Introduction

Camellia oleifera is a small evergreen tree belonging to the family Theaceae Camellia, which is a unique high-quality woody plant in China1. The oil extracted from C. oleifera seed kernels has high nutritional value, and its fatty acid composition is very similar to that of olive oil. As a result, it has been dubbed "oriental olive oil"2. In China, approximately one-seventh of the population uses tea seed oil for cooking. Tea seed oil contains more than 80% unsaturated fatty acids, such as oleic acid, linoleic acid and linolenic acid. It is characterized by its unique flavor, long storage life and easy absorption by the human body3,4. Tea seed oil is also rich in functional active substances, such as carotenoids, polyphenols, flavonoids, β-sitosterol, squalene and vitamins5, and these substances have a variety of physiological functions, such as delaying aging and inhibiting free radicals6,7. Long-term consumption of tea seed oil can effectively prevent coronary heart disease7, arteriosclerosis8, skin diseases9 and tumors10.

The low-heat valley area of Guizhou Province is located in the sloped transition from the Yunnan-Guizhou Plateau to the low mountains and hills in Guangxi, China. The area has a warm and humid subtropical monsoon climate. Wangmo red ball Camellia oleifera is a special type of Camellia oleifera distributed in this area. It has been rated as an excellent farm variety in China since the 1970s and 1980s11. Wangmo C. oleifera is characterized by large fruits, smooth and red peels, strong disease and insect resistance, early fruiting and high yield; moreover, it possesses excellent adaptability to the mountain climate of dry and hot river valleys. In recent years, researchers have examined a variety of aspects of various types of C. oleifera seed oil, including its nutritional components9,12, functions7 and applications13. The influence of the oil-making process on the quality of seed oil has also been investigated14. As the value of Wangmo C. oleifera has gradually become clear, the number research reports on its characteristics has also increased. However, only the analysis of fruit characteristics has been reported in the literature15, and the reports on oil quality have not been systematic and comprehensive. Thus, in this study, research regarding the quality of the oil of Wangmo C. oleifera was carried out to provide basic data for enhancing the quality of the oil.

In recent years, with the increasing demand for fast food and the high fat content diet predominantly found in modern countries, the importance of finding more high-end edible oils from natural resources has been highlighted. An in-depth understanding of the oil quality of Wangmo C. oleifera would form the basis for selective breeding of this species with the goal of generating of high-quality edible oil. In addition, a comprehensive assessment of Wangmo C. oleifera plants would serve as the basis for the utilization of excellent germplasm resources. A number of methods have been used for the comprehensive assessment of excellent plants, such as the fuzzy comprehensive evaluation method16, the integrated index method17, principal component analysis (PCA)18, the reasonableness and satisfaction index method19 and the gray correlation degree method20. All these methods have their own merits and drawbacks. The quality of oil fat is measured by several indices, and therefore, univariable methods fail to establish models for the relationship among samples. For the gray correlation degree method and the fuzzy comprehensive evaluation method, determination of the weight coefficients of the factors plays a critical role in assessing the quality of Camellia seed oils. Thus, to provide a reliable theoretical basis, researchers must determine the weight coefficient distribution of the factors from multiple perspectives. PCA is a statistical analysis method that reduces the dimension of multiple groups of data to several independent information groups. The advantage of this method over other evaluation methods is that it can eliminate the interference among the evaluation indices without the need to artificially assign weights or create ideal weights for the factors to ensure the objectivity of the evaluation. Similar to the reasonableness and satisfaction index method, PCA also adopts weighted summation to calculate the comprehensive score; however, the weights originate from the ratio of the contribution rate of the extracted principal component to the total contribution rate. For this reason, it has been extensively employed in the comprehensive evaluation of drugs21, grain22, water23 and Camellia.

In this study, we performed a comprehensive assessment of the fatty acid composition and functional nutrients of Wangmo C. oleifera seed oil using PCA. The plants were 20 disease/pest-resistant Wangmo C. oleifera germplasm resources of high and stable production based on five-year observations. The final results of this study identified germplasms with excellent comprehensive performance as candidate resources for cultivar selection and also provide basic data for the development of high-end edible oils, establishing a theoretical basis for the development of economic forest species in low-heat valley areas across the world.

Results

Yield per unit crown width

As shown in Table 1 (supplementary File 1), the yields per unit crown width of the 20 studied plants ranged from 4.46 to 7.72 kg/m2, with an average of 5.57 kg/m2. The yield per unit crown width of GH28 was the highest (7.72 kg/m2), while that of GH61 was the lowest (4.46 kg/m2).

Table 1 Crown width yield and ranking of 20 superior strains of Wangmo red ball Camellia oleifera.
Full size table

Fat content

The fat contents of the 20 plants differed greatly (Fig. 1, Supplementary File 1), ranging from 35.03 to 53.47%, with an average of 47.03%. The highest fat content was observed in GH33 (53.47%), and the lowest was found in GH28 (35.03%). Only one plant had a fat content lower than 40%, 15 plants had fat contents between 40 and 50%, and 4 plants had fat contents between 50 and 60%.

Figure 1
figure 1

Seed fat contents of 20 strains of Wangmo red ball Camellia oleifera. Note the same letters in the same line indicate no significant difference (P > 0.05), while different lowercase letters in the same line indicate a significant difference (P < 0.05).

Full size image

Fatty acid composition

As shown in Table 2 (supplementary File 1), the fatty acids of Wangmo C. oleifera included palmitic acid, palmitoleic acid, stearic acid, oleic acid, linoleic acid, linolenic acid and cis-11-eicosenoic acid, and the plants differed significantly in fatty acid composition (P < 0.05). Both saturated fatty acids (SFAs) and unsaturated fatty acids (UFAs) were found in C. oleifera seed oil; the UFAs included both monounsaturated fatty acids (MUFAs) and polyunsaturated fatty acids (PUFAs), with MUFAs as the main type.

Table 2 Fatty acid components in the seed oil of 20 Wangmo red ball Camellia oleifera germplasms.
Full size table

The SFAs included palmitic acid and stearic acid and the SFA content of the plants ranged from 8.880 to 11.027%. The content of palmitic acid ranged from 6.773% (GH61) to 9.040% (GH8), with an average of 7.845%. The content of stearic acid ranged from 1.360% (GH7) to 2.560% (GH61), with an average of 2.066%. The MUFAs included oleic acid, palmitoleic acid and cis-11-eicosenoic acid, and the MUFU content of the plants ranged from 78.616 to 84.868%. Oleic acid was the main type of MUFA; the content of oleic acid ranged from 77.967% (GH8) to 83.733% (GH61), with an average of 80.785%. The highest palmitoleic acid content was observed in GH61 (0.640%), while the lowest was observed in GH36 (0.006%). The content of cis-11-eicosenoic acid ranged from 0.488 to 0.644%. The PUFA content of the selected plants ranged from 5.705 to 11.210% and included linoleic acid and linolenic acid. The content of linoleic acid ranged from 5.417% (GH56) to 10.900% (GH36), with an average of 8.313%. The content of linolenic acid ranged from 0.270% (GH33) to 0.507% (GH28), and the average was 0.348%.

The plants of Wangmo C. oleifera also differed greatly in the coefficients of variation of (CV) the fatty acid components, and the CVs ranked from highest to lowest were as follows: C16:1 (129.64%), C18:2 (21.32%), C18:3 (18.24%), C18:0 (17.46%), C20:1 (8.30%), C16:0 (8.19%), and C18:1 (2.47%). The CVs of the SFA, MUFA and PUFA contents were 5.24%, 2.52% and 20.75%, respectively. The SFA content in all samples was below 12%, while the UFA content was as high as 90%, indicating that Wangmo C. oleifera seed oil is an excellent edible oil rich in UFAs.

Nutrient composition analysis

Carotenoids, total polyphenols, total flavonoids, β-sitosterol, squalene and vitamin E were the main nutrients in Wangmo C. oleifera in the low-heat valley (Table 3, Supplementary File 1), and their contents differed significantly (P < 0.05). Among the 20 germplasms considered, the carotenoid content ranged from 5.067 mg/kg (GH34) to 11.850 mg/kg (GH43), with an average of 7.404 mg/kg. Polyphenols are natural antioxidants that scavenge free radicals in the body and have strong anti-inflammatory and antibacterial effects24,25. The content of total polyphenols varied from 10.333 mg/kg (GH33) to 23.800 mg/kg (GH58), with an average of 16.062 mg/kg. GH7 had the lowest total flavonoid content (0.180 g/100 g), while GH59 had the highest (0.630 g/100 g). The average total flavonoid content was 0.401 g/100 g. GH58 had a significantly higher squalene content than any other germplasm (230.567 mg/kg), whereas GH62 had the lowest squalene content (61.800 mg/kg). The average squalene content of all germplasms was 129.315 mg/kg. Vitamin E has four configurations: α-Ve, β-Ve, γ-Ve and δ-Ve. However, only α-Ve was detected in Wangmo C. oleifera. The most bioactive structure in vitamin E26, α-Ve can promote fertility, improve lipid metabolism, promote the secretion of sex hormones, and delay aging27. In this study, GH36 had the highest α-Ve content (28.767 mg/100 g), while GH43 had the lowest content (16.457 mg/100 g). The average α-Ve content of all germplasms was 21.505 mg/100 g. Therefore, Wangmo C. oleifera seed oil can be regarded as a natural α-Ve supplement. As shown in Table 3, the CVs of the six nutrients in the 20 germplasms ranged from 16.79 to 33.86%. β-Sitosterol had the smallest CV (16.79%), indicating that β-sitosterol had the smallest improvement potential. In contrast, squalene had the highest CV (33.86%), which indicated a rather high improvement potential. According to the CV values, the nutritional substances were rated in terms of improvement potential as follows: squalene > total flavonoids > total polyphenols > carotenoids > α-Ve > β-sitosterol.

Table 3 Nutritional composition of seeds and oils of 20 strains of Wangmo red ball Camellia oleifera.
Full size table

Correlation analysis of nutritional indices

The correlation analysis of the 11 quality indices of Wangmo C. oleifera seed oil in the low-heat valley of Guizhou Province showed variability in the correlations among them (Fig. 2). There was a significant negative correlation between fat and PUFA contents, and the correlation coefficient was − 0.48. The SFA content exhibited a very significant negative correlation with the MUFA content, with a correlation coefficient of − 0.58. The MUFA content had a very significant negative correlation with the PUFA content and α-Ve, with correlation coefficients of − 0.97 and − 0.64, respectively.

Figure 2
figure 2

Correlation matrix of quality indices of Wangmo red ball Camellia oleifera. Note *: significant correlation was found at the level of 0.05 (bilateral); **: extremely significant correlation was found at the 0.01 level (bilateral).

Full size image

The positive correlation between PUFA content and α-Ve was significant, with a correlation coefficient of 0.71. Although other nutritional indices also had different degrees of correlation, none of the correlations were statistically significant. The nutritional components of different Wangmo C. oleifera germplasms varied, and the correlations between the components also differed. Therefore, evaluating the nutritional quality of Wangmo C. oleifera seed oil only based on a single index would be an incomplete assessment.

PCA

In the evaluation of the nutritional components of Wangmo red ball Camellia oleifera seed oil in the low-heat valley of Guizhou, a comprehensive, systematic and scientific assessment of the quality of Camellia oleifera seed oil should be considered, focusing on more than one or even several nutritional components. PCA is an effective method for comprehensive evaluation. It is a multivariate statistical analysis method that can select a small number of important variables via the linear transformation of multiple variables28. In this study, the fat, SFA, MUFA, PUFA, carotenoid, total polyphenol, total flavonoid, β-sitosterol, squalene and α-Ve contents were selected for PCA. Among these considered indices, the lower the value of saturated fatty acids was, the better the quality of the seed oil. To ensure that all parameters were positively correlated with the quality of seed oil, the saturated fatty acid was set to a negative number, and the original values of each index were standardized.

Four principal components with eigenvalues greater than 1 were extracted by PCA. As shown in Table 4, the cumulative variance contribution rate of the first four principal components was 76.734%, indicating that these components contained most of the information for the nutritional indices of Wangmo red ball C. oleifera seed oil. These four components could be used to represent the original 10 evaluation indices for the analysis and evaluation of the nutritional components of Wangmo red ball C. oleifera seed oil.

Table 4 Component load matrix of each nutritional index.
Full size table

As shown in Table 4, PC1 contained the comprehensive indices of fat, SFA, PUFA and α-Ve contents. The eigenvalue of PC1 was 3.294, and its contribution rate was 32.943%. Among the indices, PUFAs contributed the most to PC1, with a load value of 0.949. The characteristic value of PC2 was 1.799, with a contribution rate of 17.986%. This component only contained squalene, with a load value of 0.669. PC3 included SFAs, total polyphenols and β-sitosterol, with a characteristic value of 1.404. The load values of the three factors were 0.550, − 0.657 and 0.582, respectively. PC4 had a high negative correlation with total flavonoids, and the load value was − 0.704, indicating that total flavonoids was the main evaluation index in the fourth component.

With the proportion of the contribution rate corresponding to the four principal components in the total contribution rate as the weight, the comprehensive evaluation function of C. oleifera seed oil in the low-heat valley area was as follows: F = 0.429PC1 + 0.234PC2 + 0.183PC3 + 0.153PC4. The principal component score, comprehensive score and ranking of each excellent germplasm were calculated (Table 5). In PC1, GH8 scored the highest (2.510). GH43 scored the highest in PC2 and PC4 (3.194 and 2.299, respectively). The highest score in PC3 was that of GH31 (2.126). The top five nutritional qualities among the 20 excellent single plant tea seed oils were observed for GH7, GH43, GH28, GH8 and GH31, among which GH7 and GH43 had the top two comprehensive nutritional values.

Table 5 Principal component score, comprehensive score and ranking of all superior strains of Wangmo red ball Camellia oleifera.
Full size table

Discussion

Wangmo red ball C. oleifera is a high-quality woody edible oil tree species, but its development and utilization remains in the early stages. The quantitative analysis and evaluation of the seed oil quality indices can provide basic data for the selection of high-quality edible oil. The yield per unit crown width and fat content are important indicators for measuring the yield of C. oleifera29. The crown width yield of Wangmo red ball C. oleifera in the low heat valley of Guizhou was significantly higher than that of "Changlin" series C. oleifera (0.76 ~ 0.9 kg/m2)30 (Table 1). The high performance of Wangmo C. oleifera can be attributed to its varieties and is also closely related to the low-heat valley area. This area is located in the hot subtropical climate zone of the low-middle mountain canyon, with alternating dry and wet seasons, and its climatic conditions are conducive to the growth of C. oleifera fruit. This study showed that the crown yield of Wangmo C. oleifera was negatively correlated with the fat content, which was consistent with the results of Liu Wei31 who reported the correlation between yield and kernel fat content. The fat content of C. oleifera seed oil directly affects the promotion, utilization and market value of the species. The fat content of olive oil is 43.7% to 56.7%32, that of soybean oil is 14.51% to15.51%33, that of sunflower oil is 35.1% to 43.0%34, and that of mustard seed oil is 33.0 to 38.0%35. In this study, the fat content of Wangmo C. oleifera seed oil was between 35.03 and 53.47% (Fig. 1). Thus, compared with the fat content of many edible oils, the fat content of Wangmo C. oleifera seed oil was high. This finding indicates that Wangmo C. oleifera has satisfactory practical value and development potential.

Edible oils rich in MUFAs are easily absorbed by the human body, and the digestibility can exceed 95% (Table 2). These oils can effectively promote human metabolism and reduce the incidence of cardiovascular and cerebrovascular diseases. The composition and content of fatty acids in edible oil can regulate human health. In the fatty acids of Wangmo C. oleifera, the main component was UFAs, with a content of approximately 90%; the SFA content was approximately 10%. The composition and content of the fatty acids in Wangmo C. oleifera satisfied the requirements for SFA content based on international nutritional standards36. The SFA content in tea oil increases with increasing latitude of origin, while the UFA content decreases with increasing latitude37,38. In this study, Wangmo C. oleifera was obtained from a low-latitude area, and the UFA content was higher than that in plants growing in high-altitude areas39,40. Therefore, it is reasonable to assume that the high content of UFAs in Wangmo C. oleifera is associated with the low-latitude region. Oleic acid is a single UFA. The content of oleic acid is an important indicator to evaluate the quality of edible oil and to judge the stability of the oil. The high content of oleic acid in tea seed oil means that it has high stability and digestibility, as well as health-promoting effects such as reducing blood pressure, cholesterol and triglycerides. In this study, we found that the content of oleic acid in Wangmo C. oleifera oil was 80.79%, which had not been previously reported in the literature. Ceheng C. oleifera and Wangmo C. oleifera belong to the same Camellia species, and they grow in regions with consistent climatic and geographical conditions. According to the literature39, Ceheng C. oleifera oil has an oleic acid content of 80.19%. Our findings indicate that Wangmo C. oleifera produces oil with high stability and high nutritional value. The oleic acid contents in different editable oils vary: 74.20% in olive oil, 58.19% in rapeseed oil, 25.90% in corn oil, 79.71% in sunflower oil, 39.46% in peanut oil, 22.40% in bean oil and 28.06% in pumpkin oil. This study showed that the oleic acid content of Wangmo C. oleifera seed soil was as high as 80%, which was higher than that of other edible oils, including the oils produced from different C. oleifera species39,40,41. In terms of the oleic acid content, Wangmo C. oleifera seed oil has an absolute advantage over ordinary vegetable oil and a higher nutritional value than most C. oleifera seed oils. Therefore, based on the analysis of the fatty acid composition and oleic acid content, Wangmo red ball C. oleifera seed oil is a high-quality woody edible oil.

Functional components, such as carotenoids, polyphenols, flavonoids, β-sitosterol, squalene and vitamins, can also serve as important indicators for the evaluation of the quality and function of tea oil. According to the current study, the contents of β-sitosterol, α-Ve and total flavonoids in Wangmo red ball C. oleifera seed oil were 265.087 mg/kg, 21.505 mg/100 g and 0.401 g/100 g, respectively (Table 3). Among C. oleifera varieties, the levels of β-sitosterol, α-Ve and flavonoids in Wangmo C. oleifera seed oil were higher than average based on the values reported in the literature39,40,42. The content of total flavonoids in Wangmo C. oleifera was as high as 0.401 g/100 g in this study, which was far higher than that in Camellia (0.360–0.540 mg/g), as reported in the literature43. The reason might be related to the growing conditions. Wangmo C. oleifera grows in Guizhou, where differences in region of production and species have a great influence on the content of active substances in Camellia seed oil. The content of total flavonoids in Cyclocarya paliurus exhibits an increase with altitude to a point, followed by a decrease with altitude44. In Yang et al.’s study43, the Camellia material was from Taiwan, and the plants were primarily distributed at an altitude of approximately 2000 m. In contrast, the material in our study was from the Wangmo region in Guizhou, with an altitude of approximately 800 m. Camellia weiningensis grows at 1800 m, and its total flavonoid content was 0.540 mg/g39. From the perspective of the influence of altitude, our findings regarding Wangmo C. oleifera and those reported for other Camellia species support the trend in the content of total flavonoids which increases and then decrease with altitude. The climate in Wangmo is humid and hot with sufficient sunlight. studies on the mechanism underlying the formation of functional active substances in oil fat have been limited. Therefore, whether the high content of total flavonoids in Wangmo C. oleifera seed oil is related to the unique climate characteristics requires further research. In Wangmo C. oleifera, the coefficients of variation of the six nutritional substances studied ranged from 16.79 to 33.86% (Table 2). All of the indices exhibited some improvement potential, and Wangmo C. oleifera may be improved according to the international demands for functional nutrients in edible oils.

In this study, we considered four oil quality indices which reflected most of the information on the oil quality of Wangmo C. oleifera., and the cumulative variance contribution rate was 76.734% (Table 4). The factor scores and corresponding weights were ranked and summed. The top five germplasms in terms of oil nutritional quality were GH7, GH43, GH28, GH8 and GH31 (Table 5). This ranking provides a guideline for promoting the excellent Wangmo red ball C. oleifera plants with high nutritional quality. The final results of this study demonstrated that Wangmo red ball C. oleifera seed oil is rich in UFAs, carotenoids, polyphenols, flavonoids, β-sitosterol, squalene and vitamins, and it has dual nutritional and medicinal effects and great potential to be developed into high-quality edible oil.

Materials and methods

Experimental site and materials

The experimental site was located in the Camellia oleifera germplasm resource nursery in the low-heat valley area of Guizhou University. The climate type was that of the middle and low mountain canyon and the warm and humid climate of the southern subtropical zone. Summer and autumn are controlled by the southwest and southeast monsoons. The summer is rainy and hot, and continuous drought occurs in winter and spring.

The materials used in this study were 20 wild Wangmo C. oleifera trees that were observed by the C. oleifera team of Guizhou University for five consecutive years (2016–2021). The trees were healthy and free of pests and disease (or with only a few pests) and had high pollination and seed setting rates, stable yield and stable performance. The fruit was collected in late November 2020 in the mature stage, and slight cracking of the peel was a sign considered indicative maturity. Fresh fruits from the selected plants were collected, bagged and labeled for later use.

We adhered to all local, national and international guidelines and legislation for the use of plants in this study.

Study design

Figure 3 shows a schematic diagram of the study design, and the fatty acid composition and functional nutrients of Wangmo C. oleifera seed oil were assessed comprehensively with the goal of finding a germplasm with excellent comprehensive performance. Briefly, the crown yield of C. oleifera was investigated when the fruit was ripe. The fresh fruit was sampled and taken back to the laboratory for drying, and then, the fatty acid and nutritional components were analyzed. Three replicates were performed for each sample.

Figure 3
figure 3

Schematic diagram of the study design.

Full size image

Experimental indices

Unit crown yield

During the fruit ripening period, the crown width of each C. oleifera plant was measured. Fresh fruit was picked and converted to the fresh fruit yield of the whole plant. The yield per unit crown width of each plant was calculated according to the following formula:

$${\text{Yield per unit crown width}}\left( {{\text{kg}}/{\text{m}}^{{\text{2}}} } \right) = \,{\text{Fresh fruit weight of whole}}\;C.\, oleifera\;{\text{plant}}/{\text{crown width of the plant}}.$$

Fat content

The sampled fruits were naturally dried. The seeds were collected and then dried in an oven at 40 °C to a constant weight, peeled and then ground with a mortar. Approximately 10 g of the sample was weighed and wrapped in a dried filter paper bag. The bag containing the samples was placed into a Soxhlet extractor (SOX606) for 10-h extraction with petroleum ether. The fat content was calculated as follows:

$${\text{Fat content }}\left( \% \right)\, = \left( {{\text{mass of the extracted fat}}/{\text{that of the sample kernels}}} \right) \times 100\%$$

Fatty acid composition analysis

Approximately 100 mg of oil was transferred into a test tube with a ground glass plug, and the oil was methylated using the potassium hydroxide-methanol method. The samples were mixed on a vortex mixer and then oscillated at 40 °C for 30 min. The methyl ester was extracted with 2 ml of n-hexane (analytical grade), and the aqueous phase was discarded. The extract was washed with distilled water and dried with anhydrous sodium sulfate. After filtration, the samples were analyzed with a Shimadzu GC2014 gas chromatograph.

The gas chromatographic conditions were as follows: detector, FID; column type and specification, SP2340; chromatographic column, 60 mx0.25 111/11 × 0.2 um; heating program, 50 °C (initial temperature; retention, 2 min) followed by 170 °C (10 °C/min; retention, 10 min), 180 °C (2 °C/min; retention, 10 min) and 220 °C (4 °C/min; retention, 22 min); injection port temperature (°C), 250; detection temperature (°C), 300. Nitrogen was used as the carrier gas with a shunt ratio of 1:50 and a volume of 1 µL.

The composition of fatty acids was determined based on a comparison with the retention time of various fatty acid standard samples. The relative content of each fatty acid was calculated by the peak area normalization method. The experiment was repeated three times, and average values were obtained.

Nutritional substance determination

  1. (1)

    Vitamin E. Oil samples at 0.7 g were accurately weighed, and 10 ml of n-hexane was added for dissolution to a constant volume. The solution was filtered through a 0.22 μM organic filtration membrane for high-performance liquid chromatography (YLSZJ-SB-203).

The liquid chromatographic conditions were as follows: chromatographic column, water amino column; column temperature, 30 °C; mobile phase, n-hexane + isopropanol = 98 + 2; flow velocity, 1.2 mL/min; fluorescence detection wavelength, 295 nm (excitation wavelength) and 333 nm (emission wavelength); amount, 10 μL.

  1. (2)

    Total polyphenols Separate preparations of gallic acid at concentrations of 10, 20, 30, 40 and 50 µg/mL were used. A 2.00 g sample was accurately weighed, dissolved with n-hexane and then passed through the glycol-based column. After purification with 10 mL of n-hexane, the sample was eluted with 10 ml of methanol and then bathed at 45 °C for drying with nitrogen. The sample was redissolved with 50% methanol and then oscillated, chilled and centrifuged. The supernatant was collected. A gallic acid working solution (1 mL), distilled water (1 mL) and the sample to be determined (1 mL) were transferred into a scaled tube. Folin-phenol reagent (0.5 mL), 7.5% sodium carbonate solution (2 mL) and water (6.5 mL) were added. The mixture was subjected to vortex oscillation for 1 min and then bathed at 70 °C for 30 min. The sample was placed on a 10-mm cuvette, and the absorption was read at a wavelength of 750 nm. A standard curve was plotted according to the absorbance of the gallic acid working solution and the concentration of the working solution. Based on the standard curve, the polyphenol concentration was determined, and the content was calculated.

  2. (3)

    Squalene and β-sitosterol An oil sample of 0.4 g was applied to a conical bottle, and 10 mL of KOH-ethanol (2 mol/L) was added. The solution was saponified at 80 °C for 50 min. The unsaponified substances were extracted with n-hexane, washed to neutral, dried and then brought to 2 ml. Gas chromatography was used to calculate the contents of squalene and β-sitosterol with the external standard method.

The chromatographic conditions were as follows: detector, FID; chromatographic column: Agilent HP-5; injection port temperature: 220 °C; heating program, 160 °C for 1 min, which was increased by 15 °C/min to 280 °C for 5 min, followed by an increase by 5 °C/min to 300 °C for 7 min; injection volume: 1 μL.

  1. (4)

    Total flavonoids The 10 mg rutin standard was accurately weighed and brought to 500 mL with 70% ethanol to prepare a 0.2 mg*mL−1 standard solution. Standard rutin solutions at 1, 2, 3, 4, 5 and 6 mL, as well as 70% distilled water, were separately prepared. They were transferred into seven 25-ml volumetric flasks. Approximately 1 mL of 5% NaNO2 was added. The mixture was oscillated and then allowed to stand for 6 min. Approximately 1 mL of 10% AI(NO3)3 was added, and the mixture was oscillated and then allowed to stand for 6 min. Afterward, 10 mL of 4% NaOH was added, and the mixture was brought to volume with distilled water to 25 mL. The mixture was allowed to stand for 15 min. The absorption was read at a wavelength of 510 nm, and the absorption-concentration curve and regression equation were obtained.

The solution to be determined was centrifuged at 6000 r min−1 for 10 min. The supernatant (2 mL) was transferred into a 25-mL volumetric flask. As described above, 1 mL of 5% NaNO2, 1 mL of 10% AI(NO3)3 and 10 mL of 4% NaOH were added sequentially. The mixture was scaled to 25 mL, and then an absorption reading was performed. The formula for the total flavonoid content was as follows:

$$X\left( {g/100g} \right) = \frac{A \times C \times 10}{{m \times v \times 1000}}$$

where X is the content of the total flavonoids in the sample, A is the content of total flavonoids in the sample extract calculated based on the standard curve, C is the size of the sample, V is the volume of the solution to be determined, and m is the mass of the sample.

  1. (5)

    Carotenoids One gram of oil sample was weighed, and 5 mL of petroleum ether and 5 mL of acetone were added. The absorption of the sample was read at 445 nm, with β-carotene as the standard.

Statistical analysis

Data were sorted using Excel 2013, and multiple comparisons were performed using SPSS 26.0. The t test was used for comparisons between groups. A difference of P < 0.05 was considered statistically significant. After data standardization, principal component analysis (PCA) was performed using SPSS 26.0. The number of principal component factors was determined based on the principle that the eigenvalue of the factor must be greater than 1.00 and the cumulative variance contribution rate must be greater than 80%. After the principal components were extracted, the comprehensive evaluation was carried out, and the comprehensive score F was obtained by summing the scores of each principal component with the weight of the variance contribution rate of the corresponding eigenvalue. Origin2018 and ArcGIS were used for plotting.