Diversity of Mango (Mangifera Indica L.) Cultivars Based on Physicochemical, Nutritional, Antioxidant, and Phytochemical Traits in South West Nigeria

ABSTRACT

Evaluation of chemical characteristics of fruit crops has been successfully used for the selection of cultivars for breeding programs. In a bid to access useful information for the utilization of mango genetic resources in Nigeria, 10 mangoes (Mangifera indica L.) cultivars were evaluated for physicochemical, nutritional, antioxidant, and phytochemical characteristics at the National Horticultural Research Institute, Ibadan, Nigeria. Cultivars showed significant (P ≤ 0.05) variation for quantitative and chemical traits. Madoe, Saigon, Ogbomoso, and Haden had the highest values for TSS (19.25% brix), vitamin C (11.68%), pH (5.91), and shelf life (12.00), respectively. Alphonso had the highest values for starch – 1.73; reducing sugar – 0.61; amylose – 11.78 g/100 g, and some minerals (K – 20.15; Mn – 7.79; Ca – 51.50; Fe – 4.99; Zn – 32.95; Cu – 11.10, and Cr – 0.47 mg/100 g) composition, but low values for phytochemicals (phytate – 0.86 and tannin – 0.54 mg/100 g). The highest antioxidant values were recorded for Julie (DPPH inhibition 85.24%; reducing power – 2.46) and Saigon (total antioxidant 322.14 mg/100 g vitamin C equivalent). However, the least genetic similarity (2.41) was recorded between Palmer and Kent, while the highest genetic similarity (6.68) was between Ogbomoso and Edward. The total variation explained by the first four principal component (PC) axes was 71%. Results obtained from this study indicate that the mango cultivars are divergent and can be useful genetic resources for mango improvement through breeding.

KEYWORDS:

• Variation
• genetic diversity
• germplasm
• cultivars
• chemical characteristics

Introduction

Mango (Mangifera indica L.) is one of the most cultivated fruit in the tropical and subtropical regions. Mango is native to South Asia, wherefrom it has spread worldwide to become one of the most extensively exploited fruits (Hoque et al., 2018). It belongs to the family Anacardiaceae and is considered as the King of all fruits. In 2017, world mango production was about 50.65 million metric tons (FAOSTATS, 2019; Statista, 2019), occupying an area of over 4.37 million ha. Mango is now grown commercially in more than 111 countries (Rymbai et al., 2013) with Nigeria ranking ninth among these mango-producing countries (Ugese et al., 2012) and Benue state-ranking first (1st) in the league of states that produce mangos in Nigeria (Ubwa et al., 2014). The success of mango cultivation in Nigeria can be attributed to the diverse and favorable environmental conditions across the country.

Mango is a valuable and popular fruit possessing rich dietary source (carbohydrates, fiber minerals), antioxidants such as vitamin C, carotenoids, and phenolic compounds which have shown various health benefits (Gámez et al., 2017; Liu et al., 2014). The increasing rate of malnutrition in Nigeria (USAID, 2017) has resulted in a growing interest in functional foods that can provide not only the basic nutritional and energy requirements but also additional physiological benefits. A functional food is known to produce a beneficial effect in one or more physiological functions and protect against disease via several mechanisms (Apak et al., 2007; Rymbai et al., 2013). Mango has been assigned as a functional food because it provides the human diet with macro- and micronutrients along with a large pool of bioactive compounds (Jash and Brahmachari, 2015). Various parts of the plant such as fruit (pulp, peel, and stone), leaves, and bark are rich sources of fiber and bioactive compounds (Monribot-Villanueva et al., 2019; Rymbai et al., 2013). The fruits are specifically rich in antioxidants known to reduce the risk of cardiac disease with the potential of anticancer and antiviral activities (Jash and Brahmachari, 2015).

Mango has been reported to have extensive diversity due to alloploidy, outbreeding, continuous grafting, and phenotypic differences arising from varied agro-climatic conditions in different mango-growing regions (Ravishankar et al., 2000). Several works have been reported on genetic diversity among species of mango using SSR markers (Kumar et al., 2013; Sennhenn et al., 2014), RAPD markers (Neetu et al., 2017; Souza et al., 2011), and morphological markers (Avendaño-Arrazate et al., 2017; Carolyn et al., 2015). Meanwhile, the chemical composition of mango fruits varies with regard to different cultivars, area of production, and stage of maturity (Abbasi et al., 2011; Omer and Soad, 2014; Tharanathan et al., 2006). Therefore, assessment of genetic diversity based on chemical composition among mango would assist in identifying and selecting cultivars with desirable traits for incorporation into crop improvement program. Thus, this study was conducted in order to access the genetic diversity of 10 mango cultivars in Nigeria based on their physicochemical, nutritional, antioxidant, and phytochemical characteristics. Results obtained in this study would be a useful guide for the identification of potential parents in mangobreeding programs.

Materials and Methods

The experiment was carried out in 2016 at the National Horticultural Research Institute, Ibadan, Nigeria. Ten commonly grown mango cultivars were selected for sampling at the Mango orchard of the institute. Tagging of 50 inflorescences in each tree was made using labels showing dates of flowering and the fruits were monitored to maturity. A sample size consisting of 30 fruits per cultivar (5 representative fruits from 6 trees per cultivar) were picked at stage 4 of maturity, characterized by the full shoulder at the stem end and yellow-orange color on 25% of the fruit surface. The maturity stages were defined according to days after flowering: Stage 1 = 100 days after flowering; Stage 2 = 110 days after flowering; Stage 3 = 120 days after flowering; Stage 4 = 130 days after flowering; Stage 5 = 140 days after flowering. Only fruits that did not show any visual signs of bruises, cuts, infestation were selected for the study.

The fruits were labeled according to the trees of origin, transported to the laboratory, and were washed thoroughly. Fifteen healthy fruits of each cultivar were stored at room temperature out of which five was used for the morphological evaluation. They were evaluated for eight quantitative traits (fruit width (cm), fruit thickness (cm), fruit length (cm), fruit weight (g), Peel (%), Pulp (%), Fiber length (mm), Stone (%), and fruit color following the mango IPGRI descriptor) (IPGRI, 2006). Ten fruits were peeled to remove the skin, and pulps were sliced into thin sizes and divided into two portions. A portion was homogenized in a blender (Sharp blender model; EM-11) for 5 min and the speed switch was set at number II. The resulting juice was stored in plastic bottles and preserved in a refrigerator at 15°C for the determination of total soluble solids (TSSs), vitamin C, pH, and moisture content. The other portion was dried in a convection oven at 55°C until a constant weight was obtained and subsequently ground to a smooth powder. Milled samples were stored in airtight containers and kept at −20°C for proximate, antioxidant activity, and phytochemical analysis. Chemicals for antioxidant assays were supplied by Sigma Aldrich, UK. Standard analytical procedures with references to support were used for all analyses. TSSs were determined as °Brix using a refractometer ATAGO handheld type (ELCOMP (PTY) LTD Midrand, South Africa), which was calibrated with distilled water and maintained at a constant temperature of 25°C. The pH was measured using a digital pH meter (Delta 320 pH meter, Mettler Toledo Instruments Co., Ltd., Shanghai, China).

The moisture content of the juiced pulp was determined using standard methods (AOAC, 2000). Vitamin C content was determined by titration using 2,6-dichlorophenol solution as described by AOAC (2000). Results from this analysis were expressed per fresh weight (FW).

Mango shelf life was determined using fruits harvested at the hard green stage of maturity. The fruits at commercial maturity with uniform size, shape, and free of any visible defects, disease symptoms, and insect infestations were transported to the laboratory and used.

Dried milled samples were carried out for proximate analysis to determine the starch, reducing sugar, and amylose as per the method of Nzikou et al. (2010), ash contents (Egbekunle et al., 2017). Crude fat and crude fiber were determined using the standard analytical method of the Association of Official Analytical Chemists (AOAC, 2000).

The mineral composition of the samples was determined according to the Association of Official Analytical Chemists methods (AOAC, 2005). Phosphorus was determined colorimetrically using a spectrophotometer (Spectronic 20, Gallenkamp, London, U.K.) (Egbekunle et al., 2017) with KH₂PO₄ as the standard. Calcium, magnesium, manganese, iron, chromium, zinc, and copper were determined using an atomic absorption spectrophotometer (Model SP9, Pye Unicam Ltd., Cambridge, U.K.). Potassium was determined using a flame photometer (410 Corning) according to the method of AOAC (2005).

Methanol extracts of dried mango pulp flour of 10 mg·mL⁻¹ (dry weight) were used at appropriate concentrations to assess antioxidant activity. The 2,2,-diphenyl-1-picrylhydrazyl (DPPH) radical scavenging activity (RSA) was determined by the method of Zhang and Hamauzu (2004). RSA was expressed as percentage inhibition to DPPH radical and calculated using the formula:

% RSA = [1 – (AC/AD)] × 100, where AC is the absorbance of the solution at 517 nm when the extract was added; and AD is the absorbance of the DPPH solution.

Total Antioxidant capacity assay was based on the reduction of Mo (VI) to Mo (V) by the extract and subsequent formation of a green phosphate/Mo V complex at acid pH. The absorbance of the solution was measured at 695 nm against a blank (Banerjee et al., 2015).

Reducing power was determined according to the method of Egbekunle et al. (2017). Various concentrations of methanolic extracts (2.5 mL) were mixed with 2.5 mL of 200 mmol·L⁻¹ sodium phosphate buffer (pH 6.6) and 2.5 mL of 1% potassium ferricyanide. The mixture was incubated at 50°C for 20 min. Absorbance was measured at 700 nm using the UnicoTM UV-2100 spectrophotometer.

Total phenolic (TP) content was determined using the Folin-Ciocalteu colorimetric method (Egbekunle et al., 2017). Absorbance was measured at 760 nm and TPs content was expressed as mg/100 g DW of gallic acid equivalent (GAE).

Total flavonoid (TF) content was determined according to a colorimetric assay (Egbekunle et al., 2017). Absorbance was read at 510 nm with a spectrophotometer (model 6400/6405, Jenway, Missouri City, Tex, U.S.A.). The TFcontent was expressed as mg/100 g DW of catechin equivalent (CE).

Phytate and tannin content was determined by the method of Egbekunle et al. (2017). For phytate, 3% trichloroacetic acid was added to 5 g of sample. Phytate phosphorus was calculated from the ratio of Fe:P in the ferric phytate as 4:6 (Fe:P). Phytic acid content was determined by assuming the empirical formula C₆P₅O₂₄H₁₈.

For tannin content, finely ground samples (0.5 g) were defatted with 5% ethyl ether for 15 min. Tannin in defatted samples was extracted with methanol, and absorbance at 760 nm was determined.

Statistical Analysis

Data were subjected to Analysis of Variance using SAS (version 9.3, 2011) and means were separated using Duncan multiple range test at P ≤ 0.05. Multivariate Analysis of variance (MANOVA) was employed using PROC GLM in SAS. The principal component analysis was used to summarize the standardized data from the biochemical composition analysis. Data matrix of the cultivars was submitted to SAS (version 9.3, 2011) to investigate similarity among the genotypes using Gower genetic distance.

Results

Results showed significant (P ≤ 0.01) variation among cultivars for all the components measured except for DPPH (Table 1).

Table 1. Analysis of variance for quantitative, physicochemical, proximate, minerals, antioxidants, and phytochemical composition of mango

Source	Fruit length	Fruit width		Fruit thickness		Fruit weight		Peel		Pulp		Stone		Fiber length
Quantitative
Cultivars	9.87**	0.53**		5.27**		95990.72**		28.12**		152.61**		71.478*		3249.20**
Mean	13.4	2.69		8.07		432		19.82		65.4		11.61		59.8
Cv	2.82	4.23		3.21		2.75		9.11		4.23		2.45		3.16
Source	Tss	Vitamin C		Ph		Shelf life
Physicochemical
Cultivars	20.17**	7.77**		0.77**		9.63**
Mean	15.1	9.42		5.23		10.1
Cv	5.23	3.45		4.13		2.90
Source	Starch		Reducing sugar		Amylose		Fat		Moisture		Ash	Crude fiber
Proximate
Cultivars	0.19**		0.12**		0.49**		0.21**		2.04**		0.75**	0.06**
Mean	1.14		0.43		10.99		0.45		79.14		1.90	1.03
Cv	5.68		2.46		1.91		18.31		4.98		9.19	8.23
Source	K		Mg		Ca		Fe		Zn		Cu		P		Cr	Mn
Minerals
Cultivars	1007015.40**		106829.13**		27415.67**		50534.53**		126.79**		13.54**		3790.46**		4.39**	5728.89**
Mean	1299.95		295.40		372.15		173.42		18.60		7.90		242.60		2.00	104.37
Cv	1.01		1.08		2.25		1.04		1.32		2.48		1.82		10.05	1.02
Source	DPPH		Total Antioxidant		Reducing power		Total Polyphenol		Flavonoid		Phytate		Tannin
Antioxidants and Phytochemicals
Cultivars	80.15ns		0.84**		0.28**		00.02**		0.04**		0.30**		0.004**
Mean	76.52		2.34		1.79		1.15		0.18		1.45		0.583
Cv	9.86		3.65		6.31		2.00		18.82		6.95		3.127

Statistical significance was calculated by Tukey @ P = 0.05.

Table 2 shows variation among the quantitative traits of mango. Lipen had the highest mean values for fruit width (3.34 cm) and fruit weight (683.33 g), while the highest mean value for fruit thickness was observed in Kent (10.19 cm). Palmer had the highest fruit length (16.70 cm) when compared with the others. Among the 10 mango cultivars, Saigon had the highest percentage peel (23.45%), while Julie had the highest percentage of pulp (76.26%) and fiber length (115.00 mm). However, Ogbomoso had the highest value for stone (22.62%).

Table 2. Variability among the genotypes for some quantitative traits of mango

Cultivars	Fruit width (cm)	Fruit thickness (cm)	Fruit length (cm)	Fruit weight (g)	Peel (%)	Pulp (%)	Fiber length (mm)	Stone (%)
Ogbomoso	2.22i ± 0.06	6.83i ± 0.07	12.47g ± 0.08	221.12i ± 1.12	22.63c ± 0.09	51.60j ± 0.02	35.00g ± 1.23	22.62a ± 0.87
Madoe	2.43g ± 0.07	7.19g ± 0.02	11.57j ± 0.05	302.97g ± 1.78	23.22b ± 0.04	63.88g ± 0.05	75.00c ± 1.06	11.35d ± 0.45
Palmer	2.72f ± 0.04	8.08f ± 0.05	16.70a ± 0.09	533.77d ± 1.56	21.25e ± 0.21	63.71h ± 0.07	35.00g ± 1.09	9.18f ± 0.24
Saigon	2.07j ± 0.09	6.01j ± 0.060	12.50f ± 0.03	218.86j ± 1.87	23.45a ± 0.24	65.68e ± 0.09	26.00h ± 1.31	12.72c ± 0.28
Lipen	3.34a ± 0.02	9.89b ± 0.01	13.90d ± 0.05	683.33a ± 0.98	18.25g ± 0.56	65.53f ± 0.04	59.00e ± 0.98	11.22e ± 0.16
Kent	3.15b ± 0.02	10.19a ± 0.07	16.07b ± 0.07	663.29b ± 1.45	16.25i ± 0.78	72.93b ± 0.05	36.00f ± 1.26	8.64g ± 0.31
Edward	3.04c ± 0.05	8.81c ± 0.08	13.33e ± 0.01	570.00c ± 1.24	18.38f ± 0.12	57.41i ± 0.02	113.00b ± 1.08	7.23j ± 0.67
Julie	2.88d ± 0.01	8.41d ± 0.03	14.13c ± 0.06	485.08e ± 1.38	15.17j ± 0.17	76.26a ± 0.06	115.00a ± 0.71	7.65i ± 0.18
Haden	2.75e ± 0.07	8.13e ± 0.04	11.73h ± 0.04	383.33f ± 1.69	22.13d ± 0.19	67.01d ± 0.09	69.00d ± 1.27	8.29h ± 0.47
Alphonso	2.34h ± 0.08	7.17h ± 0.09	11.67i ± 0.06	258.29h ± 1.18	17.49h ± 0.20	69.99c ± 0.03	35.00g ± 1.08	17.19b ± 0.17

Means within the same column with different alphabets are significantly (p ≤ 0.05) different.

Values are mean ±  Standard error.

Saigon had the least fruit width (2.07 cm), fruit thickness (6.01 cm), fruit weight (218.86 g), and fiber length (26.00 mm). Lowest fruit length (11.57 cm), peel (15.17%), pulp (51.60%), and stone (7.23%) were obtained in Madoe, Julie, Ogbomoso, and Edward, respectively.

Genetic diversity is assessed in fruit crops based on differences in qualitative traits. In many cases, the qualitative traits have been used as a powerful tool in the classification of cultivars. Mango fruitt of the different variety varied in shape and color as seen in Table 3. In this study, the color of Mango fruit skin at maturity varied from green in Ogbomoso, Saigon to yellow green in Julie to red blush in Haden. Skin coloration of mature fruits may be due to anthocyanins that develop when tisssues are exposed to light.

Madoe, Saigon, Ogbomoso, and Haden had the highest values for TSS (19.25% brix), vitamin C (11.68%), pH (5.91), and shelf life (12.00), respectively (Table 4). In general, most varieties had a shelf life of above 9 days.

Table 3. Fruit morphological characteristics at maturity

Mango	Color of skin at maturity	Fruit shape
Ogbomoso	Green	Oval
Madoe	Reddish green	Oval
Palmer	Reddish green	Oblong-ovate
Saigon	Green	Oval
Lipen	Yellowish green with blush	Oval
Kent	Reddish green	Round-ovate
Edward	Reddish yellow	Oval
Julie	Yellowish green	Oval
Haden	Green with red blush	Round-ovate
Alphonso	Golden yellow	Oval

Table 4. Physicochemical properties of mango

Cultivars	TSS ^a (% brix)	Vitamin C (%)	pH	Shelf life
Ogbomoso	18.50b ± 0.03	10.93c ± 0.32	5.91a ± 0.65	9.00d ± 0.98
Madoe	19.25a ± 0.34	10.89d ± 0.23	4.60j ± 0.47	11.00b ± 1.24
Palmer	11.25j ± 0.03	8.14h ± 0.14	5.38e ± 0.41	11.00a ± 0.91
Saigon	14.25f ± 0.09	11.68a ± 0.56	5.86b ± 0.61	11.00b ± 0.83
Lipen	15.25e ± 0.06	8.44f ± 0.18	4.67i ± 0.28	9.00d ± 0.89
Kent	12.50i ± 0.02	9.26e ± 0.13	5.52d ± 0.71	10.00c ± 0.64
Edward	13.25h ± 0.05	7.51j ± 0.41	4.74h ± 0.19	12.00a ± 0.75
Julie	15.50d ± 0.03	11.28b ± 0.25	5.14f ± 0.23	10.00c ± 0.49
Haden	17.25c ± 0.04	7.72i ± 0.16	5.72c ± 0.42	12.00a ± 0.83
Alphonso	14.00g ± 0.09	8.35g ± 0.32	4.85g ± 0.25	6.00d ± 0.67

TSS = Total soluble solid.

Means within the same column with different alphabets are significantly (p ≤ 0.05) different.

Values are mean± Standard error.

Palmer, Edward, Madoe, and Alphonso had the lowest values for TSS (11.25% brix), vitamin C (7.51%), pH (4.60), and shelf life (6.00), respectively.

Ogbomoso had the highest moisture (80.44%) and crude fiber (1.24 g/100 g) contents (Table 5). Palmer and Edward had the highest ash (2.67 g/100 g) and fat (0.90 g/100 g) contents, respectively. Alphonso had the highest starch (1.73 g/100 g), reducing sugar (0.61 g/100 g), and amylose (11.78 g/100 g) contents. Edward, Madoe, and Palmer, respectively, had the lowest moisture (77.87%), starch (0.92 g/100 g), and crude fiber (0.82%) contents. The lowest ash (1.47 g/100 g) and fat (0.19 g/100 g) contents were recorded in Alphonso, while Ogbomoso had the lowest reducing sugar (0.09 g/100 g) and amylose (10.32 g/100 g) contents.

Table 5. Variability among the genotypes for the proximate composition of mango (g/100 g)

Cultivars	Moisture	Ash	Starch	Reducing sugar	Amylose	Crude fat	Crude fiber
Ogbomoso	80.44a ± 2.08	2.66a ± 0.12	0.94e ± 0.06	0.09 f ± 0.00	10.32 f ± 0.64	0.58b ± 0.13	1.24a ± 0.24
Madoe	78.96f ± 1.24	1.88b ± 0.13	0.92e ± 0.05	0.10 f ± 0.02	11.15bc ± 0.12	0.32c ± 0.01	0.87i ± 0.18
Palmer	79.70c ± 1.73	2.67a ± 0.27	0.93e ± 0.11	0.27e ± 0.01	10.68e ± 0.03	0.55b ± 0.06	0.82j ± 0.15
Saigon	79.42d ± 1.45	1.69bc ± 0.20	0.98de ± 0.06	0.40d ± 0.01	11.05dc ± 0.03	0.21c ± 0.03	1.05f ± 0.43
Lipen	78.28i ± 1.85	1.54c ± 0.23	1.10c ± 0.04	0.55 c ± 0.01	11.42b ± 0.03	0.22c ± 0.02	0.98g ± 0.65
Kent	79.20e ± 73	1.57c ± 0.12	1.07cd ± 0.06	0.55bc ± 0.00	10.90cde ± 0.01	0.83a ± 0.08	1.11d ± 0.19
Edward	77.87j ± 1.42	2.46a ± 0.22	1.13c ± 0.08	0.56bc ± 0.01	10.76de ± 0.10	0.90a ± 0.17	1.17b ± 0.27
Julie	78.83g ± 1.47	1.55c ± 0.07	1.28b ± 0.01	0.57b ± 0.01	10.92cde ± 0.02	0.19c ± 0.05	1.16c ± 0.61
Haden	80.21b ± 1.63	1.49c ± 0.16	1.31b ± 0.04	0.60a ± 0.02	10.88cde ± 0.03	0.50b ± 0.10	0.93h ± 0.22
Alphonso	78.50h ± 1.29	1.47c ± 0.15	1.73a ± 0.08	0.61a ± 0.01	11.78a ± 0.04	0.19c ± 0.02	1.06e ± 0.46

Means within the same column with different alphabets are significantly (p ≤ 0.05) different.

Values are mean± Standard error.

Significant variation was also recorded for mineral composition among mango cultivars (Table 6). Alphonso had significantly (p ≤ 0.05) higher values for most minerals, except P and Mn. The lowest value for K was recorded in Madoe, Mg and Zn in Kent; Ca, Cu, P, and Mn in Haden; and Fe and Cr compositions in Saigon (Table 6).

Table 6. Variability among the genotypes for the mineral composition of mango (mg/100 g)

Cultivars	K	Mg	Ca	Fe	Zn	Cu	P	Mn	Cr
Ogbomoso	16.85c ± 8.66	3.21d ± 0.58	38.05e ± 0.29	1.79c ± 0.29	23.05b ± 0.14	8.44d ± 0.06	247.00d ± 0.01	0.70h ± 0.35	0.36b ± 0.05
Madoe	6.88i ± 4.33	3.33c ± 1.44	49.20b ± 4.33	1.11g ± 1.44	23.20b ± 0.23	8.37d ± 0.03	298.50a ± 2.02	1.25d ± 0.58	0.16de ± 0.03
Palmer	16.50d ± 2.89	2.07g ± 1.15	33.00f ± 4.62	1.20ef ± 1.15	16.30e ± 0.06	6.24e ± 0.01	246.00d ± 4.04	0.73g ± 0.23	0.13ef ± 0.00
Saigon	16.35d ± 2.89	2.32f ± 0.58	26.50g ± 2.31	0.74i ± 0.78	13.40f ± 0.06	10.60b ± 0.12	271.00c ± 0.58	1.10e ± 0.29	0.09g ± 0.02
Lipen	9.23h± 7.51	1.54i ± 1.15	40.95d ± 10.10	1.18f ± 0.29	19.75c ± 0.14	8.15d ± 0.13	230.00e ± 1.73	1.79a ± 0.87	0.13ef ± 0.05
Kent	11.19g ± 0.01	1.10j ± 0.29	33.50f ± 4.04	2.92b ± 1.73	10.75g ± 0.09	5.95ef ± 0.10	219.50f ± 0.87	1.29c ± 0.87	0.12fg ± 0.02
Edward	11.45f ± 8.66	1.81h ± 0.87	33.45f ± 4.91	0.87h ± 1.70	16.00e ± 0.06	5.77f ± 0.03	229.00e ± 2.31	0.68i ± 0.09	0.18d ± 0.32
Julie	17.20b ± 17.32	2.73e ± 4.91	44.55c ± 6.64	1.34d ± 0.01	17.50d ± 0.12	9.49c ± 0.23	285.00b ± 5.77	1.60b ± 1.15	0.14ef ± 0.05
Haden	14.20e ± 5.77	3.65b ± 0.58	21.50h ± 1.15	1.22e ± 0.87	13.10f ± 0.23	4.91g ± 0.08	176.50g ± 1.44	0.44j ± 0.17	0.23c ± 0.14
Alphonso	20.15a ± 0.01	7.79a ± 1.73	51.50a ± 0.00	4.99a ± 0.01	32.95a ± 0.14	11.10a ± 0.15	223.50ef ± 0.29	0.83f ± 0.58	0.47a ± 0.04

Means within the same column with different alphabets are significantly (p ≤ 0.05) different.

Values are mean± Standard error.

Similarly, significant differences (p ≤ 0.05) were recorded for antioxidant and phytochemical compositions among cultivars (Table 7). Julie exhibited the highest DPPH activity (85.24%), reducing power (2.46), and TP content (123 mg/100 g), while Saigon had the highest total antioxidant activity. The highest TF content was obtained in Palmer, while phytate and tannin compositions were highest in Ogbomoso. Haden recorded the lowest DPPH and total antioxidant activities. Ogbomoso had the lowest value for reducing power and TF content. Kent had the lowest TPs, while Alphonso had the lowest phytate and tannin contents.

Table 7. Variability among the genotypes for the antioxidant and phytochemical composition of mango

Cultivars	DPPH inhibition (%)	IC 50	Total antioxidant (mg/100 g vitamin C equivalent)	IC 50	Reducing power (absorbance)	IC 50	Total Polyphenol (GAE equivalent mg/100g)	Flavonoid (CE equivalent mg/100g)	Phytate (mg/100g)	Tannin (mg/100g)
Ogbomoso	71.46b ± 4.40	1.34	182.00gh ± 0.02	0.856	1.35e ± 0.01	0.496	112.00bc ± 0.01	6.00g ± 0.00	1.87a ± 0.05	0.63a ± 0.00
Madoe	72.33b ± 2.30	1.32	190.05fg ± 0.03	0.820	1.50de ± 0.00	0.551	113.00bc ± 0.01	8.00fg ± 0.00	1.85a ± 0.09	0.61ab ± 0.01
Palmer	72.69ab ± 7.46	1.32	294.25b ± 0.02	0.529	1.62cd ± 0.01	0.595	115.00b ± 0.01	48.00a ± 0.01	1.67b ± 0.04	0.61ab ± 0.03
Saigon	79.71ab ± 3.69	1.20	322.14a ± 0.05	0.484	1.75c ± 0.02	0.643	120.00a ± 0.01	23.00b ± 0.00	1.50c ± 0.03	0.62ab ± 0.01
Lipen	79.46ab ± 3.40	1.21	236.32d ± 0.03	0.659	1.81c ± 0.00	0.665	121.00a ± 0.02	21.00bc ± 0.06	1.49c ± 0.13	0.59bc ± 0.01
Kent	79.83ab ± 8.02	1.20	205.16e ± 0.09	0.759	1.76c ± 0.03	0.647	96.00d ± 0.02	19.00bcd ± 0.00	1.42c ± 0.03	0.57cd ± 0.01
Edward	81.55ab ± 0.37	1.17	263.91c ± 0.06	0.590	2.067b ± 0.17	0.757	122.00a ± 0.01	16.00cde ± 0.00	1.42c ± 0.05	0.55d ± 0.00
Julie	85.24a ± 0.74	1.12	284.15b ± 0.07	0.548	2.46a ± 0.08	0.904	123.00a ± 0.01	15.00de ± 0.00	1.39c ± 0.03	0.55d ± 0.00
Haden	70.97b ± 3.84	1.35	170.86h ± 0.03	0.912	1.73c ± 0.05	0.636	121.00a ± 0.02	14.00de ± 0.01	1.04d ± 0.03	0.54d ± 0.01
Alphonso	71.96b ± 2.13	1.33	203.11ef ± 0.05	0.767	1.80c ± 0.04	0.661	110.00c ± 0.01	13.00ef ± 0.00	0.86e ± 0.03	0.54d ± 0.00

Means within the same column with different alphabets are significantly (p ≤ 0.05) different.

Values are mean± Standard error.

Similarities among the cultivars based on their quantitative and chemical traits are presented as 47 paired similarity coefficients (Table 8). High genetic similarities ranging from 2.41 to 6.68 were observed in the pairwise combination. Least similarity (2.41) was obtained between Palmer and Kent, while the highest similarity (6.68) was between Ogbomoso and Edward. Other cultivars with high genetic similarity are Ogbomoso and Alphonso, Ogbomoso and Julie Ogbomoso and Kent, Ogbomoso and Lipen, Saigon and Edward, Edward and Alphonso, Haden and Alphonso.

Table 8. Similarity indices of different pairs of the 10 cultivars

Cultivars	Ogbomoso	Madoe	Palmer	Saigon	Lipen	Kent	Edward	Julie	Haden
Madoe	4.59
Palmer	4.96	4.21
Saigon	3.72	3.58	2.90
Lipen	5.35	3.09	3.21	4.11
Kent	5.58	4.47	2.41	3.30	2.83
Edward	6.68	4.16	3.88	5.29	2.94	4.27
Julie	6.41	3.95	4.60	4.59	3.38	3.13	4.19
Haden	4.65	3.57	2.82	3.38	3.93	3.54	4.40	4.42
Alphonso	5.39	4.70	4.14	4.73	2.39	3.39	5.02	4.49	5.29

The total variation explained by the first four principal component (PC) axes was 71% (Table 9). PC1 alone explains 24% of the total variation and was mainly due to the influence of fruit width which had a positive loading. The second PC which accounted for 19% of the total variation was predominantly a function of Fe and Phytate with Fe having a negative loading. The third component with 16% variance separated these cultivars based on Ca, Cu, P, and Mn all of which had a positive loading. PC 4 accounted for 12% of the total variation and this was attributed to amylose which had a positive loading.

Table 9. Principal component axis of the 10 mango cultivars

Variables	PC1	PC2	PC3	PC4	PC5	PC6	PC7	PC8
Fruit length	−0.13	−0.01	−0.13	−0.09	−0.12	0.44	−0.02	0.30
Fruit width	0.30	0.07	−0.05	−0.18	−0.10	−0.01	−0.01	0.26
Fruit thickness	0.27	0.05	−0.06	−0.22	−0.05	−0.09	−0.09	0.24
Fruit weight	0.28	0.13	−0.04	−0.19	0.00	0.02	−0.06	0.17
Peel	−0.27	0.14	−0.14	0.11	−0.10	0.08	−0.01	−0.12
Pulp	0.20	−0.15	0.09	0.17	0.03	0.03	−0.37	0.13
Fiber length	0.19	0.01	0.08	0.11	−0.29	−0.03	0.33	−0.09
Stone	−0.28	−0.10	0.09	−0.15	0.11	−0.08	0.10	0.16
TSS	−0.19	−0.03	0.05	0.07	−0.39	−0.16	0.08	0.25
Vitamin C	−0.15	0.08	0.27	0.22	0.01	−0.21	−0.10	0.07
Ph	−0.13	0.05	−0.16	0.19	0.30	−0.26	−0.03	0.22
Shelf life	0.07	0.25	−0.18	0.21	−0.14	−0.07	0.07	−0.22
Moisture	−0.21	0.05	−0.19	0.12	0.11	−0.18	−0.11	0.25
Ash	−0.06	0.24	−0.01	−0.27	0.05	0.16	0.27	0.13
Starch	0.06	−0.37	−0.02	0.05	0.05	0.06	0.06	−0.08
Reducing sugar	0.23	−0.23	−0.09	0.14	0.12	−0.06	0.05	−0.09
Amylose	0.05	−0.11	−0.21	0.30	−0.24	0.02	−0.20	−0.04
Fat	0.12	0.15	−0.20	−0.27	0.02	−0.22	0.14	−0.14
Crude fiber	0.03	−0.08	0.15	−0.03	0.16	−0.40	0.35	0.08
K	−0.05	−0.07	−0.05	0.14	0.40	0.17	0.25	0.19
Mg	−0.10	−0.17	−0.25	0.20	−0.21	0.01	0.05	0.17
Ca	−0.04	−0.11	0.31	−0.18	−0.18	0.09	−0.06	−0.12
Fe	−0.01	−0.32	0.04	−0.19	0.15	−0.04	−0.17	−0.10
Zn	−0.15	−0.23	0.17	−0.18	−0.08	0.20	0.08	−0.06
Cu	−0.08	−0.07	0.36	0.18	0.08	0.07	0.00	0.16
P	−0.08	0.14	0.34	0.12	−0.06	0.04	−0.06	−0.23
Cr	−0.16	−0.29	−0.02	−0.16	0.04	−0.01	0.18	−0.02
Mn	0.13	0.08	0.30	−0.02	−0.09	0.01	−0.22	0.37
DPPH	0.25	0.09	0.21	0.12	0.08	−0.11	0.11	0.07
Reducing power	0.27	−0.08	0.14	0.18	0.02	0.03	0.18	0.05
Total Phenolic	0.03	0.06	0.01	0.23	−0.14	0.29	0.40	0.21
Total antioxidant	0.10	0.16	0.14	0.20	0.28	0.24	0.08	−0.14
Flavonoid	0.08	0.17	−0.10	0.00	0.30	0.34	−0.15	−0.07
Tannin	−0.23	0.26	0.09	−0.04	0.08	0.01	−0.11	0.04
Phytate	−0.13	0.32	0.12	−0.08	−0.06	−0.10	0.00	−0.06
Eigen value	8.44	6.61	5.59	4.19	3.30	2.72	2.39	1.07
Proportion %	24	19	16	12	9	8	7	3
Cumulative %	24	43	59	71	80	88	95	98

Discussion

Most of the traits of interest in plant breeding are quantitative in nature. Quantitative classifications offer a divergence among individuals which enable breeders to understand the racial affinities and evolutionary pattern in various species of cultivated plants (Rajan et al., 2009). They also help in making a decision in the selection of best parental combinations in the hybridization program as it serves as a sound basis for grouping any two or more genotypes based on minimum divergence or resemblance between them (Rajan et al., 2009). The results obtained in this study for weight of peel (15.17–23.45%), stone (7.23–22.62%), and flesh (51.60–76.26%) are comparable to values obtained in some mango cultivars in which peel weight ranged from 13% to 22%, seed weight ranged from 12% to 20%, and fruit pulp ranged from 58% to 75% (Abdualrahm, 2013; Anila and Radha, 2003; Pleguezuelo et al., 2012).

Aguoru et al. (2016); Krishnapillai and Wilson (2016); Mitra (2016) explained that fruits are highly capable of influencing morphological variability among mangos. The variability, therefore, observed in this study for quantitative traits would provide useful information for breeders in selecting the best parental combinations in the mango hybridization program.

TSS, vitamin C, pH, and shelf life of mango fruits are important parameters which should be taken into consideration when planning a mango breeding program because they are indispensable contributors to the quality of mango. The pH values (4.60–5.91) obtained in this study are in agreement with the requirement of jam quality control (Brandão et al., 2018) and are similar to pH (4.35–4.71) obtained for mango cultivars in Bangladesh (Ara et al., 2014). TSSs obtained in this study are also similar to 12.87–21.05% (Ara et al., 2014), 17.31–20.75% (Ubwa et al., 2014), and 14.5–30.0% reported by Othman and Mbago (2009). TSS in fruit is an index used to determine the maturity of fruits and it is a strong indication of the harvesting time. TSSs contents are directly related to the acidity of fruits. Generally, the acidity of fruit decreases and TSSs increase during maturity and ripening stage of fruit (Sajib et al., 2014). Differences observed in the TSS content of fruits in this study could be as a result of the differences in cultivars. Rahman et al. (2010) noted that the TSS content of fruit is greatly affected by differences in cultivars.

Mango fruits, being climacteric, have a short life and quickly ripens between 3 and 9 days after harvest (Kour et al., 2018). This short period seriously restricts long-distance marketing. Also, sensitivity to disease, high temperature, and perishability due to faster ripening or softening of the fruits limit its potential in terms of storage, packaging, and transport (Kour et al., 2018). Proper postharvest treatments and packaging are required for maintaining better quality, extended shelf life, and having access to international markets (Anwar and Malik, 2007). To this effect, Hoque et al. (2018), in their experiment, recorded an average shelf life of 8.25 days for untreated mango when compared with mango treated with botanicals like neem and garlic extracts which had extended shelf life of 15 days. The shelf life of 9 to 12 days recorded for most mango cultivars in this study is slightly lower than those gotten by Hoque et al. (2018) for mango treated with preservatives but higher than untreated mango cultivars. Information gotten from this result could give a good insight into breeding for mango varieties with extended shelf life.

Vitamin C contents (7.51–11.68%) obtained in this study are similar to 6.04–11.23% obtained in three mango varieties (Ubwa et al., 2014) and comparable with 5.1–25.2% obtained in some mango cultivars (Othman and Mbago, 2009). Vitamin C, a water-soluble vitamin, is usually not stored in large amount in the body. Therefore, including vitamin C-rich food in our diet is important for growth, repair of bones, teeth, skin, tissues, and also for the prevention of cell damage and protection from infections by keeping the immune system healthy. The fruit pulps in this result meet the minimum vitamin C requirement of (15 mg/100 g and 80 mg/100 g) recommended by NAFDAC (2013) for fruit groups. Cultivars in this study having good levels of TSS, high vitamin C, long shelf life, and pH could be selected for future breeding purpose.

Significant differences were observed among all the mango cultivars for proximate composition, indicating some degree of variation. This finding indicates that cultivars displayed some degree of genetic diversity for proximate composition. Thus, selection based on any of these characteristics will lead to an improvement in Nigeria mango germplasm. Proximate analysis is widely accepted as a basis for nutritional evaluation (Alege and Mustapha, 2013). Moisture content determination is one of the most important analyses performed on food product because it determines the quality of that product. The higher the moisture content of a product, the more it is susceptible to spoilage by microbial action (Seema, 2015). The moisture content in this study ranged from 77.87% to 80.44%; this is similar to results (77.4–78.70%) obtained for some mango cultivars (Abdualrahm, 2013), 77.85–82.22% obtained for three mango varieties (Ubwa et al., 2014), and 74.58–86.36% reported by Uddin et al. (2006) Most fruits are composed of 70% to 90% of water (Ueda et al., 2000); our results, therefore, fall within the range reported for most fruits. In this study, Edward that had the highest shelf life also had the lowest moisture content with other varieties having moderate moisture content. Therefore, varieties developed from these cultivars are likely to have a longer shelf life and stability. Ash content of food refers to the inorganic residues remaining after ignition or complete oxidation of organic matter which reflects the mineral content in the food (Harbers, 2002). The ash contents (1.47–2.67%) obtained in this study are higher than 1.02–1.05% obtained in some mango cultivars (Ahmed and Ahmed, 2012), 0.32–0.34% obtained for mango (Abdualrahm, 2013), and 0.31–0.66% obtained for some mango cultivars (Ubwa et al., 2014).

In the unripe stage, the main carbohydrate in mango fruits is starch. During the ripening process, it is hydrolyzed to glucose and this monosaccharide participates in the biosynthesis of fructose and sucrose. In mango, when starch hydrolysis is carried out during ripening of the fruit, some oligosaccharides, maltose, and glucose are produced, and they are quantified as reducing sugars. Additional fructose is produced by the fruit and it is quantified inside this group. The starch contents of the mango cultivars in this study which ranged from 0.92% to 1.73% are similar to results presented in the table of food composition from ICBF (2015) in which the content of glucose, fructose, sucrose, starch, and pectin in mango pulp were 3.9, 1.0, 8.8, 1.8, and 8.2 g/100 g of pulp, respectively. The starch content obtained in this study is low when compared with 11.41% obtained for unripe mango (Bello-Pérez et al., 2007) but comparable to values (3.37% and 0.36%) obtained during storage of mango at the 4th and 6th day after harvest, respectively (Bello-Pérez et al., 2007). Thus, the low starch values may be due to the physiological state of the mango fruit, as a decrease in starch content has been observed during fruit ripening since it involves a series of physiological and biochemical changes (Bello-Pérez et al., 2007; Tharanathan et al., 2006). Starch hydrolysis occurs by the great increase in the activity of enzymes α -amylase, β -amylase, and starch phosphorylase which transforms starch into reducing sugars (Hussain et al., 2012; Lester and Grusak, 2000). The reducing sugar contents (0.09–0.61%) obtained in this study are lower than 1.5% obtained for unripe mango var. Tommy Atkins (De Oliveira et al., 2001); 4.5–4.8% obtained for some mango varieties (Abdualrahm, 2013) and 3.59% obtained for Dodo mango of Tanzania (Othman and Mbago, 2009). Thus, during ripening, glucose, fructose, and sucrose generally increase (Bernardes et al. 2008). The increase in these monosaccharides and disaccharides during maturation has been observed in Alphonso (Yashoda et al., 2006) and Dashehari (Ernesto et al., 2018). The crude fat content (0.19–0.90%) is in agreement with 0.20% obtained for Dodo mango of Tanzania (Othman and Mbago, 2009); 0.29–0.0.38% reported by (Abdualrahm, 2013) for some mango cultivars and 0.13–1.20% obtained for mango (Ara et al., 2014). It was reported that the crude fat content of different fruits is usually not greater than 1.0% (Marles, 2017). Therefore, the low levels of crude fat in these varieties of mango indicate that they are not good sources of fat. Crude fiber contents (0.82–1.24%) obtained in this study are similar to 0.84–1.11% obtained for mango cultivars (Ubwa et al., 2014); 0.85–0.87% obtained for some mango varieties of Tanzania (Othman and Mbago, 2009), but lower than 3.7% reported by Mamiro et al. (2007) and 4.2–4.5% obtained for some mango cultivars (Abdualrahm, 2013). Mango consumption is associated with better diet quality and higher nutrient intake (O’Neil et al., 2013), breeding mango with good proximate values could, therefore, contribute to improving nutritional status.

Minerals are needed for the proper function of the body, growth, and development and for maintaining good health. Potassium, magnesium, and calcium are macro-elements which act as electrolytes for ionic and osmotic balance, to strengthen cells and the endoskeleton (Shrimanker and Bhattarai, 2020). Adequate calcium intake reduces the risk of fractures, osteoporosis, and diabetes (Beto, 2015). Calcium in these cultivars is higher than what is obtained in most fruits (White and Broadley, 2005); therefore, breeding mangos with high calcium content could go a long way in ameliorating calcium deficiency-related ailments. Phosphorus was the most abundant of the mineral elements evaluated in this study followed by calcium; this is not in agreement with the report of Othman and Mbago (2009) in which potassium was the predominant element. A similar result was obtained in mango by USDA (2018) in which potassium was the most abundant mineral element followed by magnesium. Ara et al. (2014) reported that the most abundant element observed in some mango cultivars was sodium. These variations could be a result of environmental influences.

In this study, Alphonso outperformed the other cultivars for most of the mineral contents as it had the highest amounts of K, Mg, Ca, Fe, Zn, Cu, and Cr.

Iron has been considered as the most common deficient element among micronutrients (Bjørklund et al., 2017). Iron deficiency is associated with decreased levels of Cr and Zn in the body, suggesting a synergy between Cr, Zn, and Fe (Angelova et al., 2014; Bjørklund et al., 2017). The iron content (0.74–4.99 mg/100 g) obtained in this study is comparable to 0.6–1.4 mg/100 g reported by Bello et al. (2016); 1.27–1.53 mg/100 g obtained in some mango varieties (Othman and Mbago, 2009) but higher than 0.09–0.41 mg/100 g reported by USDA (2018). Zinc content ranged from 10.75 to 32.95 mg/100 g in this study and is much higher than 0.11–0.31 mg/100 g obtained by Othman and Mbago (2009); 0.06–0.15 mg/100 g reported by (USDA, 2018) and 0.0–0.01 mg/100 g reported by (ICBF, 2015). Among the mineral elements evaluated in this study, chromium which is a heavy metal was the least predominant (0.12–0.47 mg/100 g) in all the mango cultivars. Plants are capable of absorbing heavy metals from soil and some plants naturally absorb these metals; furthermore, accumulation in plants depends on plant species, growth stages, type of soil and metals (Ona et al., 2006). Chromium is a subject of growing interest, mammals need trivalent chromium to maintain balanced glucose metabolism (Golubnitschaja and Yeghiazaryan, 2012), and chromium at low concentration may facilitate insulin action (Lukaski, 2018). Therefore, the discovery of the genetic value of these cultivars with low chromium content could enhance the specific breeding program.

The potassium and magnesium content obtained in this study is comparable to those reported by Ara et al. (2014) and Bello et al. (2016), but lower than values reported by Othman and Mbago (2009). Mamiro et al. (2007) reported magnesium content of 17.09 mg/100 g for Dodo mango from Tanzania. Also, the amount of copper obtained ranged from 4.91 to 11.10 mg/100 g and is higher than 0.16–0.21 mg/100 g found in some mango cultivars (Othman and Mbago, 2009). Therefore, identification of cultivars with a good level of micro- and macro-elements to meet the RDA is important as potential sources of useful genes for breeding programs.

The importance of polyphenol-rich diet has long been underlined because of their radical scavenging action, as well as anti-carcinogenetic properties (Marianna et al., 2017). Fruits are rich sources of bioactive phytochemicals which exhibit good antioxidant properties and are, therefore, regarded as an essential component in the daily diet (Liu, 2003). There is increased interest in nutrients capable of counteracting oxidative stress; hence, breeders need to concentrate effort on breeding crops of high antioxidant activities. In this study, Julie and Saigon were observed to have the highest antioxidant activities and phytochemical compositions.

The total phenol content (96–123 mg/100 g) obtained in this study is comparable to 54.70–111.70 mg/100 g obtained for some mango cultivars (Jolayemi, 2019); 48.40–208.70 mg/100 g reported for three mango varieties (Ribeiro et al., 2007); 102.7 mg/100 g obtained in Ao mango (Dars et al., 2018); and lower than 145.52 mg/100 g obtained in Xiangya mango juice (Dars et al., 2018). Among the groups of phenols, flavonoids are the most potent antioxidants and anthocyanin is the group of flavonoid most common in mango pulp (Abbasi et al., 2015).

DPPH method is a quick method to analyze the free radical activity of natural compounds (Shahzor et al., 2015). The antioxidant activity of a substance can be expressed as its ability to scavenge the DPPH free radical. The DPPH scavenging activity of fresh fruit and vegetable juices is related to phenolic contents. With the increasing 1C₅₀ value, the decreasing trend of antioxidant activity was found in fruit and vegetable juices, and vice versa (Dars et al., 2018). In this study, the highest IC₅₀ value (1.35 mg/ml) and the lowest antioxidant activity as % inhibition (70.97) were found in Haden, followed by Ogbomoso with IC₅₀ value of 1.34 mg/ml and antioxidant activity as % inhibition of 71.46. A similar trend was reported by Dars et al. (2018) in which the highest IC₅₀ value (0.032 mg/ml) with the lowest antioxidant activity was recorded in Ao mango juice when compared with Xiangya mango.

Reducing power is an important indicator to determine the potential antioxidant activity. The reducing power of test samples and their actions are usually monitored by the formation of Perl’s Prussian blue at 700 nm (Shahzor et al., 2015). In this study, the reducing power with the highest EC₅₀ value (0.90 mg/ml) and lowest antioxidant activity (2.46) was recorded in Julie, followed by Edward with an EC₅₀ value of 0.76 mg/ml and antioxidant activity of 2.07. Dars et al. (2018) reported that among the mango cultivars evaluated, the reducing power with the highest EC₅₀ value (0.032 mg/ml) and lowest antioxidant activity was found in Ao mango as compared to Xiangya mango. Therefore, selecting suitable parents with good antioxidant and phytochemical content for further evaluation could enhance the breeding for cultivars capable of counteracting oxidative stress.

Tannin interferes with protein digestibility thereby adversely influencing the bioavailability of non-heme iron, leading to poor iron and calcium absorption (Egbekunle et al., 2017). However, its antinutritional effect depends largely on dose (Fekadu, 2013). Tannin content (0.54–0.63 mg/100 g) obtained in this study is lower than those reported by Umeobika et al. (2015) and values obtained in crops like Pomegranate (Maniyan et al., 2015), Apple, and Grapes (Duda-Chodak and Tarko, 2007). The tannin content is also lower than the maximum acceptable tannic acid daily intake (560 mg/100 g) for humans (Fekadu, 2013).

Phytate is a known anti-nutrient that forms insoluble complexes with minerals such as zinc, calcium, magnesium, and iron in the body resulting in mineral deficiencies (Akin-Idowu et al., 2017; Bello et al., 2008). However, there are evidences that dietary phytate below 6 g/100 g may play beneficial roles as antioxidants and anti-carcinogen (Elinge et al., 2012). The phytate content (0.86–1.87 mg/100 g) obtained in this study is much lower than 500 mg/100 g reported by Umeobika et al. (2015).

The distance relationship observed among the 10 mango cultivars reveals the presence of divergent genetic resources among genotypes. This will be useful in the selection of genetic materials for advancement and improvement. In general, genotypes with high diversity such as Ogbomoso, Alphonso, Saigon, Julie, and Haden also had high mineral contents. The selection of genotypes having wider genetic background and using them as parents in the hybridization program helps to generate new variants with heterotic significance in addition to increasing germplasm quantity (Adewale et al., 2013; Rahim et al., 2010).

PCA was performed to understand the underlying interrelationships and to determine the best linear combination of measured traits contributing to variation. The results showed that 71% of the observed variability was explained by four components. This indicates that variation in mango fruit was multi-directional, demonstrating that mango fruit is influenced by multiple traits. The selection of these PC axes for the description of mango cultivars was justified by their significant Eigenvalues. Eigenvalues greater than 1 are considered significant and component loading greater than ±0.3 are considered meaningful (Felix et al., 2015; Yanamandram and White, 2006). Fruit length, crude fiber, and Mn were important in at least two PCs indicating their relative importance to the variation among the cultivars. Similarly, Felix et al. (2015) found genetic variation among macro- and micronutrients in pearl millet. The positive and negative loading observed in this study agrees with Felix et al. (2015) and Abe et al. (2013).

Conclusion

The present study indicates that all cultivars of mango are rich sources of vitamin C, fiber, total soluble solids, reducing sugar, and important minerals and are safe from heavy metal contamination. Cultivars also possess a wide variation for various traits. Among the 10 cultivars, Alphonso had an outstanding performance for mineral and proximate composition with low fat and antinutrient composition. Julie, Saigon, and Palmer were identified with the highest antioxidants and phytochemical composition coupled with high shelf life.

Therefore, trait-based selection and genetic advancement may be achieved through a concerted crossbreeding program.