absolute values) may originate from detection errors. The mean GTA of ​​109.98%​​ confirmed paternal genetic dominance, albeit with reduced effect magnitude relative to other metabolites (Table S5).

Total flavonoids

As the core medicinal component, ginkgo flavonoids exhibit multiple pharmacological activities including microcirculation improvement, neuroprotection, and antitumor effects [32], constituting key economic-value metabolites. This study demonstrated that paternal total flavonoid content (​​13.21 ± 0.15 mg·g^-1​​) was significantly lower than maternal means (​​14.38 ± 0.16 mg·g^-1​​) (Fig. 8c). Progeny populations showed a ​​28.99% increase​​ in mean content (​​17.79 ± 1.12 mg·g^-1​​). The CK achieved peak values (​​21.31 ± 1.98 mg·g^-1​​, P = 0.041), while A1 (​​14.34 ± 1.05 mg·g^-1​​), A2 (​​15.27 ± 1.67 mg·g^-1​​), and A3 (​​15.97 ± 1.40 mg·g^-1​​) formed low-content clusters, with only A1 falling below maternal lines B (​​15.12 mg·g^-1​​) and C (​​14.44 mg·g^-1​​).

Genetic feature analysis shows, Progeny flavonoids exhibited a mean CV of ​​23.16%​​, peaking in CK (​​29.95%​​) and reaching minima in A1 (​​17.53%​​) (Table S5). ​​77.12%​​ of progeny surpassed mid-parental values, with a positive parent-progeny deviation index of ​​+7.80​​, significantly exceeding soluble protein metrics—confirming ​​maternal inheritance dominance​​ in flavonoid accumulation. The GTA reached ​​127.53%​​, highlighting the significant enhancement effect of hybrid breeding on flavonoid content.

Terpenoid lactones

As unique bioactive constituents of ginkgo, terpene lactones act as potent platelet-activating factor antagonists with neuroprotective effects, exhibiting highly species-specific biosynthetic pathways in the plant kingdom [34, 35]. This study analyzed the content of terpenoids in the hybrid parent and offspring populations and found that the average content of terpenoids in the offspring population (11.27 ± 1.76 mg·g^-1) was significantly higher than that of the parents (9.86 ± 0.99 mg·g^-1), with the maternal parents (10.33 ± 1.09 mg·g^-1) systematically higher than the paternal parents (9.39 ± 0.15 mg·g^-1) (Fig. 8d). The highest A3 content was found in the offspring population (12.62 ± 1.50 mg·g^-1), which showed significant advantages over the lowest A2 (10.71 ± 1.01 mg·g^-1), C2 (10.66 ± 0.94 mg·g^-1), and C3 (10.28 ± 0.97 mg·g^-1) populations (P < 0.05). Notably, C3 progeny content, though lower than maternal lines A (10.54 ± 0.15 mg·g^-1​​) and C (10.27 ± 0.89 mg·g^-1​​), still exceeded paternal means.

Genetic variation analysis showed that the coefficient of variation of terpenoid lactone content in the offspring population reached 28.81% (range 24.19%−34.69%), which is similar to the variation pattern of flavonoid content (Table S5). 65.22% of offspring have individual content exceeding the median parental value, and the deviation between parents and offspring is positive (mean 2.90). The significant increase in maternal values compared to paternal values suggests the possible influence of maternal inheritance. Genetic transmission analysis showed that the average transmission efficiency of terpenoid lactone content was 114.93%, further confirming its potential for super parental inheritance.

The correlation between the growth and development of hybrid F₁ seedlings and physiological and biochemical indicators

Plant growth and development constitute a holistic, systematic, and coordinated process. Correlation analysis of growth parameters (seedling height, ground diameter, shoot growth) and physiological-biochemical indices across F₁ progeny populations revealed significant positive associations between ginkgo growth metrics and total chlorophyll content/net photosynthetic rate (P_n) (Fig. 9). Shoot growth demonstrated a ​​highly significant positive correlation​​ with P_n (P < 0.01). Notably, no significant correlation between the main contents of ginkgo leaves, indicating that ginkgo growth status under standard environmental conditions exerts minimal influence on the production and accumulation of foliar secondary metabolites.

Fig. 9.

Correlation heat map between growth and physiological and biochemical indicators. * Indicates a significant correlation at P < 0.05, ** indicates a significant correlation at P < 0.01

Discussion

Morphological characters of ginkgo seeds show significant discriminatory value in different genetic backgrounds, and the polysaccharides and toxic components contained in them as medicinal and food resources have become a focus of research [36]. The present study reveals that genetic variation in this species during hybridization shows a matrilineal dominant pattern, in contrast to the paternal plastid DNA inheritance commonly observed in angiosperms [37]. The phenomenon of Segregation distortion (SD), in which parental alleles are unequally divided in the progeny during hybridization-progressive penetrance, is widespread in nature. Segregation distortion has not been noticed since Mendel’s discovery of the law of segregation. Segregation distortion was first reported in the genetic linkage between gametophyte factors and starch endosperm in maize [38]. Ginkgo has long been in the ‘’natural hybridization field’’ of free pollination, and there was no significant difference in RLW and fresh weights of the same mother plant compared with that of a single parent. This shows that different parents do not affect the genetic variation of RLW and fresh fruit weight, and the genetic regulation of the two is mainly determined by the mother, which is consistent with the genetic pattern of wheat and other species [39]. It is worth noting that the nonlinear relationship between the development of the cavity at the end of the seed stalk and fresh weight suggests that the morphological characteristics cannot be used as a single basis for weight assessment, and the biological function of this structure needs to be analyzed in depth.

Physiological compositional analysis showed that ash content showed significant differences between mothers (mean value 8.55%, range of variation 14%), and its intra-strain variation was less than inter-strain variation, further verifying the dominance of maternal inheritance. Although the total starch content was not statistically different among hybrid combinations, i.e., there was no obvious superiority or inferiority difference with the varieties [40], the branched-chain starch percentage (71.88%) showed maternal specificity, in which the sub-sweet varieties with smaller RLW showed a significantly higher proportion of branched-chain starch, and the proportion of branched-chain starch to the total starch in the smaller and more rounded white fruits with smaller and more round RLW was higher, and the higher the glutinosity was, which was consistent with the people’s preference for This is consistent with people’s tendency to consume small and round white fruits. Toxicity analysis showed that the total ginkgotoxin content followed the same matrilineal pattern of inheritance, while the MPN/MPNG ratio remained stable among varieties. In view of the structural similarity between MPN and vitamin B6 and their antagonistic effects [41], combined with the presence of multiple toxicity factors such as hydrocyanic acid [42] and sensitizing proteins [43], the establishment of a standardized detoxification process that integrates traditional concoctions and modern biotechnology has become the key to industrial development.

The generation transmission process of ginkgo hybridization generates a large amount of genetic variation, which is ultimately reflected in the differences in growth and development, physiological and biochemical indicators of the offspring population. The growth and development analysis of hybrid offspring revealed significant intra strain variation characteristics: the intra strain variation coefficients of seedling height (25.02% vs. 11.49%), ground diameter (24.83% vs. 12.22%), and new shoot growth (31.80% vs. 13.80%) were significantly higher than the inter strain differences. The analysis of photosynthetic parameters found that the genetic transmission of net photosynthetic rate in offspring was as high as 310.73%, while the genetic transmission of total chlorophyll content was only 98.83%, indicating that non pigment factors such as stomatal conductance or enzyme activity may dominate the differences in photosynthetic efficiency [44, 45]. There is a significant positive correlation between the content of soluble sugars and soluble proteins in plants and their resistance [46, 47]. There was no significant difference in soluble sugar and soluble protein content between the hybrid parents and offspring of ginkgo, and existing studies [48] have shown that male ginkgo plants have higher resistance than female plants. Research on resistance related substances shows that the content of soluble sugars (135.95%) and proteins (109.98%) exhibits a paternal genetic tendency, consistent with the positive impact of paternal resistance on offspring in species such as jujube [49], indicating the special value of male plants in stress resistant breeding (Fig. 10). Research suggests that cytoplasmic inheritance in gymnosperms is generally dominated by patrilineal inheritance [50], and some scholars have proposed a trend of “parental inheritance-patrilineal inheritance dominant, matrilineal inheritance secondary-matrilineal inheritance dominant, patrilineal inheritance secondary” in genetic evolution in higher plants [51]. The cytoplasmic inheritance of plastids and mitochondria in ginkgo hybrid studies is mostly characterized by maternal inheritance, as is the case with Cycas revoluta [50]. It indicates that ginkgo has undergone hundreds of millions of years of evolution, and its genetic system has become relatively advanced and sophisticated.

Fig. 10.

Schematic diagram of the genetic transmission ability in the process of ginkgo hybirdization

Flavonoids and lactones are one of the main active components of ginkgo extracts, which are important for ginkgo selection breeding and industrial development. ginkgo flavonoid and lactone contents are affected by many factors, including fertilization, soil, temperature, light, rejuvenation, and UV-B irradiation [52–56]. The total flavonoid and terpene lactone contents of hybrid parents were overall smaller than those of hybrid offspring, as influenced by age factors. Unlike resistance substances, ginkgo total flavonoids and terpene lactones had a strong maternal inheritance effect in generation inheritance, i.e., hybrid parents contributed more to the accumulation of total flavonoids and terpene lactones in the offspring. Comparison of the content differences between the hybrid parents revealed that the total flavonoids and terpene lactones content was also higher in the female plants than in the male plants, again tentatively suggesting that the female plants have a higher capacity to accumulate total flavonoids and terpene lactones than the male plants. This genetic heterogeneity of metabolites provides a basis for directed breeding: strains with high photosynthetic efficiency should be preferred for garden applications, while medicinal development should focus on maternal flavonoid accumulation capacity. The study also highlights the need to establish a standardized sex identification system and expand the sample size of parents to accurately analyze the mechanism of metabolic differences between male and female.

Conclusion

This study analyzed the genetic laws of seed traits and physiological metabolism in ginkgo. Matrilineal inheritance plays a dominant role in the morphology and component accumulation of ginkgo, while paternal inheritance has no significant effect on RLW, fresh weight, and ash content. Maternal genotype affects the proportion of amylopectin and total toxic content of ginkgo; The RLW is significantly negatively correlated with the proportion of amylopectin, and the synergistic effect of multiple toxic factors highlights the urgency of standardized detoxification processes. The genetic variation within the F₁ generation is significant, with coefficients of variation for seedling height, ground diameter, and new shoot growth being 25.02%, 24.83%, and 34.25%, respectively, and no specific parental genetic tendency; The photosynthetic parameters exhibit super parental inheritance, and the net photosynthetic rate of offspring significantly exceeds the mid parental value and is consistently above the mid parental value; There are differences in the genetic patterns of chlorophyll components. Total chlorophyll, chlorophyll a, flavonoids, and terpenoids are significantly influenced by the maternal parent, while chlorophyll b, soluble sugars, and proteins show a paternal bias. This study provides a reference for the selection of parents in ginkgo directional breeding.

Abbreviations

RLW

Ratio of nut length to width (fruit type coefficients)

CV

Coefficients of variation

CK

Control group

MPN

Meurotoxic 4'-O-methylpyridoxine

MPNG

Scott detected 4'-o-methylpyridoxine-5'-glucoside

HPLC

High performance liquid chromatography

Pn

Net photosynthetic rate

Tr

Transpiration rate

Gs

Stomatal conductance

Ci

Intercellular CO₂ concentration

PAR

Photosynthetically active radiation

P_nmax

Maximum net photosynthetic rate

Rd

Dark respiration rate

GTA

Genetic transmission ability

Author contributions

QG and YH designed the experiment. XG, YH, PJ conducted material preparation and hybridization experiments. XG, PJ and YH analyzed the data and wrote the manuscript. PJ, FL, TZ and QG edited the manuscript.All authors reviewed the manuscript.

Funding

This research was funded by the National Natural Science Foundation of China (31971648), and the Talent Introduction Project Study of Nanjing Forestry University (GXL2018001) on Ginkgo biloba and other important tree germplasm resources, and the Postgraduate Research &Practice Innova-tion Program of Jiangsu Province (KYCX25_1340).

Data availability

The materials and dataset generated and/or analyzed during the study will provide from the corresponding author upon the request.

Declarations

I declare that this study is my original work and has not been submitted to any other institution anywhere for the award of any academic degree, diploma, or certificate. All sources of materials used for the study have been duly acknowledged.

Ethics approval and consent to participate

All plant materials collected in this study were in compliance with the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) and relevant Chinese laws and regulations. Although Ginkgo biloba L. is listed as a national first-class protected plant, the cultivated varieties used in this study meet the requirements of the protection regulations.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Clinical trial number

Not applicable.