Phenolic Compound Content of Leaf Extracts from Different Roselle (Hibiscus sabdariffa) Accessions

Article information

Plant Breed. Biotech.. 2020;8(1):1-10
Publication date ( electronic ) : 2020 March 1
doi :
1Advanced Radiation Technology Institute, Korea Atomic Energy Research Institute, Jeongeup 5622, Korea
2Division of Plant Biotechnology, College of Agriculture and Life Science, Chonnam National University, Gwangju 61186, Korea
3Department of Life Resources, Graduate School, Sunchon National University, Suncheon 57922, Korea
*Corresponding author Soon-Jae Kwon,, Tel: +82-63-570-3312, Fax: +82-63-570-3813
received : 2019 October 15, rev-recd : 2019 November 12, accepted : 2019 November 20.


The leaves of roselle (Hibiscus sabdariffa L.) have been used as a traditional folk medicine that has diuretic and mild laxative effects. Roselle is cultivated in many countries for medicines and food. However, studies on the variation of functional compounds in different accessions are relatively limited. In this study, we investigated the phenolic compound content of leaf extracts from 49 different roselle accessions from a worldwide collection by ultra-high performance liquid chromatography mass spectrometry. The aim was to chemically classify and assess the pharmacological properties of the phenolic compounds in the leaves. The phenolic compounds were identified as neochlorogenic acid, cryptochlorogenic acid, rutin, rutin isomer, isoquercitin, kaempferol-3-O-rutinoside, kaempferol-3-O-glucoside, quercetin, quercetin isomer, and kaempferol. These compounds were present at significantly different levels among the accessions. The most abundant phenolic compound was rutin, which was found in 45 accessions where the total phenolic compound content was 18.75-46.51 mg/g (average 31.52 mg/g). The phenolic compounds were classified into four groups. The two accessions with the highest total phenolic compound content were from Ghana (PI286316) and Senegal (PI275413). The hierarchical cluster analysis of the 49 roselle accessions showed that they formed five groups according to their phenolic compound content. Our results will be useful for the selection of roselle genotypes with improved functional compounds.


Roselle (Hibiscus sabdariffa L.), which belongs to family Malvaceae, is a valuable medicinal crop originating from West Africa that contains a variety of functional compounds (Ali et al. 2005; Wang et al. 2014; Wang et al. 2016). Roselle is cultivated widely in Africa, Asia, and America for its leaves, seeds, stem, and calyces, which are used as food (Babatunde 2003; Sayago-Ayerdi et al. 2007; Ochani and D’Mello 2009). Roselle leaves have three to five lobes and they are arranged alternately on the stem (Daudu et al. 2015). The leaves are rich in flavonoids, which may contribute to their antioxidant capacity (Mohd-Esa et al. 2010; Wang et al. 2014; Wang et al. 2016). Roselle leaves are used traditionally for their diuretic, choleretic, febrifugal, and hypotensive effects (Lin et al. 2007; Kuo et al. 2012; Guardiola and Mach 2014). Extracts from roselle leaves have a variety of biological activities, including antioxidant, antitumor, antihyperammonemic, antiatherosclerotic, antifilarial, and antihyperlipidemic activities (Fernández-Arroyo et al. 2011; Yang et al. 2011; Chen et al. 2013; Guardiola and Mach 2014). Roselle leaves are used as a vegetable in soup and salad mainly in Africa, but are disregarded as a food source in many countries (Wang et al. 2014; Daudu et al. 2015; Wang et al. 2016).

Plants are used widely as sources of food and drugs and the phytochemical content of many of them has been investigated (Wang et al. 2016; Ryu et al. 2016; Kwon et al. 2018; Ryu et al. 2019). Phytochemicals are bioactive compounds that provide abundant health benefits (Zhen et al. 2016; Ryu et al. 2017a). The classification and analyses of phytochemicals have revealed new medicinal and industrial resources among naturally available materials that could be developed into profitable crops (Obouayeba et al. 2015; Ryu et al. 2017a; Kwon et al. 2018). Flavonoids and phenolic acid are bioactive compounds that have important pharmacological effects. They are abundant in roselle leaves where the major flavonoids are rutin, quercetin, and kaempferol and its derivatives (Chen et al. 2013; Wang et al. 2016; Zhen et al. 2016).

Breeders have developed new cultivars of various medicinal crops that are more nutritious and have improved levels of bioactive compounds (Sheela and Sheena 2014; Ryu et al. 2017a). Although the flavonoid content of roselle leaves has been studied (Wang et al. 2014; Wang et al. 2016; Zhen et al. 2016), a limited number of accessions was used and the flavonoid content of roselle grown in Korea has not been investigated so far. The aim of the present study was to investigate the flavonoid content of roselle leaves from worldwide accessions that could be grown commercially in Korea.


Plant materials

Forty-nine roselle accessions were used in this study (Table 1, Fig. 1). These accessions were obtained from the United States Department of Agriculture (USDA) in 2016 and were cultivated from 2016 to 2017 at the Korean Atomic Energy Research Institute (Jeongeup, Korea). The leaves of each accession were harvested in August 2017 for functional compound analysis. The leaves were handpicked from plants in three separate plots on the same plantation. Seeds were planted in plots (3 × 4.2 m) with alternate row spacings of 20 and 60 cm. Fertilizer (N:P:K 4:2:2 w/w/w) was applied at 600 kg/ha shortly after seedling. Manure was spread before planting, and the plants were not fertilized after planting.

Origin and leaf morphological characteristics of the 49 roselle accessions used in this study.

Fig. 1

Leaf morphological characteristics of the 49 roselle accessions used in this study. The numbers in the boxes correspond to the numbers in column 1 of Table 1.

Ultra-high performance liquid chromatography (UPLC) mass spectrometry (MS) analysis

Phenolic compounds were analyzed using a UPLC system (CBM-20A, Shimadzu Co., Kyoto, Japan) with two gradient pump systems (LC-30AD, Shimadzu), a UV-detector (SPD-M30A, Shimadzu), an auto sample injector (SIL-30AC, Shimadzu), and a column oven (CTO-30A, Shimadzu). Separation was achieved on an XR-ODS column (3.0 × 100 mm, 1.8 mm; Shimadzu, Japan) using a linear gradient elution program with a mobile phase containing solvent A (0.1% (v/v) trifluoroacetic acid in distilled deionized water) and solvent B (0.1% (v/v) trifluoroacetic acid in acetonitrile). Samples for UPLC analysis of phenolic compound content were ground using a grinder immediately prior to analysis. All samples were ground to achieve a particle size that passed through a 500 mL sieve. For the UPLC analysis, ground samples (1 g) were extracted in 5 mL water for 16 hours and filtrated through a 0.45 mm membrane filter. The phenolic compounds were separated using the following gradient: 0-5 minutes from 10%-15% solvent B; 5-10 minutes from 15%-20% solvent B; 10-15 minutes from 20%-30% solvent B; 15-25 minutes from 30%-50% solvent B; 25-30 minutes from 50%-75% B; 30-35 minutes from 75%-100% solvent B; 35-40 minutes from 100%-5% solvent B; and 40-45 minutes from 5%-0% solvent B. The phenolic compounds and anthocyanins were detected by MS at 280 nm and 520 nm, respectively. Chlorogenic acid (Sigma, USA), caffeic acid (Sigma, USA), delphinidin-3-sambubioside (Sigma, USA), delphinidin-3-glucoside (Sigma, USA), cyanidin-3-sambubioside (Sigma, USA) were determined by MS according to the retention times of commercial standards and UV–visible spectral characteristics.

Statistical analysis

The chemical analysis data were assessed by analysis of variance using a multiple comparisons method in the SPSS version 12 statistical software package (SPSS Inc., Chicago, IL, USA). Differences were considered to be significant at a = 0.05. When the treatment effect was significant, means were separated using Duncan’s multiple range test.

The clustering analysis of samples from the leaves of the 49 roselle accessions was performed using the complete linkage method in the SPSS software. The phenolic compounds were visualized as z-values in a heatmap.


Leaf morphological characteristics

The evaluation of leaf morphological characteristics is presented in Table 1 and Fig. 1. The leaf shapes of the roselle accessions were divided into five types: 24 accessions with palmatifid leaves, 10 accessions with palmatisect, 8 accessions with semi-palmatifid leaves, 6 accessions with trifid leaves, and 1 accession with entire leaves. All the roselle accessions had green leaves, except PI286312, which had purple leaves.

Phenolic compounds

The phenolic compounds in the leaves of the forty-nine accessions were analyzed by UPLC-MS. The chemical compounds identified in the roselle accessions are listed in Table 2. Fifteen compounds, delphinidin-3-O-sambubioside (D3S), neochlorogenic acid (NCA), cyanidin-3-O-sambubioside (C3S), cyanidin-3-O-glucoside (C3G), delphinidin-3-O-glucoside (D3G), cryptochlorogenic acid (CCA), rutin (RTN), rutin isomer (RTN-I), isoquercitin (IQN), kaempferol-3-O-rutinoside (K3R), kaempferol-3-O-glucoside (K3G), quercetin (QRN), quercetin isomer (QRN-I), kaempferol, and unknown compounds were detected. Four anthocyanins (D3S, C3S, C3G, D3G) were detected only in PI286312. Significant differences in phenolic compound content were detected among 11 roselle accessions (Table 3). RTN, NCA, and K3R were the most abundant phenolic compounds in roselle leaves from all the accessions. The NCA content ranged from 4.43-18.16 mg/100 g, and the highest level was in PI268100. The CCA content was highest in PI268100 (4.17 mg/100 g) and lowest in PI500724 (1.12 mg/100 g). The RTN content differed significantly among the roselle accessions and ranged from 0.66 mg/100 g (PI468409) to 19.47 mg/100 g (PI275413) (i.e., a 29.5-fold difference among accessions). The RTN-I content was highest in PI291128 and PI295592 (7.16 and 7.15 mg/100 g). The IQN content was highest in PI468409. The K3R content was highest in PI274247 (17.51 mg/100 g), and lowest in PI365477 (1.48 mg/100 g). The K3G content ranged from 0.38-3.06 mg/100 g (mean 1.23 mg/100 g) across all the accessions. The QRN content was highest in PI207920 and PI468409 (1.24 and 1.21 mg/100 g) and was relatively low in the other accessions. The QRN-I content ranged from 0-0.84 mg/100 g across all the accessions. The KFL content ranged from 0.05-0.96 mg/100 g across all the accessions and was highest in PI500734. The unknown compound content was approximately 2.9-fold higher in PI275413 than in all the other roselle accessions.

Details of phytochemical compounds identified in leaf extracts from the 49 roselle accessions used in this study.

Chemical hierarchical cluster analysis

The results of the hierarchical cluster analysis are presented in Fig. 2. The 49 roselle accessions clustered into five groups, which formed two independents supergroups (Groups I and II, and Groups III, IV, and V). Group I contained 16 accessions (PI500719, PI500734, PI468413 and PI500729, PI500725, PI500731, PI500727, PI500721, PI500732, PI591549, PI365477, PI291128, PI500752, PI500723, PI638933, and PI468409). Group II contained five accessions (PI256038, PI273388, PI263224, PI500724, and PI267778). Group III contained 17 accessions (PI180026, PI207920, PI500705, PI500747, PI500740, PI591551, PI256039, PI273391, PI500699, PI500720, PI500736, PI273459, PI500701, PI500706, PI265319, PI468412, and PI669506). Group IV contained six accessions (PI274245, PI500698, PI500737, PI496938, PI286312, and PI295592). Group V contained five accessions (PI268100, PI275413, PI286316, PI500713, and PI274247).

Fig. 2

Hierarchical cluster analysis of the 49 roselle accessions according to their phenolic compound content.

The chemical hierarchical cluster analysis divided the nine chemical compounds into four clusters. Cluster I contained four compounds (K3G, KFL, IQN, and RTN-I). Cluster II contained three compounds (QRN-I, Unknown, and K3R) compounds; Cluster III contained two compounds (QRN and RTN) as well as total content. Cluster IV contained two compounds (CCA and NCA).


Roselle originated in West Africa and has been cultivated for about 6000 years. It was introduced to Asia and South America from the 17th to 20th century (Ali et al. 2005; Da-Costa-Rocha et al. 2014; Daudu et al. 2015). Roselle leaves are a promising economic product, but information about their morphological traits and chemical composition is limited, which significantly impacts their affective applications. Morphological studies of roselle have been limited to the calyx of native Nigerian species (Daudu et al. 2015), and accessions collected from various countries are considered to be optimized in each environment (Wang et al. 2016). This is the first study at the functional compound level of roselle plants in Korea. In the kenaf plant, a Hibiscus species, its leaves also divided into three types, such as non-lobed (entire), shallow-lobed, deeply lobed (palmate). Ryu et al. (2017a) reported that kenaf has presented five major phenolic compounds, these highly correlated flowering date, hypocotyl and petal color rather than type of leaves. Our study, also observed no direct relationship among phenolic compounds level and leaf shape types.

Roselle is considered an important medicinal plant in Africa and Asia (Chen et al. 2013), Its leaves are used to treat colds, toothaches, and hangovers and have long been used in traditional medicines in Africa (Ali et al. 2005; Kuo et al. 2012; Guardiola and Mach 2014). The phenolic compounds are very important biological activates (Min et al. 2015; Deng et al. 2018; Ryu et al. 2019). Previous studies concluded that variations in phenolic compound content depend on the origin, harvest time, and cultivation conditions (Sheela and Sheena 2014; Wang et al. 2016; Ryu et al. 2016). High-performance liquid chromatography analysis of roselle leaves detected eight flavonoids (RTN, RTN-I, IQN, K3R, K3G, QRN, QRN-I, and KFL), four anthocyanins (D3S, C3S, C3G, and D3G) and two phenolic acids (NCA and CCA) compounds. Wang et al. (2016) identified nine compounds in the leaves of 31 different roselle accessions, including six flavonoids (RTN, IQN, K3R, K3G, QRN, and KFL) and three phenolic acids (NCA, CCA, and chlorogenic acid). In addition, the liquid chromatography/quadrupole-time-of-flight mass spectrometry method successfully identified five compounds (NCA, CCA, chlorogenic acid, RTN, and RTN-I) in eight roselle accessions (Wang et al. 2014). In the present study, the highest total phenolic compound content was found for PI275413. The phenolic compound RTN exhibits strong antioxidant properties in roselle leaves (Ochani and D’Mello 2009; Chen et al. 2013; Wang et al. 2014). We found that the RTN content was highest for PI275413, which was from Senegal, and PI286316, which was from Ghana, compared to the other accessions. Generally, environmental conditions are important factors that affect the phytochemical content of plants. Phenolic compound content is considerably influenced by soil, temperature, light, fertilizer conditions, and genetic background (Min et al. 2015; Ryu et al. 2017a; Deng et al. 2018). In this study, the plants were cultivated under the same conditions and the leaves were collected at the same time for phytochemical analysis, so any differences in phenolic compound content are likely genotype dependent, which may have implications for selection in breeding programs.

Anthocyanins play important roles in the response of plants to oxidative stress because of their ability to scavenge reactive oxygen species (Ryu et al. 2017b; Ryu et al. 2018). In this study, the anthocyanins D3S, C3S, C3G, and D3G were found in the roselle leaves, which may prevent cell damage and could be beneficial in the treatment of various human diseases (Obouayeba et al. 2014; Ryu et al. 2016; Ryu et al. 2018). This is the first report on the anthocyanin composition of roselle leaves. The predominant anthocyanins in roselle petals are D3S and C3S (Obouayeba et al. 2014), and two anthocyanins (C3G and D3G) were identified in calyces of roselle (Obouayeba et al. 2015). The identification of Hibiscus species as a source of functional compounds may result in an increase in consumption (Sheela and Sheena 2014; Wang et al. 2016; Ryu et al. 2018). Breeding novel cultivars with high flavonoid and anthocyanin content may be possible. Our results suggest that PI286312 is the most appropriate resource for food and drug industries because it contains high levels of anthocyanins.

The hierarchical cluster analysis grouped the 49 roselle accessions according to their chemical similarities. The chemical hierarchical cluster analysis provided information that can be used to select for chemotype in breeding programs and other useful information (Sheela and Sheena 2014; Ryu et al. 2019). In this study, we found high levels of chemical diversity among the 49 roselle accessions analyzed. Roselle accessions could be divided into five major groups, Group I mainly contained Zambia accessions and high isoquercitin content, Group II included five accessions with exhibited lack of all phenolic compounds. In particular, the highest RTN content and total phenolic content were found in Groups III and V, and RTN and NCA content was the main feature highlighted in the cluster analysis. This suggests that the RTN and NCA content of roselle leaves could be used as a marker to assess chemotypes. Although Group I was apparent pattern of geographical origin, but other groups were not exhibited. These results could be used in breeding programs to develop roselle cultivars with improved functional compounds.

Phenolic compoundz) content in leaf extracts from roselle accessions.


This work was supported by the research program of KAERI, Republic of Korea.


Ali BH, Al Wabel, Blunden G. 2005;Phytochemical, pharmacological and toxicological aspects of Hibiscus sabdariffa L.: a review. Phytother. Res. 19:369–375.
Babatunde FE. 2003;Intercrop productivity of Roselle in Nigeria. Af. Crop Sci. J. 11:43–47.
Chen JH, Wang CJ, Wang CP, Sheu JY, Lin CL, Lin HH. 2013;Hibiscus sabdariffa leaf polyphenolic extract inhibits LDL oxidation and foam cell formation involving upregulation of LXRa/ABCA1 pathway. Food Chem. 141:397–406.
Da-Costa-Rocha I, Bonnlaender B, Sievers H, Pischel I, Heinrich M. 2014;Hibiscus sabdariffa L. A phytochemical and pharmacological review.. Food Chem. 165:424–443.
Daudu OAY, Falusi OA, Dangana MC, Abubakar A, Yahaya SA, Abejide DR. 2015;Collection and evaluation of Roselle (Hibiscus sabdariffa L.) germplasm in Nigeria. Afr. J. Food Sci. 9:92–96.
Deng M, Deng Y, Dong L, Ma Y, Liu L, Huang F, et al. 2018;Effect of storage conditions on phenolic profiles and antioxidant activity of litchi pericarp. Molecules 23:E2276
Fernández-Arroyo S, Rodríguez-Medina IC, Beltrán-Debón RL, Pasini F, Joven J, Micol V, et al. 2011;Quantification of the polyphenolic fraction and in vitro antioxidant and in vivo antihyperlipemic activities of Hibiscus sabdariffa aqueous extract. Food Res. Int. 44:1490–1495.
Guardiola S, Mach N. 2014;Therapeutic potential of Hibiscus sabdariffa: a review of the scientific evidence. Endocrinol. Nutr. 61:274–295.
Kuo C, Kao E, Chan K, Lee H, Huang T, Wang C. 2012;Hibiscus sabdariffa L. extracts reduce serum uric acid levels in oxonate-induced rats. J. Funct. Foods 4:375–381.
Kwon SJ, Kim DK, Kim JM, Kang KY, Lee MK, Hong MJ, et al. 2018;Phytochemical compounds and antioxidant activity in the grain of selected faba bean (Vicia faba) genotypes. Plant Breed. Biotech. 6:65–73.
Lin T, Lin H, Chen C, Lin M, Chou M, Wang C. 2007;Hibiscus sabdariffa extract reduces serum cholesterol in men and women. Nutr. Res. 27:140–145.
Min K, Freeman C, Kang H, Coi SU. 2015;The regulation by phenolic compounds of soil organic matter dynamics under a changing environment. Biomed. Res. Int. 2015:825098
Mohd-Esa N, Hern FS, Ismail A, Yee CL. 2010;Antioxidant activity in different parts of roselle (Hibiscus sabdariffa L.) extracts and potential exploitation of the seeds. Food Chem. 122:1055–1060.
Obouayeba AP, Djyh NB, Diabate S, Djaman AJ, N'Guessan JD, Kone M, et al. 2014;Phytochemical and antioxidant activity of roselle (Hibiscus Sabdariffa L.) petal extracts. Res. J. Pharm. Biol. Chem. Sci. 5:1453–1465.
Obouayeba AP, Diarrassouba M, Soumahin EF, Kouakou H. 2015;Phytochemical analysis, purification and identification of Hibiscus anthocyanins. J. Pharm. Chem. Biol. Sci. 3:156–168.
Ochani PC, D'Mello P. 2009;Antioxidant and antihyperlipidemic activity of Hibiscus sabdariffa Linn. leaves and calyces extracts in rats.. Indian J. Exp. Biol. 47:276–282.
Ryu J, Kwon SJ, Jo YD, Jin CH, Nam BM, Lee SY, et al. 2016;Comparison of phytochemicals and antioxidant activity in blackberry (Rubus fruticosus L.) fruits of mutant lines at the different harvest time. Plant Breed. Biotech. 4:242–251.
Ryu J, Kwon SJ, Kim DG, Lee MK, Kim JM, Jo YD, et al. 2017a;Morphological characteristics, chemical and genetic diversity of kenaf (Hibiscus cannabinus L.) genotypes. J. Plant Biotechnol. 44:416–430.
Ryu J, Kwon SJ, Jo YD, Choi HI, Kang KY, Nam BM, et al. 2017b;Fruit quality and chemical contents of hybrid boysenberry (Rubus ursinus) lines developed by hybridization and gamma irradiation. Plant Breed. Biotech. 5:228–236.
Ryu J, Kwon SJ, Ahn JW, Kim SH, Lee SY, Kim JB, et al. 2018;Development of a stem-color mutant kenaf (Hibiscus cannabinus L.) cultivar, 'Jeokbong', and analysis of its functional compounds. Korean J. Hortic. Sci. Technol. 36:77–84.
Ryu J, Nam B, Kim BR, Kim SH, Jo YD, Ahn JW, et al. 2019;Comparative analysis of phytochemical composition of gamma-irradiated mutant cultivars of Chrysanthemum morifolium. Molecules 24:3003
Sayago-Ayerdi SG, Arranz S, Serrano J, Goni I. 2007;Dietary fiber content and associated antioxidant compounds in roselle flower (Hibiscus sabdariffa L.) beverage. J. Agric. Food Chem. 55:7886–7890.
Sheela VL, Sheena A. Tomlekova NB., Kozgar M., Wani MR, editors. 2014. Novel trends and achievements in breeding of tropical ornamental crops especially orchids and anthuriums: the mutation breeding approach. Mutagenesis: exploring genetic diversity of crops Wageningen Academic Publishers. Wageningen, Netherlands: p. 141–158.
Wang J, Cao X, Jiang H, Qi Y, Chin KL, Yue Y. 2014;Antioxidant activity of leaf extracts from different Hibiscus sabdariffa accessions and simultaneous determination five major antioxidant compounds by LC-Q-TOF-MS. Molecules 19:21226–21238.
Wang J, Cao X, Ferchaud V, Qi Y, Jiang H, Tang F, et al. 2016;Variations in chemical fingerprints and major flavonoid contents from the leaves of thirty-one accessions of Hibiscus sabdariffa L. Biomed. Chromatogr. 30:880–887.
Yang MY, Huang CN, Chan KC, Yang YS, Peng CH, Wang CJ. 2011;Mulberry leaf polyphenols possess antiatherogenesis effect via inhibiting LDL oxidation and foam cell formation. J. Agric. Food Chem. 59:1985–1995.
Zhen J, Villani TS, Guo Y, Qi Y, Chin K, Pan MH, et al. 2016;Phytochemistry, antioxidant capacity, total phenolic content and anti-inflammatory activity of Hibiscus sabdariffa leaves. Food chem. 190:673–680.

Article information Continued

Fig. 1

Leaf morphological characteristics of the 49 roselle accessions used in this study. The numbers in the boxes correspond to the numbers in column 1 of Table 1.

Fig. 2

Hierarchical cluster analysis of the 49 roselle accessions according to their phenolic compound content.

Table 1

Origin and leaf morphological characteristics of the 49 roselle accessions used in this study.

Lines No. Accession number Origin Leaf shape Leaf color
1 PI180026 India Palmatified Green
2 PI207920 Cuba Palmatisect Green
3 PI256038 Bangladesh Palmatisect Green
4 PI256039 Bangladesh Palmatified Green
5 PI263224 Zambia Palmatisect Green
6 PI265319 Cuba Palmatisect Green
7 PI267778 Sudan Trifid Green
8 PI268100 Nigeria Palmatified Green
9 PI273388 Taiwan Palmatisect Green
10 PI273391 Taiwan Palmatified Green
11 PI273459 Transvaal, Trifid Green
12 PI274245 Nigeria Semi-palmatified Green
13 PI274247 Poland Semi-palmatified Green
14 PI275413 Senegal Palmatified Green
15 PI286312 Ghana Palmatified Purple
16 PI286316 Ghana Palmatisect Green
17 PI291128 Ghana Trifid Green
18 PI295592 Niger Palmatisect Green
19 PI365477 Thailand Trifid Green
20 PI468409 United States Palmatisect Green
21 PI468412 United States Entire Green
22 PI468413 United States Palmatified Green
23 PI496938 Sudan Palmatified Green
24 PI500698 Zambia Palmatified Green
25 PI500699 Zambia Palmatified Green
26 PI500701 Zambia Semi-palmatified Green
27 PI500705 Zambia Palmatified Green
28 PI500706 Zambia Palmatified Green
29 PI500713 Zambia Palmatified Green
30 PI500719 Zambia Palmatified Green
31 PI500720 Zambia Palmatified Green
32 PI500721 Zambia Semi-palmatified Green
33 PI500723 Zambia Trifid Green
34 PI500724 Zambia Palmatified Green
35 PI500725 Zambia Palmatified Green
36 PI500727 Zambia Semi-palmatified Green
37 PI500729 Zambia Palmatified Green
38 PI500731 Zambia Semi-palmatified Green
39 PI500732 Zambia Semi-palmatified Green
40 PI500734 Zambia Palmatified Green
41 PI500736 Zambia Palmatified Green
42 PI500737 Zambia Palmatified Green
43 PI500740 Zambia Palmatified Green
44 PI500747 Zambia Palmatified Green
45 PI500752 Zambia Trifid Green
46 PI591549 Zambia Semi-palmatified Green
47 PI591551 Zambia Palmatified Green
48 PI638933 South Africa Palmatisect Green
49 PI669506 Tanzania Palmatisect Green

Table 2

Details of phytochemical compounds identified in leaf extracts from the 49 roselle accessions used in this study.

No. RT M+H+ Identification
1 12.89 597 Delphinidin-3-O-sambubioside (D3S)
2 12.97 355 Neochlorogenic acid (NCA)
3 13.05 581 Cyanidin-3-O-sambubioside (C3S)
4 13.20 449 Cyanidin-3-O-glucoside (C3G)
5 13.40 465 Delphinidin-3-O-glucoside (D3G)
6 13.56 355 Cryptochlorogenic acid (CCA)
7 14.64 611 Rutin (RTN)
8 14.91 611 Rutin isomer (RTN-I)
9 15.08 465 Isoquercitin (IQN)
10 15.24 595 Kaempferol-3-O-rutinoside (K3R)
11 15.75 449 Kaempferol-3-O-glucoside (K3G)
12 16.02 303 Quercetin (QRN)
13 16.91 303 Quercetin isomer (QRN-I)
14 18.37 287 Kaempferol (KFL)
15 19.79 449 Unknown

RT: Retention time, M+H+: Molecular ion value.

Table 3

Phenolic compoundz) content in leaf extracts from roselle accessions.

1 8.60c 1.39bc 16.75c 0.70de 1.81ij 3.78ef 2.60ab 0.67cd 0.26d 0.17cd 0.06de 36.79bc
2 7.47de 1.81bc 17.49ab 0.45ef 2.29ij 4.18de 0.71de 1.24a 0.44c 0.10de 0.02de 36.19bc
3 5.27gh 1.42bc 9.46fg 0.41ef 1.53ij 2.97fg 0.49de 0.57de 0.25d 0.07e 0.02de 22.47fg
4 5.95gh 1.80bc 16.48c 0.75cd 1.79ij 3.60ef 0.58de 0.78cd 0.24d 0.19cd 0.05de 36.19bc
5 5.51gh 1.91bc 6.74g 0.29f 1.45ij 3.37fg 0.51de 0.24f 0.12d 0.05e 0.01e 20.21fg
6 6.20fg 1.32bc 15.26cd 1.17cd 1.84ij 3.00fg 0.54de 0.99b 0.34cd 0.31bc 0.08de 31.05cd
7 6.51fg 1.45bc 7.14g 1.73bc 1.57ij 2.89fg 0.49de 0.05f 0.02d 0.37bc 0.13cd 22.36fg
8 18.16a 4.17a 8.60fg 0.64ef 1.49ij 3.01fg 0.46de 0.61cd 0.20d 0.13de 0.01e 37.47bc
9 4.98h 1.22c 8.40fg 0.87c 1.20j 2.49fg 0.38e 0.75cd 0.24d 0.19cd 0.06de 20.78fg
10 7.27de 1.76bc 17.42ab 1.60cd 2.09ij 3.89de 0.54de 0.48de 0.15d 0.53bc 0.14cd 35.73bc
11 6.86ef 1.71bc 17.40ab 1.01cd 3.85fg 4.44de 0.99cd 0.78cd 0.30cd 0.31bc 0.09de 37.75bc
12 9.00c 2.45b 7.55fg 1.91cd 5.14ef 3.68ef 1.89bc 0.08f 0.04d 0.84ab 0.29b 32.87cd
13 6.61fg 2.25bc 5.42g 2.19b 1.71ij 17.51a 3.06a 0.35f 0.84a 0.51bc 0.85a 41.31ab
14 10.95b 2.20bc 19.47a 1.92bc 3.80fg 5.68c 0.91cd 0.60cd 0.21d 0.62ab 0.16cd 46.51a
15 8.87c 1.92bc 12.05de 0.64ef 3.54gh 4.48de 1.47cd 0.29f 0.11d 0.19cd 0.04de 33.60cd
16 9.27c 1.81bc 19.33a 1.62bc 5.30ef 4.26de 0.94cd 0.04f 0.00e 0.79ab 0.15cd 43.52ab
17 5.51gh 1.71bc 1.21h 7.16a 4.69fg 2.43fg 2.07ab 0.07f 0.15d 0.59ab 0.20cd 25.79ef
18 4.43h 1.32bc 10.42ef 7.15a 4.70fg 2.45fg 2.10ab 0.67cd 0.19d 0.20cd 0.04de 33.61cd
19 7.22de 1.96bc 0.85h 1.71bc 6.60cd 1.48g 1.25cd 0.15ef 0.17d 0.89ab 0.16cd 22.44fg
20 6.31fg 1.72bc 0.66h 1.35cd 10.91a 7.04b 1.37cd 1.21a 0.06d 0.78ab 0.10de 31.49cd
21 4.62h 1.27bc 15.29cd 0.81cd 2.09ij 2.81fg 0.55de 0.21ef 0.23d 0.30bc 0.05de 28.23de
22 6.98e 1.89bc 6.39g 1.56bc 6.67cd 1.99g 1.56cd 0.12f 0.03d 0.74ab 0.18cd 28.10de
23 8.77c 2.05bc 10.48ef 1.55bc 8.77b 2.39fg 1.52cd 0.22ef 0.78ab 0.75ab 0.13cd 37.43bc
24 7.88d 2.16bc 9.05fg 1.58bc 7.61bc 2.44fg 1.53cd 0.20ef 0.06d 0.89ab 0.18cd 33.59cd
25 6.33fg 1.91bc 17.16b 1.51bc 2.55hi 4.75cd 0.86de 0.17ef 0.07d 0.72ab 0.23bc 36.26bc
26 7.32de 2.12bc 17.22b 1.87bc 2.98hi 4.23de 0.83de 0.15ef 0.13d 0.82ab 0.22bc 37.89bc
27 7.38de 1.89bc 17.67ab 0.72de 3.03hi 4.26de 1.18cd 0.02f 0.03d 0.28bc 0.07de 36.54bc
28 7.50de 2.28bc 18.12ab 1.76bc 2.77hi 5.01cd 0.99cd 0.25ef 0.22d 0.75ab 0.23bc 39.87ab
29 10.44b 2.15bc 17.43ab 0.92cd 3.27gh 5.64c 1.15cd 0.13f 0.10d 0.21cd 0.07de 41.51ab
30 7.00ef 1.92bc 6.49g 0.98cd 6.17de 2.34fg 1.63cd 0.16ef 0.13d 0.47bc 0.12de 27.40ef
31 6.92ef 1.98bc 16.58c 0.72de 2.47hi 4.42de 0.79de 0.02f 0.11d 0.22cd 0.06de 34.29cd
32 6.92ef 2.24bc 6.10g 1.21cd 6.32de 2.74fg 2.59ab 0.04f 0.05d 0.52bc 0.15cd 28.88de
33 6.33fg 1.87bc 6.05g 1.17cd 6.21ef 2.67fg 1.89bc 0.16ef 0.14d 0.61ab 0.19bc 27.28ef
34 5.61gh 1.12c 7.11g 0.56ef 1.48ij 2.13fg 0.41de 0.09f 0.08d 0.13de 0.03de 18.75g
35 5.66gh 1.32bc 6.63g 1.42cd 6.35de 1.59g 1.08cd 0.11f 0.09d 0.77ab 0.14cd 25.15ef
36 6.77ef 1.71bc 6.94g 1.26cd 7.94bc 1.79g 1.33cd 0.11f 0.08d 0.60ab 0.11de 28.64de
37 6.50fg 1.91bc 7.13g 1.43cd 6.63cd 2.15fg 1.90bc 0.04f 0.04d 0.63ab 0.14cd 28.50de
38 6.50fg 1.64bc 6.32g 0.86cd 6.81cd 1.91g 1.83bc 0.04f 0.04d 0.39bc 0.09de 26.43ef
39 5.74gh 1.64bc 6.17g 1.41cd 6.02ef 1.90g 1.86bc 0.04f 0.04d 0.79ab 0.18cd 25.78ef
40 6.06g 1.68bc 6.37g 1.51bc 6.52cd 2.09fg 1.57cd 0.13f 0.12d 0.96a 0.24bc 27.25ef
41 7.12de 1.67bc 16.81bc 0.70de 2.58hi 3.74ef 0.61de 0.08f 0.08d 0.21cd 0.05de 33.66cd
42 6.44fg 2.07bc 7.83fg 1.32cd 8.29b 2.18fg 2.28ab 0.06f 0.04d 0.88ab 0.19bc 31.57cd
43 8.58c 1.68bc 17.58ab 0.88cd 2.67hi 4.18de 0.73de 0.13f 0.12d 0.24bc 0.06de 36.86bc
44 7.45de 2.16bc 17.19b 1.31cd 2.54hi 4.38de 0.85de 0.18ef 0.16d 0.37bc 0.10de 36.69bc
45 6.09g 1.53bc 6.16g 0.71de 6.08de 2.05fg 1.55cd 0.12f 0.09d 0.34bc 0.08de 24.78ef
46 5.50gh 1.45bc 6.23g 0.61ef 6.28de 1.88g 1.61cd 0.04f 0.03d 0.27bc 0.06de 23.97ef
47 7.09de 1.75bc 18.92ab 0.66de 3.26gh 3.25ef 0.58de 0.14f 0.12d 0.22cd 0.03de 36.03bc
48 6.60fg 1.52bc 5.49g 1.29cd 5.71ef 1.66g 1.13cd 0.10f 0.06d 0.49bc 0.10de 24.17ef
49 6.76fg 1.45bc 12.83d 1.14cd 1.73ij 2.76fg 0.57de 0.15ef 0.68b 0.37bc 0.09de 28.53de

NCA: Neochlorogenic acid, CCA: Cryptochlorogenic acid, RTN: Rutin, RTN-I: Rutin isomer, IQN: Isoquercitin, K3R: Kaempferol-3-O-rutinoside, K3G: Kaempferol-3-O-glucoside, QRN: Quercetin, QRN-I: Quercetin isomer, KFL: Kaempferol. Data are shown as the mean values; n = 3.