Sunflower (Helianthus annuus L.), an annual herb of Helianthus in Compositae, is generally 1~3 m high and is one of the four major oil crops in the world, which has anti-cancer effects and has a positive effect on improving nutrition and preventing diseases (Schilling and Heiser, 1981; Flagella et al, 2002; Liang, 2016, Nong Min et al., 13: 149-149; Wang et al., 2019). At present, sunflowers are widely distributed in Northeast, Northwest and North China. As oil crops, sunflowers are planted in more than 40 countries around the world (Cao, 2008), which can be divided into edible sunflower, oil sunflower and ornamental sunflower according to its usage (Huang et al., 2009). However, due to the serious degradation of varieties in actual production, it is urgent to select and update varieties. At present, the main breeding aims are to improve the yield and commodity of oil and edible sunflower (Zhang et al., 2018).

In the genetic research of sunflower agronomic characters, the multivariate genetic analysis method of crop quantitative traits has been widely used in the genetic research of sunflower agronomic characters. In edible sunflower, Li et al. (2010) used 275 hybrid combinations of edible sunflower as test materials and performed coefficient of variation, correlation and principal component analysis on the 15 agronomic characters of edible sunflower. Through the analysis, it was found that in the selection of edible sunflower hybrids and parents, the focus should be on yield, plant height, plant type, grain, commodity quality and other agronomic characters. Jia et al. (2019) used 20 edible sunflower varieties as materials to analyze the relationship between several agronomic characters and yield by using correlation analysis, grey correlation degree, regression analysis and principal component analysis. The results showed that the characteristics of high-yield sunflower varieties mainly included appropriate growth period, abundant leaves, plant height within a reasonable range, strong stems, large disk diameter and heavy 100-grain weight. In oil sunflower, Zhang et al. (2019) performed genetic analysis on 12 important agronomic characters of oil sunflower, and the results showed that the agronomic characters of oil sunflower affected and correlated with each other. Jia et al. (2009) analyzed the agronomic characters under different salt environments. The results showed that under different soil salinity conditions, the disk diameter and dry matter were the two main factors affecting yield, and the number of leaves had the least effect on yield.

In the genetic analysis method, the main purpose of principal component analysis method is to transform the multi index problem into several comprehensive indexes that are not related to each other, so as to more intuitively map the main information of the original indexes (Yuan and Zhou, 2002). Among sunflowers, Feng et al. (2018) analyzed the phenotypic character diversity of Guizhou local sunflower germplasm resources through principal component and cluster analysis. Sunflower resources were divided into two major categories and four subcategories. In addition, principal component analysis was widely used in soybean, wheat, rice and other crops (Li et al, 2012; Govindasamy et al, 2017; Thapa et al, 2017; Yang et al, 2019).

Zhejiang Province is located in the southeast coast. The climate is characterized by high temperature and high humidity. This environment is unfavorable to the pollen spreading, seed setting and pest control of sunflower during flowering. Therefore, there has been doubt whether sunflower is suitable for planting in Zhejiang Province. At present, no systematic study on the adaptability of sunflower in Zhejiang Province has been carried out. Based on the above situation, this study intends to carry out the difference and genetic analysis of main agronomic characters between oil sunflower and edible sunflower in Zhejiang Province. 77 oil sunflower materials and 59 edible sunflower materials were planted in Haining and Huzhou experimental bases, the important agronomic characters of which were investigated and counted to clarify the genetic variation, composition and correlation of the main agronomic characters of oil sunflower and edible sunflower in Zhejiang Province, to explore the possibility of planting oil sunflower and edible sunflower in Zhejiang Province, and to provide data support for sunflower breeding in Zhejiang Province in the future.

1 Results and Analysis

1.1 Phenotypic variation analysis

The genetic variation of oil sunflower and edible sunflower planted in the two places was analyzed. The results showed that the tested varieties had great differences in agronomic characters. The variation rate of 12 traits of oil sunflower planted in Haining was 0.06%~0.71%, and the variation rate of the total seed number was highest, which was 0.71%, while the variation rate of the mature period was lowest, which was only 0.06%; The variation rate of oil sunflower planted in Huzhou was 0.01%~0.58%, and the variation rate of the total seed number and the mature period was still the highest and lowest, respectively. The results showed that there were great differences in seed setting among different varieties, but there was little difference during the mature period (Table 1). In addition, the variation rate of 12 traits of edible sunflower planted in Haining and Huzhou was 0% ~ 0.27% and 0% ~ 0.33%, respectively, and the variation rate of the setting rate and the mature period was the highest and lowest, respectively (Table 2). In addition, the variation rate of the total seed number, setting rate and 100-grain weight of edible sunflower and oil sunflower planted in these two locations was relatively higher, but the variation rate of the mature period was lower, indicating that genetic improvement can effectively improve yield traits, but it was less effective for traits with relatively low variation rate such as mature period, flowering time, seed width, seed length and so on.

1.2 Genetic analysis

Further genetic analysis was conducted on the phenotypes of oil sunflower and edible sunflower. The results of normal distribution (Table 1; Table 2) showed that the skewness of 12 traits in oil sunflower was -6.62 ~ 1.97 and the kurtosis was -1.24~44.15. However, except for the mature period and seed length, the absolute value of skewness and kurtosis of the other 10 traits was relatively small, which basically accorded with the characteristics of normal distribution (Table 1). Similar to the results of oil sunflower, the skewness of 12 traits of edible sunflower was -4.8 ~ 6.30 and the kurtosis was -1.29 ~ 39.81. Except for the mature period and seed length, the other 10 traits basically accorded with the characteristics of normal distribution (Table 2), indicating that most agronomic characters of oil sunflower and edible sunflower were mainly controlled by quantitative traits, while the mature period and seed length may be controlled by the modes of major + minor polygene model.

In addition, the heritability results showed that among the tested oil sunflower varieties, the heritability of 12 agronomic characters in a single environment ranged from 0.31 to 0.94. Except that the heritability of disk size in flowering (without ligulae flowers) was 0.46 (Haining) and 0.31 (Huzhou), the heritability of other agronomic characters was higher than 0.6, indicating that these agronomic characters were mainly regulated by genotypes. However, among multiple environments, the heritability of 12 agronomic characters ranged from 0.11 to 0.86, indicating that the effects of environment on different agronomic characters were different. The heritability of flowering time, mature period, plant height, total seed number and seed length were higher than 0.6, indicating that the environment had less effect on these 5 characters, but had a greater effect on the other characters (Table 1). Meanwhile, in edible sunflower, the heritability of 12 agronomic characters in a single environment ranged from 0.39 to 0.96. The single-point heritability of disk size at maturity, plant height, total seed number, seed length, seed width and seed thickness was higher than 0.6 in two environments, indicating that the variation was mainly regulated by genotypes; However, among different environments, the heritability of 12 characters ranged from 0.00 to 0.99, of which only the heritability of seed length and seed thickness was higher than 0.6, indicating that the environment had less effect on these 2 characters, but had a greater effect on the other 9 characters (Table 2).

1.3 The correlation analysis for agronomic characters

The correlation of 12 characters in oil sunflower and edible sunflower was further analyzed. The results showed that (Figure 1) 8 pairs of characters in oil sunflower planted in Haining showed extremely significant correlation (p<0.01), the correlation coefficient was 0.386 ~ 0.846, 5 pairs of characters showed significant correlation (p<0.05), and the correlation coefficient was -0.315~0.297 (Figure 1a). While in oil sunflower planted in Huzhou, 12 pairs of characters showed extremely significant correlation (p<0.01), the correlation coefficient was 0.404~0.801, 10 pairs of characters showed significant correlation (p<0.05), and the correlation coefficient was -0.462~0.540 (Figure 1b). Among them, there was a significant positive correlation between flowering time and mature period; There was a significant positive correlation between disk size in flowering (including the ligulate flowers), disk size in flowering (without ligulae flowers) and disk size at maturity; There was a significant positive correlation between disk size at maturity, plant height and seed thickness, and there was a significant positive correlation between seed width and seed thickness; There was a significant positive correlation between setting rate and 100-grain weight. In addition, in edible sunflower planted in Haining, 14 pairs of characters showed extremely significant correlation (p<0.01), the correlation coefficient was -0.578~0.816, 3 pairs of characters showed significant correlation (p<0.05), and the correlation coefficient was -0.316~0.316 (Figure 1c). While in edible sunflower planted in Huzhou, 14 pairs of characters showed extremely significant correlation (p<0.01), the correlation coefficient was -0.555 ~ 0.769, 9 pairs of characters showed significant correlation (p<0.05), and the correlation coefficient was -0.415 ~ 0.444 (Figure 1d). Similar to the results of oil sunflower, there was a significant positive correlation between disk size in flowering (including the ligulate flowers), disk size in flowering (without ligulae flowers), disk size at maturity and plant height; there was a significant positive correlation between plant height and total seed number; there was a significant positive correlation between setting rate and 100-grain weight. But there was a significant negative correlation between total seed number and setting rate. The above results showed that the correlation between characters should be comprehensively considered in the process of breeding. In seed selection, the positively correlated agronomic characters can be used as a whole, but the impact of the negatively correlated characters on yield should also be comprehensively considered in breeding.

In order to further clarify the differences between oil sunflower and edible sunflower, the agronomic characters of oil sunflower and edible sunflower in Haining and Huzhou were compared and analyzed. The results showed that in Haining, 7 characters, namely total seed number, flowering time, plant height, 100-grain weight, seed length, seed width and seed thickness of oil sunflower and edible sunflower were significantly different; In Huzhou, 8 characters, namely mature period, maturity stage, disk size in flowering (without ligulate flowers), plant height, total seed number, 100-grain weight, seed length and seed width of oil sunflower and edible sunflower were significantly different. Among them, 5 characters, namely plant height, total seed number, 100-grain weight, seed length and seed width were significantly different under both two environments (Figure 2).

1.5 Principal component analysis of oil sunflower and edible sunflower

The results of principal component analysis of the tested oil sunflower and edible sunflower showed that the characteristic value of axis 1 was 22.6%, the characteristic value of axis 2 was 17.4%, and the 12 characters explained 40% of the variation of population structure distribution (Figure 3). According to the length and angle of the arrow, the effects on the population structure distribution of oil sunflower and edible sunflower from large to small were as follows: 100-grain weight > disk size in flowering (without ligulate flowers) > seed width, disk size at maturity > total seed number > disk size in flowering (including the ligulate flowers) > setting rate > seed thickness > plant height > seed length > flowering time > mature period, and the eigenvector values were 4.26, 3.61, 3.03, 2.95, 2.75, 2.62, 2.21, 2.10, 2.09, 1.27, 1.23 and 0.66 respectively. It was considered that the first principal component was 100-grain weight, the second principal component was disk size in flowering (without ligulate flowers), and the third principal component was seed width. The 100-grain weight of oil sunflower and edible sunflower was significantly different under the two growth environments. Therefore, in terms of agronomic characters, 100-grain weight was very important for the classification of oil sunflower and edible sunflower, which can be considered as the main component of yield. Therefore, we should attach importance to the characters that different types of varieties should pay attention to.

2 Discussion

The yield of sunflower is affected by many factors, such as cultivation mode, sowing mode and so on. In previous studies, the results showed that the irrigation period had a significant impact on the plant height of edible sunflower, and different densities had a significant impact on the grain commodity characters of edible sunflower (Li et al., 2006); In terms of sowing methods, compared with conventional open-air direct seeding, the yield of direct seeding covered with plastic film increased by 40.37%, and the yield of seedling raising and transplanting covered with plastic film increased by 5.05% (Zhang and Ran, 2017). At the same time, the genetic mechanism of sunflower is complex, and various agronomic characters directly or indirectly affect the yield. In the oil sunflower, agronomic characters such as plant height, single disk weight, 100-grain weight and grain width were significantly correlated with yield. Agronomic characters such as disk grain number, mature period, disk stem, stem diameter and leaf number had indirect effects on final yield (Zhang et al., 2019). In the edible sunflower, the yield of edible sunflower was positively correlated with the setting rate, the mature period was positively correlated with the leaf number and stem diameter, the plant height was positively correlated with stem diameter, average disk weight and grain width, and the 100-grain weight was positively correlated with average disk weight, grain length and grain width. In our study, the correlation between agronomic characters of oil sunflower and edible sunflower was deeply analyzed. 13 and 24 pairs of significantly correlated characters were detected in oil sunflower in Haining and Huzhou respectively, and 17 and 23 pairs of significantly correlated characters were detected in edible sunflower planted in the two places respectively. At the same time, the agronomic characters leading to different groups of oil sunflower and edible sunflower were studied. The further research of the genetic characteristics of agronomic characters is of great significance to the yield improvement of sunflower.

The sunflower is tall and grows fast, which has a short growth period and absorbs nitrogen, phosphorus and potassium nutrients more and quickly (Duan et al., 2013). In the oil sunflower, the use of nitrogen fertilizers will achieve a better yield increase than that of phosphorus fertilizer and potassium fertilizer, and the yield increase were 20.5%, 14.4%, and 11.5%, respectively (Duan et al., 2013). In the edible sunflower, the amount of nitrogen fertilizer was parabola related to the yield. When the amount of nitrogen application was 189.7 ~ 194.0 kg/hm², the highest yield of sunflower can be obtained, and the economic optimal amount of nitrogen application should be 148.5 ~ 157.4 kg/hm² (Guo et al., 2019). Wang (2013) found that the use of potassium oxide at 135 kg/hm² had the most significant effect on the yield of sunflower. In the study, Wang (2013) found that when the amount of phosphorus pentoxide application was 135 kg/hm², the yield increased significantly by 15.66%, compared with no fertilization. In the current study, Haining and Huzhou experimental bases are geographically close, but there are significant differences in soil nutrient content. The basic physical and chemical properties of the soil in the Haining experimental base are as follows: pH: 7.85, available nitrogen content: 65.56 mg/kg, available phosphorus content: 14.74 mg/kg, available potassium content: 132.42 mg/kg; The basic physical and chemical properties of the soil in the Huzhou experimental base are as follows: pH: 5.68, available nitrogen content: 150.89 mg/kg, available phosphorus content: 37.18 mg/kg, available potassium content: 53.29 mg/kg (Table 2). The difference of soil nutrient content leads to some different results of the analysis of variation rate, correlation coefficient and principal component of agronomic characters of oil sunflower and edible sunflower planted in different locations. By comparing the stable genetic characteristics of oil sunflower and edible sunflower in different environments in Zhejiang Province, we provide a theoretical basis for the genetic improvement of sunflower breeding in Zhejiang Province, and also provide a practical basis for the planting of oil sunflower and edible sunflower in Zhejiang Province.

3 Materials and Methods

3.1 Test materials

The tested materials include 77 oil sunflower materials and 59 edible sunflower materials. The edible sunflower No. S001-S069 and oil sunflower No. Y032-Y041 were introduced from the Special Oil Research Group of the Institute of Oil Crops, Chinese Academy of Agricultural Sciences. The other numbered materials were provided by the Institute of Crops and Nuclear Technology Utilization, Zhejiang Academy of Agricultural Sciences, with specific accessions and origins (Table 3).

3.2 Experimental design and investigation of agronomic characters

The oil sunflower and edible sunflower were sown in Huzhou, Zhejiang and Haining, Zhejiang experimental bases on August 5^th and August 8^th respectively, and tested after harvest in early November. Haining experimental base is located at 120.68°E and 30.53°N. The basic physical and chemical properties of soil was as follows: pH: 7.85, available nitrogen content: 65.56 mg/kg, available phosphorus content: 14.74 mg/kg, available potassium content: 132.42 mg/kg (Table 4). Huzhou experimental base is located at 120.08°E and 30.90°N. The basic physical and chemical properties of soil was as follows: pH: 5.68, available nitrogen content: 150.89 mg/kg, available phosphorus content: 37.18 mg/kg, available potassium content: 53.29 mg/kg (Table 4). In this experiment, a random block design was adopted, with one ridge and two rows, the length of the plot was 8.0 m, the row spacing was 1.0 m, the plant spacing was 0.4 m, and the plot area was 8.0 m². There were 308 plots at two points in the two places, with a total area of 2464.0 m². The field management was the same as the field production management.

The sowing to flowering time (A1), mature period (A2), disk size in flowering (including the ligulate flowers) (A3), disk size in flowering (without ligulate flowers) (A4), disk size at maturity (A5), plant height (A6), total seed number (A7), setting rate (A8), 100-grain weight (A9), seed length (A10), seed width (A11) and seed thickness (A12) were investigated respectively. The character investigation methods were referred to Guidelines for Testing Specificity, Consistency and Stability of New Sunflower Varieties (NY/T 2433-2013).

3.3 Data analysis

The software Microsoft Excel 2016 was used for the calculation of basic statistics, SPSS 22.0 was used for correlation analysis, significant difference analysis and normal analysis, ICIMapping V4.1 was used for heritability analysis, and RStudio was used for principal component analysis.

Authors’ contributions

CSY was the executor of the experimental design and project research; ZH participated in data analysis; HYT and MGF participated in the analysis of test results; TY and CDD participated in the collection of test data; CRX participated in the revision of the manuscript; RGH was the conceiver of the project and guided the revision of the manuscript; CSY completed experimental design, data analysis, manuscript writing and revision. All authors read and approved the final manuscript.

Acknowledgments

This study was jointly funded by the Key Research and Development Plan (Agricultural Field) Project of Zhejiang Province (2019C02004) and the ‘San Nong Liu Fang’ Science and Technology Cooperation Plan Project of Zhejiang Province (2019SNLF003).

References

Cao M.L., Development general situation and economic value of sunflower in China, Zuowu Keji (Crop Technology), 6: 19-20

Duan Y., Tuo D.B., Zhang J., Li H.C., Zhao P.Y., An H., and Yao J.Q.,2013, Effects of N, P and K fertilizers on yield and fertilizer efficiency of oil sunflower, Zhiwu Yingyang Yu Feiliao Xuebao (Journal of Plant Nutrition and Fertilizer), 19(03): 761-771

Feng B., Li H.L., Yang S.X., Lin T., and Wei Z.F., 2018, Analysis of genetic diversity of sunflower germpasm resources in Guizhou, Zhongguo Nongye Keji Daobao (Journal of Agricultural Science and Technology), 20(7): 34-41

Flagella Z., Rotunno T., Tarantino E., Di Caterina R., and De Caro A., 2002, Changes in seed yield and oil fatty acid composition of high oleic sunflower (Helianthus annuus L.) hybrids in relation to the sowing date and the water regime, European Journal of Agronomy, 17(3): 221-230

https://doi.org/10.1016/S1161-0301(02)00012-6

Govindasamy V., George P., Aher L,Ramesh S.V., Thangasamy A., Anandan S., Raina S.K., Kumar M., Rane J., Annapurna K., and Minhas P.S., 2017, Comparative conventional and phenomics approaches to assess symbiotic effectiveness of Bradyrhizobia strains in soybean ( Glycine max L. Merrill) to drought, Scientific Reports, 7(1): 6958

https://doi.org/10.1038/s41598-017-06441-3

Guo S.J., Duan Y., Zhang J., An H., and Zhang R.S., 2019, The effect of nitrogen fertilization on sunflower yield and quality, Zhongguo Turang Yu Feiliao (Soils and Fertilizers Sciences in China), 280(02): 126-132

Huang Z.Q., Shi C.C.,Wang W., and Zhu T.L.,2009, High-yield cultivation technique of Sunflower and supporting application of agricultural machinery, Nongye Jishu Yu Zhuangbei (Agricultural Technology & Equipment), 169: 21-22  

Jia X.J., Ding B.H., Yang F, Wu X.M., and Lei W., 2019, Multivariate analysis of main agronomic traits and yield of edible sunflower, Zhongguo Nongxue Tongbao (Chinese Agricultural Science Bulletin), 35(4): 1-6

Jia X.P., Yue Y., and Chen B.D., 2009, Influence of salt stresson maturity and agronomic characters of oil sunflower, Zuowu Zazhi (Crops), (6): 45-48

Li J., Wang L.J., and L Y., 2012, The identification system of wheat pests based on pca and svm, digital manufacturing and automation (ICDMA), Third International Conference on. IEEE, DOI: 10.1109/ICDMA.2012.217

https://doi.org/10.1109/ICDMA.2012.217

Li S.P., An Y.L., Guo S.C., Nie H., and Zhang M.Y., 2010, Analysis on hereditary variation, correlation and principal components of main agronomic characters in confectionery sunflower, Heilongjiang Nongye Kexue (Heilongjiang Agrictultural Science), (09): 7-10

Li X.L., Hou H.F., Zhang F., Wei J.G., and Xu F., 2006, Effect of different cultivation way on biological characters of edible sunflower, Xinjiang Nongye Kexue (Xinjiang Agricultural Sciences), 43(02): 133-135

Schilling E.E., and Heiser C.B., 1981, Infrageneric classification of helianthus, Compositae, 30(2): 393-403

https://doi.org/10.2307/1220139

Thapa S., Prasanna R., Ranjan K., Velmourougane K., and Ramakrishnan B., 2017, Nutrients and host attributes modulate the abundance and functional traits of phyllosphere microbiome in rice, Microbiological Research, 204: 55

https://doi.org/10.1016/j.micres.2017.07.007

Wang X.Z., Jia X.P.,Liang G.S., Chen B.D., Miao X.H., and Wang L.X., 2019, Identification of disease resistance of 39 sunflower germplasm resources in Gansu province, Gansu Nongye Keji (Gansu Ag R. Sci. and Techn), (10): 57-62

Wang Z.L., 2013, Effect of different phosphate fertilizers dosage on the yield of edible sunflower in irrigated fields, Xiandai Nongye Keji (Modern Agricultural Science and Technology), (21): 83-86

Wang Z.M.,2013, Effect of different potassium fertilizers dosage on the yield of edible sunflower in irrigated fields, Xiandai Nongye Keji (Modern Agricultural Science and Technology), (23): 101-102

Yang Q., Yang Y., Xu R., Lv H.Y., and Liao H., 2019, Genetic analysis and mapping of QTLs for soybean biological nitrogen fixation traits under varied field conditions, Frontiers in Plant Science, 10

https://doi.org/10.3389/fpls.2019.00075

Yuan Z.F., and Zhou J.Y., 2002, Multivariate statistical data analysis, Beijing: Science Pres, pp.188-190

Zhang B.Y., Bai W., and Yang S.M., 2019, Analysis of the factors in the production of edible sunflower in the north of shanxi Province, Nongye Keji Tongxun (Agricultural Science and Technology Newsletter), (7): 196-199

Zhang B.Y., Bai W., Hu Y., Yang S.M., Cui J.L., Jin T., and Bai H.H., 2019, Analysis of the factors in the production of oil sunflower, Liaoning Nongye Kexue (Liaoning Agrictultural Science), (3): 18-21

Zhang L.J., and Ran S.B., 2017, Effect of sowing method on yield of edible Sunflower, Gansu Nongye Keji (Gansu Agr. Sci. and Techn), 11: 43-44