Physicochemical characterization and quantification of bioactive compounds of Antrocaryon amazonicum fruits cultivated in Brazilian Amazonia

ABSTRACT

In this study, we reported for the first time about the physicochemical characterization, as well as the bioactive compounds in the peel and pulp of A. amazonicum fruits of two varieties harvested in Northern Brazil (Pará State) from two harvest periods (beginning and end). The peel and pulp of A. amazonicum fruits presented high contents of water (75–89%) and total carbohydrates (5–12%), whereas low total energetic value ranges (41–92 kcal/100 g, wet basis). Regardless the variety of A. amazonicum fruits, phenolic compounds were the major bioactive compounds, and the peels accumulated high contents at the beginning of the harvest period (4884–6556 mg GAE/100 g, dry basis, db), and at the end for the pulps (1394–2855 mg GAE/100 g, db). The multivariate exploratory techniques evidenced that high contents of bioactive compounds are available in the peel at the beginning of the harvest period, while at the end for the pulps.

RESUMEN

En el presente estudio informamos por primera vez sobre la caracterización fisicoquímica y los compuestos bioactivos presentes en la cáscara y la pulpa de los frutos de A. amazonicum de dos variedades cosechadas en el norte del Brasil (estado de Pará) durante dos periodos de cosecha (principio y fin). La cáscara y la pulpa de los frutos de A. amazonicum muestran altos contenidos de agua (75-89%) y de carbohidratos totales (5-12%), mientras que los rangos de valor energético total son bajos (41-92 kcal/100 g, base húmeda). Independientemente de la variedad de frutos de A. amazonicum de que se trate, siempre los compuestos fenólicos son los principales compuestos bioactivos; las cáscaras acumulan altos contenidos de los mismos al principio del periodo de cosecha (4884-6556 mg GAE/100 g, base seca (db)), en tanto que las pulpas lo hacen al final (1394-2855 mg GAE/100 g, db). El uso de técnicas de exploración multivariantes permitió poner de manifiesto que se dispone de un alto contenido de compuestos bioactivos en la cáscara al principio del periodo de cosecha, mientras que al final tal contenido está disponible para las pulpas.

KEYWORDS:

• Amazonian fruit
• phenolic compounds
• carotenoids
• Principal Component Analysis
• Hierarchical Cluster Analysis

PALABRAS CLAVE:

• Fruta amazónica compuestos fenólicos
• carotenoides análisis de componentes principales
• análisis de agrupamiento jerárquico

1. Introduction

The Brazilian Amazonian region is considered a valuable natural reserve of food and medicinal plants worldwide, and about 44% of native Brazilian fruits are located in this region. Brazil has several underexploited native and exotic fruit species with potential interest to agro-industry and most of them are wild fruits or available only for local trade fairs (Anunciação et al., 2019; Neves et al., 2015).

Among the native fruits from the Amazonia, Antrocaryon amazonicum (Brazilian name: jacaiacá) belongs to Anacardiaceae family and it occurs in lower Amazon basin (Pará, Brazil), but also in Mato Grosso (Brazil), which is close to other Brazilian biomes (Cavalcante, 2010). It is also known in Brazil by names as cedar, taperebá-cedro, fruta-de-cedro, cedrorana, taperebá-açu, fruta-de-cedro and yacá-yacá, with fruits characterized as globular-flat drupes with 4–5 cm in diameter, 2–3 cm in length and average weight of 32 g (Cavalcante, 2010; FAO/WHO, 1986; Lorenzi, 2009). The fruits of A. amazonicum present a smooth and yellow-orange thin peel with a sweet pulp surrounding the hard endocarp, and they are widely appreciated in the regions of occurrence for their aromatic sensory characteristics (Teixeira et al., 2019), and they are used to prepare juices, jams, ice creams and alcoholic beverages.

The A. amazonicum fruits are promising sources of nutrients and bioactive compounds, such as phenolic compounds, carotenoids, and ascorbic acid; however, to the best of our knowledge, the contents of the major nutrients and bioactive compounds were not yet reported in the literature. The frequent intake of bioactive compounds has been associated with the prevention of several chronicle degenerative diseases, such as cancer, cardiovascular, coronary heart, atherosclerosis, and neurological disease (Lavecchia et al., 2013), and these compounds have been reported to exhibit antioxidant, anti-inflammatory, antiallergic, antimicrobial, anticarcinogenic, antiviral and antimutagenic activities (Alahyane et al., 2019).

Several studies have recognized the importance of comparing bioactive compounds variations in plants at different harvesting times (Olsson et al., 2012; Yao et al., 2016; Özcan et al, 2017), which is an useful strategy to determine the appropriate harvest period for extracting purposes, make indications for human intake and development of crop resources. Considering the lack of researches in the scientific literature about A. amazonicum fruits, the purpose of this study was to report for the first time the chemical composition of fruits of two varieties from two different locations in Northern Brazil during two different harvest periods. Furthermore, all data regarding the chemical composition of A. amazonicum fruits, including nutritional value and bioactive compound contents, were used to group and classify the fruits according to the fruit variety, harvest periods and part of the fruits by multivariate exploratory techniques.

2. Material and methods

2.1. Antrocaryon amazonicum fruits

The A. amazonicum fruits (two batches of 15 kg each) were collected at local fairs of Cametá (Latitude 2°14ʹ38ʹ’ S and Longitude 49°29ʹ45ʹ’ W) and São Caetano de Odivelas (Latitude 0°45ʹ0’’ S and Longitude 48°1ʹ12ʹ’ W), Pará State, Brazil. The access to these fruits was registered in the Brazilian National System for the Management of Genetic Heritage and Associated Traditional Knowledge (SisGen, A334963). The fruits from Cametá (CMT) and São Caetano de Odivelas (SCO) are from different varieties and they were obtained at two different harvest periods: beginning (March 2017) and end (April 2017). All the ripe fruits were sanitized by immersion in a sodium hypochlorite solution (100 μg/mL) for 10 min, and the peel and pulp were separated from the seed and each part of the fruit was ground and homogenized before vacuum-packing and stored at −18°C, protected from light, until analysis. Another part of the pulp and peel was freeze-dried for the bioactive compounds determination.

2.2. Biometrical characterization

The biometric characterization of A. amazonicum fruits was performed on randomly selected fresh fruits (n = 100/location). The selected biometric parameters were longitudinal and transverse diameters (cm), measured with a slide caliper (Vonder, Curitiba, PR, Brazil) and the weighed masses (g) of pulp, peel and seed of all the fruits were obtained with an analytical balance (model M214Ai, Bel Photonics, Osasco, SP, Brazil).

2.3. Physicochemical and proximate composition analyses

The physicochemical and proximate composition analyses were determined in the peel and pulp of A. amazonicum fruits. The contents of moisture (105°C), total ashes, total protein (total nitrogen conversion factor 6.25 to total protein), as well as pH, total acidity (equivalent of citric acid) and total soluble solids contents (°Brix) were obtained according to Association of Official Analytical Chemists (AOAC, 1998). Total lipid contents were determined according to Bligh and Dyer (1959), and the total carbohydrate contents were calculated by the difference of 100% and the sum of the percentages of moisture, ashes, lipids and proteins. The total energetic values were calculated according to the specific Atwater conversion factors for fruits: Total energetic value (kcal/100 g) = (% protein x 3.36 kcal/g) + (% lipids x 8.37 kcal/g) + (% carbohydrates x 3.6 kcal/g) (FAO/WHO, 2002). All the assays were carried out in triplicate and the results were expressed in g/100 g (%) of peel or pulp (wet basis, w.b.).

2.4. Determination of bioactive compounds

2.4.1. Ascorbic acid

Ascorbic acid was extracted from the freeze-dried peel and pulp of A. amazonicum fruits (1 g) with 1% aqueous oxalic acid solution (3 times), followed by centrifugation (3000 x g at 4°C) during 5 min and the combined supernatants were filtered (22 μm) prior injection into the HPLC-DAD system (Agilent 1260 Infinity model, Santa Clara, CA, USA) (Almeida et al., 2012). The ascorbic acid was separated on a C18 column (4 μm, 250 × 4.6 mm) with an isocratic mobile phase of sulfuric acid solution (0.001 M, pH 2.5), flow rate at 0.7 mL/min during 10 min and column temperature at 25°C. Ascorbic acid was detected at 245 nm and identified based on the retention time, co-chromatography and UV-visible characteristics compared to authentic standard of ascorbic acid analyzed under the same conditions. The ascorbic acid contents were determined by external seven-point analytical curves of ascorbic acid standard (1.56–100 mg/mL) and the results were expressed as mg/100 g (n = 3, dry basis, d.b.).

2.4.2. Total phenolic compounds and total flavonoids

The extracts used for the total phenolic compounds and total flavonoids determination were obtained from 0.5 g of freeze-dried peel and pulp of A. amazonicum fruits. The extraction was carried out five times with methanol/water (80:20, v/v) (Chisté & Mercadante, 2012).

The quantification of total phenolic compounds was carried out by the colorimetric method with Folin-Ciocalteu reagent at 750 nm (Singleton & Rossi, 1965) using seven-point analytical curves of gallic acid standard (1–100 mg/mL). The results were expressed as mg of gallic acid equivalent (GAE)/100 g (n = 3, d.b.).

The total flavonoids were quantified by spectrophotometry at 435 nm, as described by Pękal and Pyrzynska (2014), using seven-point analytical curves of quercetin standard (1.56–100 mg/mL). The results were expressed as mg of quercetin equivalent (QE)/100 g (n = 3, d.b.).

2.4.3. Total carotenoids

The carotenoids were exhaustively extracted from 3 g of the freeze-dried peel and pulp of A. amazonicum fruits with acetone and subsequent liquid-liquid partition in petroleum ether/diethyl ether (1:1, v/v), followed by spectrophotometric readings at 450 nm (Godoy & Rodriguez-Amaya, 1994). The total carotenoid contents were calculated considering the specific molar absorptivity coefficient of β-carotene in petroleum ether ($A_{1cm}^{1\mathrm{\%}}$ = 2592) (Davies, 1976) and the results were expressed as mg/100 g (n = 3, d.b.).

2.5. Statistical analysis

All the physicochemical and bioactive compounds data found for the peels and pulps of A. amazonicum fruits (mean ± standard deviation) were submitted to analysis of variance (ANOVA) and the means were compared by Tukey test at 95% of significance (p <.05). Two multivariate exploratory techniques, Principal Components Analysis (PCA) and Hierarchical Cluster Analysis (HCA), were applied for grouping and classification of both part of the fruit (peel and pulp) according to the variety and harvest periods using Statistica software.

3. Results and discussion

3.1. Biometric, physicochemical characterization, and proximate composition of Antrocaryon amazonicum fruits

The A. amazonicum fruits from both varieties (CMT and SCO) were characterized as globular and yellowish with approximately 3–4 cm in longitudinal and transverse diameters, but they showed different apparent physical forms: star-shaped (CMT) and rounded (SCO) (Figure 1). The whole fruits weighed approximately 26.73 ±5.18 g to 31.65 ±5.86 g, their seeds accounted for 32–34% of the fruit, the peel accounted for 16–32% and the pulp 34–52%, while the highest pulp yield was obtained for the fruits from SCO (52%). These biometric characteristics for both varieties were similar to the values previously reported (FAO/WHO, 1986).

Figure 1. Antrocaryon amazonicum fruits from Cametá (a) and São Caetano das Odivelas (b) and the longitudinal and transverse diameters.

Figura 1. Frutos de Antrocaryon amazonicum de Cametá (a) y São Caetano das Odivelas (b) y diámetros longitudinal y transversal.

Concerning the main physicochemical characteristics and proximate composition, the results provided an overview about the chemical composition and nutritional value of the fruits from both varieties and different harvest periods (Table 1). The pH values varied from 4.57 to 5.37 in the peel and from 2.10 to 2.41 in the pulp, evidencing the acidic characteristic of fruit pulp, as expected. However, negligible variation were observed for the total acidity content in both part of the fruits from different location and harvest periods, varying from 1.10–1.28% of citric acid in the peel and 0.93–1.2% of citric acid in the pulp. According to the literature, the pH and total acidity values found for the A. amazonicum pulps were similar to those reported for Spondias mombin (pH 2.53 and total acidity of 1.86%), another fruit from the same Anacardiaceae family (RA Mattietto et al., 2010). The acidity of fruits can vary considerably depending on fruit species, and those with total acidity between 0.1 and 0.3% (expressed as citric acid) were considered as low acidic, such as melon, banana, and avocado, while fruits with values between 3 and 8% were considered acid, such as lemon (Tiburski et al., 2011). Thus, A. amazonicum was considered a medium acidic fruit, since its total acidity was within the range of 0.9–3%, such as cherry, strawberry, raspberry and orange fruits (Mattietto et al., 2007; Tiburski et al., 2011).

Table 1. Physicochemical characterization and proximate composition of peel and pulp of Antrocaryon amazonicum fruits from Northern Brazil.

Tabla 1. Caracterización fisicoquímica y composición aproximada de la cáscara y la pulpa de los frutos de Antrocaryon amazonicum del norte del Brasil.

	CMT		SCO
	Beginning of the harvest (March 2017)	End of the harvest (April 2017)	Beginning of the harvest (March 2017)	End of the harvest (April 2017)
Peel
pH	5.07 ± 0.04^b	4.57 ± 0.01 ^c	5.37 ± 0.08^a	5.33 ± 0.02^a
Total acidity (% citric acid)	1.28 ± 0.01^b	1.34±<0.01^a	1.12±<0.01 ^c	1.10 ± 0.02 ^c
Moisture (%)	78.5 ± 0.2^b	81.5 ± 0.8^a	75.5 ± 0.4 ^c	82 ± 1^a
Total lipids (%)	1.1 ± 0.1 ^c	1.7 ± 0.1^b	2.0 ± 0.1^a	1.41 ± 0.06 ^c
Total proteins (%)	2.33 ± 0.05 ^c	4.2 ± 0.1^b	5.6 ± 0.1^a	4.6 ± 0.1^b
Ashes (%)	0.83±<0.01^a	0.54 ± 0.01 ^c	0.7 ± 0.1^b	0.56 ± 0.01 ^c
Total carbohydrates (%)	17.6 ± 0.2^a	12.18 ± 0.01 ^c	16.4 ± 0.5^a	12.4 ± 0.8^b
Total energetic value (kcal/100 g)	78 ± 2^b	62 ± 4 ^c	92 ± 4^a	64 ± 5 ^c
Pulp
pH	2.41 ± 0.08^a	2.39 ± 0.01^a	2.17 ± 0.03^b	2.10 ± 0.02^b
Soluble solids (°Brix)	10.3 ± 0.2^a	9.2 ± 0.2^b	8.20±<0.01 ^c	9.60 ± 0.01^b
Total acidity (% citric acid)	1.13 ± 0.05^a	0.93 ± 0.07^a	1.2 ± 0.1^a	1.02±<0.01^a
Moisture (%)	82.9 ± 0.7^d	88.2 ± 0.2^a	85.2 ± 0,8 ^c	89.1 ± 0.1^a
Total lipids (%)	0.95 ± 0.08^a	0.54 ± 0.2 ^c	0.93 ± 0.05^a	0.96 ± 0.03^a
Total proteins (%)	2.2 ± 0.1^d	3.3 ± 0.2 ^c	4.1 ± 0.2^a	4.2 ± 0.3^a
Ashes (%)	0.54 ± 0.01^b	0.54 ± 0.01^b	0.7 ± 0.1^a	0.27 ± 0.01 ^c
Total carbohydrates (%)	13 ± 1^a	7.0 ± 0.5^b	8.41 ± 0.01^b	5.4 ± 0.1 ^c,d
Total energetic value (kcal/100 g)	64 ± 2^a	44 ± ^b	56 ± 3^a	41 ± 2^b

CMT: Cametá (Pará, Brazil); SCO: São Caetano de Odivelas (Pará, Brazil). All the values (mean ± standard deviation, n = 3, wet basis) with the same superscript letters are not statistically different (p < 0.05) (Tukey’s test).

CMT: Cametá (Pará, Brasil); SCO: São Caetano de Odivelas (Pará, Brasil). Todos los valores (media ± desviación estándar, n = 3, base húmeda) con las mismas letras de superíndice no son estadísticamente diferentes (p < 0.05) (prueba de Tukey).

Regarding the proximate composition (Table 1), water and carbohydrates were the main constituents of both parts of A. amazonicum fruits. For both fruit parts and varieties, the contents of total carbohydrates were higher at the beginning of the harvest period than at the end, while the water content exhibited opposite behavior. The other constituents did not show expressive differences along the harvest period or between the varieties. The contents of water, total lipids, total proteins, and total carbohydrates of pulps of A. amazonicum fruits were higher than those values reported for Spondias mombin (Tiburski et al., 2011), while the peels presented similar composition to the contents found in peels of mango (Marques et al., 2010) and mangosteen at three different harvest periods (Chisté et al., 2009). Concerning the total energetic value, A. amazonicum fruits showed values in the same range that other tropical fruits, such as mango (60 kcal/100 g) and guava (68 kcal/100 g) (USDA, 2016).

3.2. Bioactive compounds of Antrocaryon amazonicum fruits

The peels of mostly fruits were reported to exhibit higher contents of secondary metabolites than their pulps (Ayala-Zavala et al., 2011). In our study, the highest contents of bioactive compounds were found in the peel of A. amazonicum fruits from both varieties (Table 2), excepting for ascorbic acid that presented the highest contents in the pulp. The ascorbic acid contents in pulps were 2–5 times higher than the values found in the peels and, in all the cases, the highest values were observed at the beginning of the harvest period in both varieties (Table 2). At the end of the harvest period, the ascorbic acid contents decreased by about 62–71% in the peel and 86–88% in the pulp as compared to the beginning. Thus, if ascorbic acid is the target bioactive compound to be obtained from these fruits, the most appropriate time for harvesting is at the beginning of the period, regardless of their variety. The ascorbic acid contents found in the peel and pulps of A. amazonicum fruits were lower than the values reported for peel and pulp of Spondias purpurea (115–130 mg/100 g) and Spondias tuberosa (152–174 mg/100 g), both fruits from the same family (Omena et al., 2012). However, the ascorbic acid in the pulp at the beginning of the harvest period is in the same range than those observed for guava (65.8 mg/100 g), kiwifruits (55–91 mg/100 g) and higher than pineapples (11–61 mg/100 g) (Valente et al., 2011). In plants, the accumulation of ascorbic acid in plant tissues changes during physiological phenomena such as senescence, cell expansion development and, various biotic and abiotic stimulations since its metabolism is associated with oxidative stress defense (Corrêa et al., 2018).

Table 2. Bioactive compounds of peel and pulp of Antrocaryon amazonicum fruits from Northern Brazil.

Tabla 2. Compuestos bioactivos de la cáscara y la pulpa de los frutos de Antrocaryon amazonicum del norte de Brasil.

	CMT		SCO
Bioactive compounds*	Beginning of the harvest (March 2017)	End of the harvest (April 2017)	Beginning of the harvest (March 2017)	End of the harvest (April 2017)
Peel
Ascorbic acid (mg/100 g)	15.6 ± 0.4^a	5.8 ± 0.3^b	26.7 ± 0.8^a	7.7 ± 0.1^a
Total phenolic compounds (mg GAE/100 g)	4884 ± 13^b	2886 ± 5^d	6556 ± 25^a	3736 ± 1 ^c
Total flavonoids (mg QE/100 g)	68.9 ± 0.4 ^c	69.8 ± 0.7 ^c	109 ± 2^b	127 ± 2^a
Total carotenoids (mg/100 g)	1.72 ± 0.07^b	2.23 ± 0.04^b	4.6 ± 0.2^a	4.5 ± 0.3^a
Pulp
Ascorbic acid (mg/100 g)	81 ± 1^a	9.7 ± 0.6^a	77 ± 1^a	10.4 ± 0.3^a
Total phenolic compounds (mg GAE/100 g)	1267 ± 10^d	1910 ± 12^b	1394 ± 1 ^c	2855 ± 27^a
Total flavonoids (mg QE/100 g)	28.6 ± 0.1^d	64.9 ± 0.4 ^c	67.7 ± 0.8^b	95.2 ± 0.1^a
Total carotenoids (mg/100 g)	0.5 ± 0.1 ^c	1.2 ± 0.1^b	1.28 ± 0,05^b	1.73 ± 0.01^a

CMT: Cametá (Pará, Brazil); SCO: São Caetano de Odivelas (Pará, Brazil). *The contents (mean ± standard deviation, n = 3, dry basis) with the same superscript letters in the same line are not statistically different (p < 0.05) (Tukey’s test). GAE: gallic acid equivalent; QE: quercetin equivalent.

CMT: Cametá (Pará, Brasil); SCO: São Caetano de Odivelas (Pará, Brasil). *El contenido (media ± desviación estándar, n = 3, base seca) con las mismas letras de superíndice en la misma línea no son estadísticamente diferentes (p < 0.05) (prueba de Tukey). GAE: equivalente en ácido gálico; QE: equivalente en quercetina.

Among the investigated bioactive compounds in this study, phenolic compounds were the most expressive. The levels of total phenolic compounds accounted for 2.8 to 6.5% of the peel composition and 1.2 to 2.8% of the pulp composition (Table 2). In the peel of A. amazonicum fruits from both varieties, the total phenolic compound contents were ≈1.7 times high at the beginning of the harvest period, and the fruits from SCO presented the highest values. The total phenolic compound contents decreased ≈ 40% in the peel of fruits from both locations from the beginning to the end of the harvest period. Conversely, the highest phenolic compound contents for the pulps were observed at the end of the harvest period for both locations (1.5–2 times), showing an increase of ≈33% in the pulp of fruits from CMT and ≈51% in the pulp of fruits from SCO from the beginning to the end of the harvest period. Similar behavior was observed for the total flavonoid contents, where the highest values were observed at the end of the harvest period in both fruit parts and locations, excepting for the fruits from CMT, which did not show statistical difference during the harvest period (Table 2).

Our results suggest that the highest yields of phenolic compounds of A. amazonicum fruits can be obtained at the beginning of the fruiting season for the peel and towards the end of the season for the pulp. The role of phenolic compounds, which are secondary metabolites, is widely discussed and among the reported functions, they were reported to exhibit antimicrobial, antiproliferative, insecticidal or herbicidal activity, as well as protection functions in plants, such as tolerance to extreme temperatures or water stresses, attraction of pollinators and defenses against herbivores and pathogens (Girardi et al., 2014; Rial et al., 2014). Antrocaryon amazonicum fruits showed to accumulate high levels of phenolic compounds in the peel at the beginning of the harvest period and this fact can be due to its high exposition to environmental variations, requiring the presence of these compounds for their maintenance, development, and defense (Miyashira et al., 2012; Motta et al., 2013; Righi et al., 2013). Vasco et al. (2008) classified the total phenolic compound contents of 17 fruits from Ecuador into three categories: low (<100 mg GAE/100 g), medium (100–500 mg GAE/100 g) and high (> 500 mg GAE/100 g), and according to this classification, both the peel and pulp of A. Amazonicum fruits can be considered high sources of total phenolic compounds.

Regarding total carotenoids, there are no statistical differences (p <.05) among the contents in the peel of A. Amazonicum fruits from the beginning to the end of the harvest period for both varieties. However, the peels of fruits from SCO showed the highest total carotenoid contents, about twice higher than fruits from CMT. For the pulps, the carotenoid contents were superior at the end of harvest period and the fruits from SCO presented the highest values. The carotenoid contents found in the pulp of A. amazonicum fruits were in the same range to those reported for plum (0.5 mg/100 g), star fruit (0.8 mg/100 g), loquat (1.4 mg/100 g), but lower than the other Amazonian fruits considered as very high carotenoid sources, such as buriti (14.2 mg/100 g) (Barreto et al., 2009), tucumã (8.4 mg/100 g) and peach palm fruits (3.2 mg/100 g) (Matos et al., 2019).

The contents of secondary metabolites in plants may be influenced by numerous factors, such as environmental conditions (temperature, rainfall, insolation), growth cycles, genotypes, harvesting times and cultivation techniques. As far as our knowledge is concerned, this is the first time the variation of phenolic compounds and carotenoids were demonstrated for A. amazonicum fruits as influenced by fruit varieties and harvest periods. According to the available 30-years climatological information of the cities where the A. amazonicum fruits were collected (Figure 2), the average values of insolation, precipitation and temperature in CMT were higher in April than March, while the insolation and temperature were also superior in April than March in SCO, but the average precipitation was higher in April.

Figure 2. Monthly 30-years average values of total insolation (h), precipitation (mm) and temperature (°C) from January to December in the Northern Brazil cities where the Antrocaryon amazonicum fruits were harvested: Cametá (CMT) and São Caetano das Odivelas (SCO)*. Fonte: Data available in Brazilian National Institute of Meterology (http://www.inmet.gov.br) and Climatempo (https://www.climatempo.com.br). *Once none climate station was available to SCO, the average insolation information was obtained with data of the nearest station (Belém, PA, ≈100 km of distance).

Figura 2. Valores medios mensuales de 30 años de insolación total (h), precipitación (mm) y temperatura (°C) de enero a diciembre en las ciudades del norte de Brasil donde se cosecharon los frutos de Antrocaryon amazonicum: Cametá (CMT) y São Caetano das Odivelas (SCO)*. Fuente: datos disponibles en el Instituto Nacional de Meteorología de Brasil (http://www.inmet.gov.br) y Climatempo (https://www.climatempo.com.br). *Cuando la SCO no disponía de ninguna estación climática, la información sobre la insolación media se obtuvo a partir de los datos de la estación más cercana (Belém, PA, ≈100 km de distancia).

In our study, regardless the varieties, the peels of A. amazonicum fruits, which are the fruit part mostly exposed to external environmental stress conditions, they presented lower total phenolic compound contents (≈ 1.5 times) in April than March (Table 2) as both the average temperature and insolation increased (Figure 2). The influence of a single factor may not produce high correlations with secondary metabolite contents, as demonstrated by Yao et al. (2016), which showed the positive correlation between the average precipitation and phenolic compounds contents was higher than the average temperature. Conversely, the total phenolic contents, including total flavonoids, increased in the pulps as the average temperature and insolation have increased in both cities. This behavior was also observed for grapes since a decrease by 15% in solar radiation has decreased up to 60% the total anthocyanins contents, suggesting the decrease of ligh exposure decreased synthesis of flavonoids due to the decrease in the activity of phenylalanine ammonia lyase, the principal involved enzyme whose activity is strongly dependent on light exposure (Mazza & Miniati, 1993). Concerning the contents of carotenoid in the peel of A.amazonicum fruits, their contents were not affected by the harvest period (p <.05), but by fruit varieties. On the other hand, the highest carotenoid contents of pulps were always found in fruits harvested at the end of harvest time (April) (Table 2) as the average temperature and insolation have increased in both cities (Figure 2). Similar results were observed in four red cultivars of carrots where the harvest date influenced the carotenoid contents with the earliest harvest date exhibited the lowest values (almost 1.5 times lower with 1 month difference) (Olsson et al., 2012). Furthermore, solar UV-B radiation during plant growing seems to influence the accumulation of carotenoids and phenolic compounds (Becatti et al., 2009; Mannucci et al., 2020); however, since not all plant species behave in the same way, it is difficult to generalize the effects of UV-B radiation on plant physiology due to different experimental acclimation conditions (Jansen, 2002).

Accordig to Yao et al. (2016), these observations are consistent with the growth-differentiation balance hypothesis in plants, which includes growth- and differentiation-related processes. These authors explained that differentiation is closely connected to the production of secondary metabolites and when resources are abundant, plants are given priority to grow, while when resources are limited, plant growth and differentiation is reduced. Thus, when an intermediate level of resources is available, which may occur in response to a mild drought, a nutrient deficit or an environmental stress condition, plants prioritize differentiation with the synthesis and accumulation of numerous secondary metabolites.

3.3. Classification of Antrocaryon amazonicum fruits by multivariate statistical analysis

The two principal components, PC1 and PC2, were responsible for explaining 80% of the total data variance, being the first main component (PC1) able to explain 54.97%, while the second (PC2) explained 25.12% of the total variance (Figure 3a). The dendogram obtained by HCA (Figure 3b) provides evidence for the formation of two groups, which can be observed in Figure 3c.

Figure 3. Classification of pulp and peel of Antrocaryon amazonicum fruits considering their physicochemical characterization and bioactive compounds contents. (a) Variable projection by Principal Component Analysis (PCA), (b) Dendrogram by Hierarchical Cluster Analysis (HCA) and (c) Dispersion plot for the Antrocaryon amazonicum fruit parts as grouped by PCA, based on HCA similiarities. Abbreviations: CMT_Peel 1 (peel of fruits from Cametá at the beginning of the harvest), CMT_Peel 2 (peel of fruits from Cametá at the end of the harvest), CMT_Pulp 1 (pulp of fruits from Cametá at the beginning of the harvest), CMT_Pulp2 (pulp of fruits from Cametá at the end of the harvest), SC_Peel 1 (peel of fruits from São Caetano de Odivelas at the beginning of the harvest), SC_Peel 2 (peel of fruits from São Caetano de Odivelas at the end of the harvest), SC_Pulp1 (pulp of fruits from São Caetano de Odivelas at the beginning of the harvest), SC_Pulp 2 (pulp of fruits from São Caetano de Odivelas at the end of the harvest). CFT (total phenolic compounds), FT (total flavonoids), CT (total carotenoids), AA (ascorbic acid) and VET (total energetic value).

Figura 3. Clasificación de la pulpa y la cáscara de los frutos de Antrocaryon amazonicum teniendo en cuenta su caracterización fisicoquímica y el contenido de compuestos bioactivos. (A) Proyección variable por análisis de componentes principales (PCA); (B) Dendrograma obtenido por análisis de agrupamiento jerárquico (HCA); y (C) Diagrama de dispersión de las partes de los frutos de Antrocaryon amazonicum agrupadas por el PCA, basado en las similitudes del HCA. Abreviaturas: CMT_Peel 1 (cáscara de los frutos de Cametá al principio de la cosecha), CMT_Peel 2 (cáscara de los frutos de Cametá al final de la cosecha), CMT_Pulp 1 (pulpa de los frutos de Cametá al principio de la cosecha), CMT_Pulp 2 (pulpa de los frutos de Cametá al final de la cosecha), SC_Peel 1 (cáscara de los frutos de São Caetano de Odivelas al comienzo de la cosecha), SC_Peel 2 (cáscara de los frutos de São Caetano de Odivelas al final de la cosecha), SC_Pulp1 (pulpa de los frutos de São Caetano de Odivelas al comienzo de la cosecha), SC_Pulp 2 (pulpa de los frutos de São Caetano de Odivelas al final de la cosecha). CFT (compuestos fenólicos totales), FT (flavonoides totales), CT (carotenoides totales), AA (ácido ascórbico) y VET (valor energético total).

The physicochemical characterization and proximate composition of peel and pulp of A. amazonicum fruits from CMT and SCO were not able to differentiate or classify the fruits according to the fruit variety or the harvest period, whereas the bioactive compounds contents were responsible to group these fruits according to the fruit parts. According to PCA and HCA analyses, the contents of bioactive compounds grouped the A. amazonicum fruits into two distinct groups: one group composed by the peel and the other group was composed by the pulp. The first group (Figure 3c) was formed by the peel of A. amazonicum fruits from both varieties and periods according to the highest contents of total phenolic compounds, total flavonoids and total carotenoids (Figure 3a). Moreover, among the peels, the fruits from SCO at the beginning of the harvest period were highlighted by the highest levels of total phenolic compounds, while the peels of fruits from SCO and CMT at the end of the harvest period showed the highest levels of total flavonoids and total carotenoids. The second group (Figure 3c) was formed by the pulps of A. amazonicum fruits from both varieties and harvest periods due to the highest contents of ascorbic acid. Furthermore, the pulps collected at the beginning of the harvest period from both locations were located near each other since they presented the highest levels of ascorbic acid among the samples (Figure 3b and c).

Altough the contents of bioactive compounds in fruits can be influenced by growing location, crop, cultivar, ripening stage, cultivation and processing practices (Sancho et al., 2011), in our study, PCA and HCA evidenced the efficient grouping of peel and pulp of A. amazonicum fruits mainly considering the contents of bioactive compounds. Summarizing, the levels of bioactive compounds were higher in the peel of fruits from SCO at the beginning of the harvest period, while the differences concerning the bioactive compounds in the pulps were evidenced between the harvest period, whereas not between the fruit variety.

4. Conclusion

This is the first report concerning the physicochemical, proximate composition and bioactive compounds characterization of A. amazonicum fruits and the influence of two fruit varieties harvested at the beginning and end of fruiting season. The peel and pulp of A. amazonicum fruits presented high contents of water and total carbohydrates, whereas low total energetic value ranges. Both parts of A. amazonicum fruits have a promising chemical composition for the research of bioactive compounds regardless their variety. Therefore, the A. amazonicum fruits from São Caetano de Odivelas (SCO; Pará, Brazil) presented high contents of bioactive compounds as compared to fruits from Cametá (CMT, Pará, Brazil) and the harvest period should be considered for prospecting purposes: at the beginning to extract high contents from the peel and at the end to obtain high levels from the pulp.

Declaration of interest

The authors have no conflict of interest.

Acknowledgements

The authors acknowledge Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq, Brazil, Project 403121/2016-7) and Federal University of Pará (UFPA, Brazil) through PROPESP/UFPA for the financial support. This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES, Brazil - Finance Code 001, Process 04319973305) through Master scholarship of Anna Paula P. Barbosa.

Additional information

Funding

This work was supported by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior [04319973305]; Conselho Nacional de Desenvolvimento Cientifico e Tecnologico [403121/2016-7].