Introduction

Xanthoceras sorbifolium Bunge, which belongs to the Sapindaceae family, is a traditional woody oil species found in northern China that has been praised as one of the most potential plant resources in the 21st century. The extracted oil has a high content of unsaturated fatty acids and nervonic acid, which can be used to produce edible oils and renewable and environmentally friendly biodiesel. Saponins extracted from it have anti-oxidative and antitumor effects, and its flavonoids are used to treat rheumatic heart disease, arthritis, and other diseases1. In addition, the plant exhibits robust windbreak and sand-fixing capacity, as well as ecological adaptability in sandy, barren, arid, and mountainous areas2. Therefore, increasing attention has been paid to this species in recent years.

However, a low seed yield is a bottleneck in the development of X. sorbifolium. A low female-to-male flower ratio is one of the main factors limiting yield. The flowers of X. sorbifolium can be divided into female and male flowers3. Typically, the terminal bud differentiated into flowers (mainly female flowers), shoots and leaves; lateral buds differentiated into flowers (mainly male flowers), shoots and leaves. Both male and female flowers are bisexual in the early stages of differentiation, but the pistils and stamens stop developing in the later stages, suggesting sex regulation. Therefore, increasing the ratio of female flowers on lateral inflorescences may be an effective way to improve seed yield.

Bud differentiation is a complex process influenced by both external (e.g., temperature, humidity, and solar radiation) and internal factors (e.g., gene expression, phytohormone levels, and nutrient status)4,5,6. N, P, K, soluble sugars, and starch are classical macronutrients required for plant development. High levels of mineral elements contribute to flower bud differentiation and sexual organ development7,8. Specifically, N and P positively affect pistil differentiation and ovule vitality, whereas high K levels contribute to the development of anthers and mastoids9,10. The content of nonstructural carbohydrates (soluble sugars and starch) changes with the development of sex organs. For instance, the least soluble sugar content is observed during flowering in both Camellia sinensis and Phalaenopsis aphrodite11,12. In addition, the soluble sugar content in the anther wall and pollen grains has been shown to increase gradually and peak at flowering in tomato Solanum lycopersicum13.

Source organs, including mature leaves and shoots, significantly affect bud differentiation and distribution14,15. Because of the varying nutrient contents in different sections of shoots, they usually differentiate into mixed, flower, or leaf buds. Nutrient levels significantly differ between mixed, flower, and leaf buds; between female and male flower buds; and at different bud locations. For example, the soluble sugar and starch contents in flower buds peaked at the physiological differentiation stage and were much higher than those in the leaf buds of Syringa oblata and Rhododendron moulmainense16,17. The accumulation of sucrose and soluble sugars in leaves is conducive to female flower bud differentiation in Morella rubra18. Reproductive buds often occur at the end of short shoots or the lateral positions of long shoots in Larix gmelinii19.

Nutrients have been found to have a crucial effect on bud differentiation in X. sorbifolium. High K, Mn, Cu, Fe, Mg, and Ca levels promote the differentiation and formation of male flowers20. The starch grain accumulation rate, fullness, and quantity of female flowers are significantly lower than those of male flowers from the mononuclear to dikaryotic stages1,21. Additionally, the sucrose content of female flowers increases with the development of pistils, whereas it does not in male flowers22. Therefore, nutrition affects not only bud differentiation but also sex differentiation.

Shoot position affects the distribution of different types of buds; in other words, different types of buds are distributed in distinct parts of the shoots. Nutrient levels differ in different parts of the shoot. However, the relationship between bud type and nutrient levels remains unclear. The present study aimed to determine the differentiation characteristics of buds on different parts of the shoot; explore the nutrient level differences among buds, shoots, and leaves on different parts of the shoot during bud differentiation; determine the nutrient differences leading to the differentiation of mixed and leaf buds and female and male flowers; and explore the effects of N supplementation on female and male differentiation. This study offers novel insights into nutritional regulation and ideas for improving the female-to-male flower ratio and seed yield of X. sorbifolium.

Materials and methods

Experiment site and plant materials

The experimental site is located in Chaoyang City, Liaoning Province, Northern China (120.45°E, 41.57°N). The area has a temperate continental monsoon climate. The annual maximum temperature is 39.8 °C in July, and the annual minimum temperature is -28.5 °C in January. The area receives approximately 2,698.2 h of annual sunshine and 456.7 mm of precipitation. Ninety well-grown, 6-year-old individuals from the same superclone were selected as the material. The trees were planted with a spacing of 3 m × 3 m. The average tree height was 1.19 m and the diameter at breast height ranges from 2.8 cm to 3.9 cm. Beijing Forestry University have permission to collect Xanthoceras sorbifolium.

Morphological observation and determination of nutrient levels

Bud differentiation into different parts of the shoots was observed in May 2020. Based on the bud differentiation results, shoots, leaves, and buds were divided into different parts (upper, middle, and lower) and sampled separately. The bud and shoot sampling intervals were 21 days from bud formation to dormancy (late May 2020 to November 10, 2020), 5 days from dormancy breaking to inflorescence elongation (February 21, 2021, to April 5, 2021), and 3 days until flowering (April 6, 2021, to May 5, 2021). Leaves were sampled between May 23 (leaf spreading period) and September 30 (deciduous period), 2020, at 21-day intervals. For the above-mentioned samples, collections are made randomly and evenly from the four directions (east, west, south, and north) of each tree, and these collections were mixed as a biological sample and subjected to three biological replicates, with each replicate 15 g. To avoid destructive sampling from affecting the nutrient distribution of the trees, a different experimental tree is used for each sampling.

The buds sampled were divided into two parts. One was used to observe external morphology, after which it was fixed in a formalin-aceto-alcohol solution for paraffin sectioning23. The other was washed thoroughly with the shoots and leaves and dried at 65 °C until constant weight, ground, and analyzed for nutrient element contents. Total N and P contents were determined using an automatic discontinuous analyzer (SmartChem® 450; AMS Alliance, Frepillon, France). K content was determined using flame photometry24. The soluble sugar content was determined from the ethanol supernatant, while the starch content was determined from the pellets of the hydroalcoholic extracts25.

Foliar N application

Based on our previous studies20,23 and the results of nutrient levels of the sampled buds, leaves, and shoots, N concentrations of 0.3, 0.5, and 0.8% were applied on April 5, 2021 (before sex differentiation). Clear water was used as a control (CK). Each sample was analyzed in triplicate. After spraying, the buds and shoots of the three parts were sampled for nutrient level determination every 3 days until flowering, and the numbers of female and male flowers and the proportion of female-to-male flowers on the terminal and lateral inflorescences were observed.

Statistical analyses

Prior to ANOVA, Shapiro-Wilk test was used to verify the normality of residuals, while homogeneity of variances was evaluated via Levene’s test. The statistical model included nitrogen application treatments and position as the fixed factor, with no random factors included in the one-way design. When ANOVA indicated significant differences (P < 0.05), multiple comparisons were conducted using the LSD (Least Significant Difference) test with a significance level set at α = 0.05. All statistical analyses were performed using R software (version 4.1.2; R Foundation for Statistical Computing, Vienna, Austria), with the `stats` package for conducting ANOVA and Shapiro-Wilk tests, the `car` package for Levene’s test, and the `agricolae` package for implementing the LSD test26.

Results

Section division of annual shoots and leaves based on bud distribution

Based on morphological observations, buds on the annual shoot can be divided into three types: upper (UBs), middle (MBs), and lower (LBs) buds (Fig. 1). Specifically, UBs are terminal buds that develop into female flowers, shoots, and leaves (Fig. 1A). MBs are lateral buds in the middle and upper sections of the shoots that develop into male flowers, shoots, and leaves (Fig. 1B). LBs are positioned between the first three nodes from the base of the annual shoot (Table S1). LBs wilt (LBW, Fig. 1C), develop into male flowers without a rachis (LBM, Fig. 1D), or sprout into leaves (LBL, Fig. 1E).

Based on the apparent positional effects of the three types of buds, the annual shoots were also divided into upper (USs), middle (MSs), and lower (LSs) shoots (Fig. 1F). Leaves sprouting from the corresponding location of the annual shoot were referred to as upper (ULs), middle (MLs), and lower (LLs) leaves (Fig. 1F).

Fig. 1
figure 1

Distribution and germination of buds on annual shoots of Xanthoceras sorbifolium. (A) The terminal bud on April 29, 2021. Female flower with three shoots germinated from the upper bud. (B) The lateral bud on May 2, 2021. Male inflorescence only or male inflorescence along with leaves emerged from the middle bud. C Non-germinating lower bud on May 2, 2021. (D) Male flowers without rachis from the lower bud on May 2, 2021. (E) Leaves emerged from the lower bud on May 4, 2021. (F) Section division of annual shoots based on the three types of bud location.

Full size image

External and internal morphology of different bud types

Previous studies have shown that the differentiation of male flowers (mainly from MBs) and female flowers (mainly from UBs) remained the same until sex organ abortion occurred in X. sorbifolium3,23 (Zhang, 2019; Zhou et al., 2019). Hence, using the onset of sex differentiation as the reference point, we compared the morphological differences between UBs and LBs before this stage (Fig. 2A) and between UBs and MBs following this stage (Fig. 2B).

The external morphology showed that UBs (Fig. 2A ②, ⑥, ⑦, and ⑧) were plumper with more bud scales than LBs (Fig. 2A ⑪, ⑫, ⑬ and ⑭). Moreover, UBs and MBs sprouted (Fig. 2A ⑨ and ⑩) earlier than LBs (Fig. 2A ⑮–⑰).

The process of UB differentiation was divided into nine stages: bud pre-differentiation (Fig. 2B ①, Stage 1), inflorescence primordium differentiation (Fig. 2B ②, Stage 2), sepal primordium differentiation (Fig. 2B ③, Stage 3), petal and stamen primordium formation and entry into dormancy (Fig. 2B ④ and ⑤, Stage 4), release from dormancy (Fig. 2B ⑥, Stage 5), pistil primordium differentiation (Fig. 2B ⑦, Stage 6), bisexual (Fig. 2B ⑧, Stage 7), sex differentiation (Fig. 2B ⑨, Stage 8), and flowering (Stage 9, after May 5, 2021).

The floral organs of LBM emerged later than those of UBs. The sepal primordium of LBM developed on March 8, 2021 (Fig. 2B ⑫), while that in UBs appeared in September of the previous year. When UBs were at the pistil primordium differentiation stage (Fig. 2B ⑦, Stage 6), LBM was at the petal primordium differentiation stage (Fig. 2B ⑬). Furthermore, another notable difference was that LBM directly differentiated into sepal primordium without inflorescence axis formation (Fig. 2B ⑫).

The difference between UBs and LBL became apparent at Stage 2 (Fig. 2B ② and ⑪). The flower primordium of UBs began to develop, and the growth cone became conical, whereas no change occurred in LBL, which had a larger leaf primordium than that of UBs. From stages 2 to 8, the leaf primordium developed fully (Fig. 2B ⑪, Stage 8) and sprouted (Fig. 2B ⑯). The morphological differences among LBL, LBM, and UBs appeared from the early differentiation stage till flowering. In LBW, no organ primordium appeared (Fig. 2B ⑰) during the whole bud development period.

At the sex differentiation stage (Stage 8), the ovary and ovule in UBs were fully developed and the pollens aborted, whereas in MBs and LBM, pollens were fully developed, and the ovary was atrophied (Fig. 2B ⑨, ⑩, ⑮).

Fig. 2
figure 2

External and internal morphology of various buds in Xanthoceras sorbifolium. External (A) and internal (B) morphology of various buds, respectively. A, anther; Gc, growth cone; Ip, inflorescence primordium; Br, bract primordium; Lp, leaf primordium (can be seen clearly); Se, sepal primordium; Pe, petal primordium; Sp, stamen primordium; Pp, pistil primordium; Pe, petal primordium; Ova, ovary; Ovu, ovules.

Full size image

Changes in mineral elements in buds, shoots, and leaves at different positions during bud differentiation

The N and P contents at the three locations on the annual shoots and buds showed similar trends (Fig. 3A, B), generally decreasing from top to bottom. At all stages, the N content of UBs was approximately 1.01–1.21-fold higher than that of MBs, except at Stage 7. The largest difference (1.21-fold) appeared at the flowering stage (Stage 9). The P content of UBs was always approximately 1.03–1.66-fold higher than that of MBs, except at Stage 2. The largest difference (1.66-fold) appeared at pistil primordium differentiation (Stage 6). In particular, at the sex differentiation stage (Stage 8), the P content in UBs and USs was significantly higher than that in MBs, LBs, MSs, and LSs. The N content in UBs and USs was also the highest among that in the three parts. High N and P levels could promote female functional differentiation of UBs. Mixed bud differentiation had a higher demand for N and P than did leaf-bud differentiation. At the stage where differences in mixed and leaf buds appeared (Stage 2), N and P contents among the three bud parts did not significantly differ, whereas those in USs were significantly higher than those in MSs and LSs.

The K content of MBs was consistently higher than that of UBs in stages 4–9 (Fig. 3C). From the bisexual to the sex differentiation stages (stages 7–8), the K content of the buds increased sharply, and that of MBs increased more than that of UBs. In the sex differentiation stage (Stage 8), the K content in MSs (11.4 mg/g) was the lowest, and that in MBs (22.5 mg/g) was the highest among the three parts. High K content might be beneficial for the development of male flowers. At the stage where the difference between mixed and leaf buds appeared (Stage 2), the K content among the three parts of the buds did not differ significantly, whereas that in LS was significantly lower than that in MSs and USs.

Fig. 3
figure 3

Dynamic changes in N, P, and K content in shoots, leaves, and buds at different stages during bud differentiation in Xanthoceras sorbifolium. N (A), P (B), and K (C) content in shoots, leaves, and buds. Different letters represent significant differences among different parts (p ≤ 0.05). The error bars indicate SEs, n = 3.

Full size image

Changes in non-structured carbohydrates in buds, shoots, and leaves of different positions during bud differentiation

The dynamic changes in soluble sugar and starch content in shoots, leaves, and buds at each stage of bud differentiation are shown in Fig. 4. At the stage where the difference between mixed and leaf buds appeared (Stage 2), the soluble sugar content of LBs was significantly higher than that of UBs and MBs. At the sex differentiation stage (Stage 8), the soluble sugar content in the UBs was significantly higher than that in the MBs, which was reversed in stages 2–7. The soluble sugar content in USs was significantly higher than that in MSs and LSs before dormancy (Stage 4), indicating that female flowers have a greater demand for soluble sugars than male flowers, as do those stored in the shoots (Fig. 4A).

No significant difference was observed in the starch content of shoots and buds among the three parts when differences in mixed and leaf buds appeared (Stage 2). The starch content of shoots gradually decreased from top to bottom at the sex differentiation stage (Fig. 4B, Stage 8). It was significantly higher in MBs (0.117 mg/g) and LBs (0.103 mg/g) than in UBs.

Fig. 4
figure 4

Dynamic changes in soluble sugar and starch content in shoots, leaves, and buds at different stages during bud differentiation in Xanthoceras sorbifolium. Soluble sugar (A) and starch content (B) in shoots, leaves, and buds. Different letters represent significant differences among different parts (p ≤ 0.05). The error bars indicate SEs, n = 3.

Full size image

Effects of foliar N application on nutrient distribution and flower sex differentiation

Higher N concentrations (0.5 and 0.8%) significantly increased the proportion of female flowers on lateral inflorescences (Fig. 5A), and the maximum proportions of female flowers on single inflorescences were 53.3 and 63.6%, respectively. In contrast, foliar N application had no significant effect on the male flower ratio in the terminal inflorescences (Fig. 5A). The 0.5% treatment significantly increased the number of flowers on the lateral inflorescence (25 flowers) to 1.3-fold higher than that of CK (20 flowers) (Fig. 5B). However, the number of flowers on the terminal inflorescence decreased as N concentration increased (Fig. 5B), with the 0.8% treatment causing a significant reduction. In addition, flowering was brought forward after N spraying, with the 0.5% treatment resulting in the earliest flowering (4 days earlier) (Fig. 5C).

Fig. 5
figure 5

Effects of foliar N application on male and female flowers flowering in Xanthoceras sorbifolium. (A) Proportion of male and female flowers on the terminal and lateral inflorescences after spraying N. (B) Number of flowers on the terminal and lateral inflorescences after spraying N. (C) Flowering time with different concentrations of N. Different letters represent significant differences among different treatments (p ≤ 0.05). The error bars indicate SEs, n = 10.

Full size image

To explore the effect of nutrients on flower sex differentiation, the nutrient contents of shoots and buds were measured after N spraying (Fig. 6). In most cases, the N content of the buds and shoots significantly increased, except in UBs (Fig. 6A). The P content of all samples, except MBs, was also significantly affected (Fig. 6B). The K content of all samples decreased significantly in most cases, except for MSs and MBs (Fig. 6C). Notably, at the sex differentiation stage (Stage 8), the N, P, and K contents of MBs and MSs somewhat increased.

After N application, the soluble sugar content of MSs and MBs significantly increased from April 14 to 23 (Figure 6D), and that in MSs increased the most—7.9-fold—with the 0.5 and 0.8% treatments. The starch content of MBs increased significantly (Figure 6E) during the sex differentiation stage (Stage 8).

Fig. 6
figure 6

Effects of foliar N application on nutrient content of the shoots and buds at different locations in Xanthoceras sorbifolium. N (A), P (B), K (C), soluble sugar (D), and starch contents. Different letters represent significant differences among different treatments (p ≤ 0.05). The error bars indicate SEs, n = 3.

Full size image

Discussion

Bud location effect in X. sorbifolium

Bud location effects are common in plants, especially in fruit trees. As a cross-pollinated species, X. sorbifolium showed bud distribution characteristics similar to those of Juglans regia, as the position on the annual shoot descends, bud size decreases and the buds located at the bottom part of the shoot cannot sprout18. In the present study, the buds were classified as UBs, MBs, or LBs according to their location. UBs are terminal buds that germinate into inflorescences (mainly female flowers), shoots, and leaves; MBs are middle buds that germinate into inflorescences (mainly male flowers), shoots, and leaves; and LBs are primarily distributed between the first three nodes from the base of the annual shoot. LBs can be divided into three types: wilted, male flowers without a rachis, and leaves (Fig. 1). Previous research has shown that female flowers on top of shoots can promote the pollination of cross-pollinated plants27,28. Hence, female flowers germinating on top of annual shoots in X. sorbifolium might help improve the pollination success rate as a result of long-term evolutionary adaptation.

The canopy microclimate may explain the bud distribution in X. sorbifolium. Light intensity in the leaf curtain of X. sorbifolium is in the following: upper outer layer > upper inner layer > lower outer layer > lower inner layer29. According to the principle of the nearby distribution of assimilates, light intensity reflects the intensity of photosynthesis30,31. The photosynthetic products obtained by the top buds are greater than those obtained by the axillary buds, which is related to the formation of different types of buds. In addition, considering that the buds at any positions will be affected by the season and the environment, this study has not yet analyzed their impact on the nutritional level of buds. Relevant analyses can be further expanded in subsequent studies to identify the key factors influencing the nutritional differences among different types of buds.

Effect of nutrient contents in buds, shoots, and leaves on different bud type formation

Female organs are more complex to develop than male organs and require more “raw materials,” including photosynthetic nutrients and cell-building materials. In particular, high N levels have a positive effect on prolonging ovule longevity and improving ovule viability9,32. Furthermore, high P levels can promote female flower formation and decrease pollen viability33. In the current study, during bud differentiation, the N and P contents of UBs were approximately 1.01–1.21- and 1.03–1.66-fold higher than those in MBs, respectively (Fig. 3A, B). The N content of the UBs was the highest among the three parts, particularly during the sex differentiation stage. Generally, high N and P levels positively affected female flower formation in X. sorbifolium. K content increases with stamen development10. This may explain why the K content of MBs was significantly higher than that of UBs from the petal primordium differentiation stage (Stage 4) to the flowering stage (Stage 9) (Fig. 3C). In addition, the K content of LBs was similar to that of MBs, leading to male flower germination in the LBM. Thus, it can be inferred that the high K content of buds is highly related to male flower formation in X. sorbifolium.

The soluble sugar and starch contents changed as floral organ development progressed. Cells actively divide and grow during the inflorescence elongation stage, and large amounts of soluble sugars are metabolized to synthesize cellulose, pectin, and hemicellulose34,35. Similarly, at the inflorescence elongation stage (Fig. 4A Stage 8–9) in X. sorbifolium, the soluble sugar content of the three bud types was lower than that in the previous stages. More importantly, the soluble sugar content in UBs was significantly higher than that in MBs at the sex differentiation stage (Stage 8). This indicates that female flowers consumed more carbon nutrients than did male flowers36. The USs stored more soluble sugars than the other two parts before dormancy (Fig. 4A, Stage 4), which might serve as a supplement for UBs sprouting into female flowers37. Starch grains first accumulate in the anthers of both sexes, with those in male flowers being plumper than those of female flowers in X. sorbifolium1,21. Starch grain accumulation in the ovules of female flowers does not occur until the abnormality of megaspore mother cells in male flowers38. In the present study, the starch content of MBs and LBs was significantly higher than that of UBs at the sexual differentiation stage (Fig. 4B, Stage 4), suggesting that male flowers require more starch than female flowers.

Nutrient levels also affect the differentiation of buds into mixed or leaf buds. In most cases, the N and P content of shoots and buds decreased from top to bottom (Fig. 3). It could be inferred that mixed bud differentiation has a higher demand for N and P than leaf-bud differentiation. In addition, the K content in LSs was significantly lower than that in MSs and USs, indicating that the K content stored in shoots corresponding to mixed buds was greater than that in shoots corresponding to leaf buds. The nutrient content of LBs was lower than that of UBs and MBs at most times, resulting in wilting or leaf development.

Effects of foliar N application on nutrient distribution and flower sex differentiation

Applying 0.5 and 0.8% foliar N significantly increased the N and P content and the number of female flowers in MBs, suggesting that N and P could promote female flower formation in X. sorbifolium. The resource redistribution hypothesis assumes that the energy cost of male flower development is lower than that of bisexual flower development and that the energy saved from male flowers may be redistributed to the development of bisexual flowers and fruits39,40. Therefore, we suggest that high N concentrations (0.5 and 0.8%) could break nutrient limits, induce more nutrient distribution to MBs, stimulate carbohydrate production, and promote pistil differentiation9. However, the 0.3% treatment was insufficient to exceed the minimum nutrient limit. Notwithstanding, based on the aforementioned differences in canopy microclimate, N supplementation could compensate for the decrease in the number of lateral inflorescence buds resulting from light restriction to a certain extent. In addition, it is worth noting that no immediate difference was observed in internal nutrients after spraying N in some samples, such as the N content of MBs and the P content of MSs. This observation is suspected to be related to the metabolic mechanism of the tree and needs to be further verified in the future. The present study provides a new theoretical basis for regulating the female-to-male ratio of X. sorbifolium from the perspective of nutrition.

Conclusion

Significant nutrient differences were observed among different locations in the buds, shoots, and leaves, resulting in different bud types. Terminal and middle buds germinated into inflorescences (mainly female and male flowers, respectively), shoots, and leaves. Lower buds, distributed between the first three nodes from the base of the annual shoot, could be divided into three types: wilted, developed into male flowers without a rachis, and developed into leaves. N and P significantly affected female flower formation, whereas K contributed to male flower generation. Female flower development may require more soluble sugars, whereas male flower development may require more starch. Mixed bud differentiation had a higher demand for N and P than leaf-bud differentiation. High N concentrations (0.5 and 0.8%) increased the number of female flowers on lateral inflorescences and improved the female-to-male flower ratio. N treatment advanced flowering time, and the 0.5% treatment showed the most apparent effect (4 days earlier). This study provides a novel perspective for modulating the yield of X. sorbifolium.