TY - JOUR
AU - Suárez, Juan Carlos
AB - 1. Introduction Cocoa (Theobroma Cacao L.) is the basic raw material to produce chocolate and other derivatives such as cocoa butter, cocoa powder and cocoa liquor (cocoa paste) [1]. This food has sustained part of the daily diet of people and is currently making inroads in the cosmetics, perfumery and pharmaceutical industries [2] with a requirement in the final quality of the product. Recent studies have reported that grain quality in sensory terms is affected by different factors such as the genetics of the material [3], environmental conditions [4], the state of maturity of the grain and practices in the postharvest stage [5], such as those related to fermentation and drying [6, 7]. The composition and quality of cocoa beans is determined by a set of physical, chemical and sensory qualities [8]. According to García et al. [9] the main biochemical changes that influence the quality of cocoa beans occur during the fermentation stage, since it is there where different microbial communities (yeasts, lactic acid and acetic acid bacteria) intervene in the metabolization of sugars and other compounds [10]. Fermentation is a process in which fresh cocoa beans are placed in a wooden box for four to seven days. During this process, the microorganisms metabolize the pulp that surrounds the beans and generate reactions that result in an increase in temperature, a decrease in pH and death of the embryo [11]. During fermentation, different biochemical reactions occur on various substances, including antioxidants, polyphenols and methylxanthines [1]. For example, polyphenols are products of the secondary metabolism of plants, characterized by aromatic rings with hydroxyl groups as substituents, which are directly related to the sensory properties of chocolate and its positive effect on human health [12]. Inadequate fermentation of cocoa produces purple and slaty beans, negatively affecting quality and thus marketing price [13]. However, it has been reported that beans without fermentation have genotype-specific sensory attributes, mainly related to bitterness, astringency and acidity, and other attributes depend directly on fermentation, drying and roasting [14]. Cocoa fermentation is mainly carried out by small producers or collection centers in an artisanal manner, with little or no technology and without monitoring of processing conditions, which results in low quality beans [15]. These variations during the fermentation process make the chocolate industry face increasing challenges in maintaining the supply of standard products with high quality organoleptic properties [16]. This condition in the final product causes the cocoa bean to be marketed as ordinary cocoa with prices already regulated by an industry monopolized by large companies, which in the case of Colombia, acquire 90% of the national production [17], which means that the payment received does not have a sufficient impact on the welfare of cocoa-growing families [18], which discourages and puts cocoa production at risk, increasing the possibility of a change in production activity [19]. Although fermentation is an indispensable stage to ensure optimal bean quality [20], there are deficiencies in the number, intensity and coverage of studies that determine the effects of the fermentation process on physicochemical properties related to the content of protein, fat, fiber, functional and antioxidant capacity of cocoa beans as well as the different sensory attributes. The study focused on these cocoa bean processing plants since the three associations of cocoa producers called ASOACASAN, COMICACAO and COMCAP are part of the strategic cooperation agreements with European companies whose objective is the production and marketing of organic chocolate under the principles of fair trade which promotes sustainability since these associations produce “organic cocoa free of deforestation”. Therefore, the objective of this study was to monitor the biochemical, physical and sensory changes during fermentation process of cocoa beans in cocoa bean processing plants in the department of Caquetá, Colombia. It is expected that the study will provide a scientific basis for decision-making based on the recognition of the local needs of each plant. With this information, the fermentation processes carried out by cocoa collection centers in the department of Caquetá will be improved and standardized, resulting in better bean quality and added value in marketing. 2. Materials and methods 2.1. Study area The study was carried out in the department of Caquetá, specifically in the cocoa bean processing plants of a) Asociación Orgánica Agrícola de San José del Fragua-ASOACASAN (ASA) (1°19′43″N 75°58′22″W), b) Comité de Cultivadores de Cacao en Sistemas Agroforestales del municipio de San Vicente del Caguán-COMICACAO (CMI) (2°06′55″N 74°46′12″W) and c) Comité de Cacaoteros de los municipios del Paujil y el Doncello-COMCAP (COC) (1°40′26″ N 75°16′48″W). The environmental conditions in which these processing plants are found mainly contrast in the amount of average annual precipitation and the relative humidity (RH), variables that can affect the drying time. For example, in ASOACAN, which is located further north in the department of Caquetá, it has an average annual rainfall of 4,200 mm with a RH of 85% and handles the mass (560 kg of cocoa beans with mucilage) in relation to 7 days of fermentation with 6 subsequent turnings of the anaerobic phase and uses 14 days in the drying process. COMICACAO, very north of the department of Caquetá, has an average annual rainfall of 2,700 mm with a RH of 83%, it handles the mass (270 kg of cocoa beans with mucilage) very similar to ASOACASAN but with the main difference is that they dry only in 10 days. Finally, COMCAP, which is located 90 km from COMICACAO to the north and 128 km to the south with ASOACASAN, presents an average annual rainfall of 3,800 mm with a RH of 82%, carries out mass management (340 kg of cocoa beans with mucilage) with 8 days of fermentation with 7 turning and 12 days of drying. All cocoa processing plants carry out the turning of the cocoa bean during fermentation every 24 hours and the process of drying the bean after fermentation in canopies at room temperature. 2.2. Monitoring fermentation and drying of cocoa beans The cocoa beans were obtained from healthy pods from trees of different universal, national and regional clones obtained from farms associated with the organizations. Both ASOACASAN and COMICACAO make mixtures of cocoa beans from clones such as CCN51, ICS1, ICS95, TSH565 and FEAR5, however the difference between these associations is the additional use in FLE2 and ICS39 for ASOACASAN and COMICACAO, respectively. On the other hand, COMCAP carries out fermentation processes for cocoa beans obtained from raw materials. This process was carried out between March and June 2023, when production peaked. The beans were then fermented and dried according to the protocol of each collection center. The cocoa beans were placed in wooden crates, which were covered with banana leaves and natural jute fiber sacks to maintain a homogeneous temperature. These crates were perforated at the bottom to allow drainage of the cocoa pulp. The fermentation process lasted 168 hours (7 days) and during this time the fermentable mass was turned every 24 hours after 48 hours of fermentation. The turning of the grain was done by passing the dough into the compartments of the fermenter box, a process that was carried out using a plastic shovel. This fermentation process was carried out in triplicate, that is, three fermentation bins were used for each storage center. During fermentation, the following parameters were monitored daily between 13:00 and 16:00 hours: mass temperature, pH of the grain pulp and pH of the cotyledon. Samples for this monitoring were taken in the center and at the ends of the bin at a depth of 30 cm. This process was carried out in triplicate. The temperature of the dough was determined using a digital thermometer Hl 145 (Hanna Instrument, Woonsocket, RI, USA), the pH of the pulp during fermentation was measured using a pH tester Hl 98108 (Hanna Instrument, Woonsocket, RI, USA) and the pH of the cotyledon was performed following the methodology used by Papalexandratou [21]. For which, 15 g of randomly selected cocoa beans were taken from the three points of the bin; then, the pulp and testa were manually removed from the beans with a knife and the cotyledon was macerated with 30 ml of deionized water in a mortar for one minute until a homogeneous sample was obtained and the pH was read using a pH tester Hl 98108 (Hanna Instrument, Woonsocket, RI, USA). 2.3. Determination of chemical characteristics of beans during the fermentation process 2.3.1. Physical analysis. At the end of the fermentation and drying process, a cutting test was carried out according to the NTC 1252 2021 edition [22] to determine the colorations and defects present. For this purpose, 100 cocoa beans were randomly selected from each fermented cocoa pod at each collection center, cut longitudinally with the help of a CocoaT Bean knife (Cocoatown, Alpharetta, Georgia, USA) and the number of beans that were completely fermented, purple, moldy, damaged by insects, and slaty were quantified. 2.3.2. Bromatological composition. Total fat content was determined by the Randall method with petroleum ether (AOAC Method No.991.36) [23], using the semi-automatic solvent extractor (SER 148 VELP Scientific, Italy) whose results were expressed as percentage of fat. Ash content was determined by incinerating the organic matter in a muffle furnace (1,100°C, 22.9A Fisher Scientific, Spain) at 550 ± 5°C until the sample was free of carbon, cooled in a desiccator and the amount of ash was calculated (AOAC Method No. 930.30) [24], the results were expressed as percentage of ash. Moisture content was determined using the methodology proposed by Nuñez et al. [25] with modifications. A crucible was brought to constant weight, 10 g of dry cocoa beans were weighed into the crucible and left in the oven (Heratherm OMH400, Fisher Scientific, Spain) at 105 ± 5°C for 4 hours. The results were expressed as percent of moisture. Titratable acidity was measured according to AOAC method 939.05 [24]. Extracts of dried cocoa beans (5 g) were homogenized in 50 mL of deionized water, subsequently filtered through filter paper (3hw, 110 mm, 65 g/m2; Boeco, Germany) and titrated with a standard 0.1 N sodium hydroxide (NaOH) solution to a pH of 8.3. The results were expressed as percent of acidity. 2.3.3. Polyphenolic compounds. First, the methanolic extract of defatted cocoa beans was obtained by weighing 3 g and depositing them in falcon tubes, adding 15 mL of methanolic solution (25% HPLC grade methanol, 24% deionized water and 1% HPLC grade formic acid). The samples were homogenized for 25 minutes using vortex and ultrasound for 20 minutes. Subsequently, they were left in darkness for 24 h and centrifuged for 15 min at 4,500 rpm (4°C) and filtered with filter paper (3hw, 110 mm, 65 g/m2; Boeco, Germany). With this extract obtained from each sample, the contents of total polyphenols, total flavonoids, DPPH, FRAP and methylxanthine contents were analyzed in triplicate. The determination of the total polyphenol content was carried out using the Folin-Ciocalteu colorimetric method [26]. Eighteen μL of the extract, 124.5 μL of deionized water, 37.5 μL of Folin-Ciocalteu reagent and 120 μL of 7.1% anhydrous sodium carbonate (Na2CO3) were taken. It was left to react for 60 min in the dark at room temperature, after which the absorbance was read at 760 nm. Gallic acid was used as a standard. Results were expressed as mg gallic acid equivalent (mg GAE)/g dried cocoa bean. Total flavonoid content was determined by reaction with aluminum chloride (AlCl3) according to the methodology proposed by Zhishen et al. [27] with slight modifications. The reaction mixture consisted of 120 μL of deionized water, then 30 μL of the extract was added, followed by 9 μL of 5% sodium nitrite (NaNO2) (waited 5 minutes), 9 μL of 10% aluminum chloride (AlCl3) (waited 5 minutes), then 60 μL of 1M sodium hydroxide (NaOH) (waited 15 minutes) and finally 72 μL of deionized water. It was left to react in the dark at room temperature for 30 minutes and the absorbance was read at 510 nm. The (+)-catechin was used as a standard for the quantification of total flavonoids. Results were expressed as mg catechin equivalent (mgCE)/g dried cocoa bean. Quantification of methylxanthines and epicatechin were developed on an Ultimate 3000 HPLC, equipped with an auto-injection system and UV-VIS detector, analytical reverse phase column (Zorbax Eclipse XDB 150mm × 2.1mm) with particle size of 5μm, at 25°C. All compounds were detected at a wavelength of 273 nm. The mobile phase was water/acetic acid (99.7/0.3 v/v) (solvent A) and methanol (Solvent B), flow rate 0.5 mL/min. The gradient was as follows: 0–10 min, 15% linear B; 10.1–18 min, 25% linear B; 18.1–25 min, 30% linear B; 25.1–30 min, 100% linear B; 30.1–35 min, 0% linear B, followed by 5 min of column re-equilibration before a new injection with an injection volume of 5μL. All analytes were identified and quantified by the external standard method using calibration curves of the standard substances [28]. 2.3.4. Antioxidant activity. The 1,1-diphenyl-2-picrylhydrazil (DPPH) radical scavenging activity was used according to the method of Brand-Williams et al. [29] with slight modifications. A stock solution of DPPH (20 mg/L) was prepared in absolute methanol; the absorbance of the radical was adjusted to 0.3 absorbance units with methanol at 4°C, then 3 μL of the extract and 297 μL of the adjusted DPPH solution were taken. It was allowed to react in the dark for 30 min at room temperature and the absorbance was read at a wavelength of 517 nm. The results were expressed as Trolox equivalent antioxidant capacity (TEAC) values in μmol of Trolox (μmol Trolox)/g of dry cocoa beans, constructing a reference curve using Trolox as an antioxidant. The reducing capacity FRAP (ferric reducing antioxidant power) evaluates the antioxidant capacity of a sample according to its ability to reduce ferric iron (Fe+3) present in a complex with 2,4,6-tri(2-pyridyl)-s-triazine (TPTZ), to the ferrous form (Fe+3) [30]. The assay was carried out in a pH 3.6 acetic acid-sodium acetate buffer containing TPTZ and FeCl3. 15 μL of the extract, 15 μL of buffer and 270 μL of FRAP solution were used as samples. It was allowed to react in the dark for 30 min at room temperature and the absorbance was read at a wavelength of 590 nm. The FRAP values were expressed as mg ascorbic acid (mg AA/g of dry cocoa bean), based on a reference curve of ascorbic acid as the primary standard. 2.3.5. Sensory analysis. The cocoa paste samples (cocoa bean roasted and milled) were analyzed by a sensory panel following the process described by ICONTEC [22]. The roasting curve was constructed according to bean size and moisture percentage; the roasted beans were also passed through a grinder and the husk was separated from the nibs. Finally, the nibs were processed and refined by a melangeur for the manufacture of cocoa paste/liquor. Sensory evaluation was carried out with the help of five trained panelists, and the main qualities were determined: cocoa, acidity, astringency, bitterness, and notes of fresh fruit, brown, floral, wood, spice and nut, as well as atypical flavors in the samples. The order of perception, intensity, residual flavor and persistence were established as scores given by the panel. The international cocoa evaluation scale of excellence [31] was used, which scores from 0 to 10 points, where 0 indicates the complete absence of the evaluated attribute and 10 indicates a very high intensity. 2.4. Data analysis and experimental design A completely randomized plot design with factorial arrangement was used (i. cocoa bean processing plants called "Sites" corresponding to ASOACASAN ASA, COMICACAO CMI and COMCAP COC; and ii. fermentation time, corresponding to samples of cocoa beans obtained at 4 and 7 days after the start of the fermentation process) using three repetitions with the objective of analyzing the impact on the physics of the grain, bromatological composition, polyphenolic compound content and antioxidant activity as sensory attributes. Each repetition corresponded to a mass of cocoa bean with mucilage carried out independently in each fermentation box at each site from which a sample was obtained for analysis 4 and 7 days after the start of the fermentation process. The data were analyzed by fitting a linear mixed model (LMM) where in the fixed factor the sites (ASA, CMI and COC) and the fermentation time (days 4 and 7) were adjusted, and within the random factor the repetition (mass of cocoa bean) within the sites. Assumptions of normality and homogeneity of variance were examined through an exploratory analysis of residuals. In addition, the LSD Fisher test (P< 0.05) was performed to determine if differences existed between fixed factors. Line graphs were also made to show the behavior of temperature and pH during fermentation. Subsequently, radar plots were made to visualize the sensory attributes of the sites by days of fermentation. In addition, Pearson’s correlation using the “corr” package [32] was employed to establish correlations between antioxidant activity, polyphenolic compounds and methylxanthines. Finally, a principal component analysis (PCA) was performed to determine the multivariate relationships between the variables evaluated and the study sites. The models were applied using the statistical software InfoStat [33], PCA and Pearson correlation were performed using the packages ade4, ggplot2 [34], factoextra [35], FactoMineR [36] in the R language software, using the RStudio interface [37]. 2.1. Study area The study was carried out in the department of Caquetá, specifically in the cocoa bean processing plants of a) Asociación Orgánica Agrícola de San José del Fragua-ASOACASAN (ASA) (1°19′43″N 75°58′22″W), b) Comité de Cultivadores de Cacao en Sistemas Agroforestales del municipio de San Vicente del Caguán-COMICACAO (CMI) (2°06′55″N 74°46′12″W) and c) Comité de Cacaoteros de los municipios del Paujil y el Doncello-COMCAP (COC) (1°40′26″ N 75°16′48″W). The environmental conditions in which these processing plants are found mainly contrast in the amount of average annual precipitation and the relative humidity (RH), variables that can affect the drying time. For example, in ASOACAN, which is located further north in the department of Caquetá, it has an average annual rainfall of 4,200 mm with a RH of 85% and handles the mass (560 kg of cocoa beans with mucilage) in relation to 7 days of fermentation with 6 subsequent turnings of the anaerobic phase and uses 14 days in the drying process. COMICACAO, very north of the department of Caquetá, has an average annual rainfall of 2,700 mm with a RH of 83%, it handles the mass (270 kg of cocoa beans with mucilage) very similar to ASOACASAN but with the main difference is that they dry only in 10 days. Finally, COMCAP, which is located 90 km from COMICACAO to the north and 128 km to the south with ASOACASAN, presents an average annual rainfall of 3,800 mm with a RH of 82%, carries out mass management (340 kg of cocoa beans with mucilage) with 8 days of fermentation with 7 turning and 12 days of drying. All cocoa processing plants carry out the turning of the cocoa bean during fermentation every 24 hours and the process of drying the bean after fermentation in canopies at room temperature. 2.2. Monitoring fermentation and drying of cocoa beans The cocoa beans were obtained from healthy pods from trees of different universal, national and regional clones obtained from farms associated with the organizations. Both ASOACASAN and COMICACAO make mixtures of cocoa beans from clones such as CCN51, ICS1, ICS95, TSH565 and FEAR5, however the difference between these associations is the additional use in FLE2 and ICS39 for ASOACASAN and COMICACAO, respectively. On the other hand, COMCAP carries out fermentation processes for cocoa beans obtained from raw materials. This process was carried out between March and June 2023, when production peaked. The beans were then fermented and dried according to the protocol of each collection center. The cocoa beans were placed in wooden crates, which were covered with banana leaves and natural jute fiber sacks to maintain a homogeneous temperature. These crates were perforated at the bottom to allow drainage of the cocoa pulp. The fermentation process lasted 168 hours (7 days) and during this time the fermentable mass was turned every 24 hours after 48 hours of fermentation. The turning of the grain was done by passing the dough into the compartments of the fermenter box, a process that was carried out using a plastic shovel. This fermentation process was carried out in triplicate, that is, three fermentation bins were used for each storage center. During fermentation, the following parameters were monitored daily between 13:00 and 16:00 hours: mass temperature, pH of the grain pulp and pH of the cotyledon. Samples for this monitoring were taken in the center and at the ends of the bin at a depth of 30 cm. This process was carried out in triplicate. The temperature of the dough was determined using a digital thermometer Hl 145 (Hanna Instrument, Woonsocket, RI, USA), the pH of the pulp during fermentation was measured using a pH tester Hl 98108 (Hanna Instrument, Woonsocket, RI, USA) and the pH of the cotyledon was performed following the methodology used by Papalexandratou [21]. For which, 15 g of randomly selected cocoa beans were taken from the three points of the bin; then, the pulp and testa were manually removed from the beans with a knife and the cotyledon was macerated with 30 ml of deionized water in a mortar for one minute until a homogeneous sample was obtained and the pH was read using a pH tester Hl 98108 (Hanna Instrument, Woonsocket, RI, USA). 2.3. Determination of chemical characteristics of beans during the fermentation process 2.3.1. Physical analysis. At the end of the fermentation and drying process, a cutting test was carried out according to the NTC 1252 2021 edition [22] to determine the colorations and defects present. For this purpose, 100 cocoa beans were randomly selected from each fermented cocoa pod at each collection center, cut longitudinally with the help of a CocoaT Bean knife (Cocoatown, Alpharetta, Georgia, USA) and the number of beans that were completely fermented, purple, moldy, damaged by insects, and slaty were quantified. 2.3.2. Bromatological composition. Total fat content was determined by the Randall method with petroleum ether (AOAC Method No.991.36) [23], using the semi-automatic solvent extractor (SER 148 VELP Scientific, Italy) whose results were expressed as percentage of fat. Ash content was determined by incinerating the organic matter in a muffle furnace (1,100°C, 22.9A Fisher Scientific, Spain) at 550 ± 5°C until the sample was free of carbon, cooled in a desiccator and the amount of ash was calculated (AOAC Method No. 930.30) [24], the results were expressed as percentage of ash. Moisture content was determined using the methodology proposed by Nuñez et al. [25] with modifications. A crucible was brought to constant weight, 10 g of dry cocoa beans were weighed into the crucible and left in the oven (Heratherm OMH400, Fisher Scientific, Spain) at 105 ± 5°C for 4 hours. The results were expressed as percent of moisture. Titratable acidity was measured according to AOAC method 939.05 [24]. Extracts of dried cocoa beans (5 g) were homogenized in 50 mL of deionized water, subsequently filtered through filter paper (3hw, 110 mm, 65 g/m2; Boeco, Germany) and titrated with a standard 0.1 N sodium hydroxide (NaOH) solution to a pH of 8.3. The results were expressed as percent of acidity. 2.3.3. Polyphenolic compounds. First, the methanolic extract of defatted cocoa beans was obtained by weighing 3 g and depositing them in falcon tubes, adding 15 mL of methanolic solution (25% HPLC grade methanol, 24% deionized water and 1% HPLC grade formic acid). The samples were homogenized for 25 minutes using vortex and ultrasound for 20 minutes. Subsequently, they were left in darkness for 24 h and centrifuged for 15 min at 4,500 rpm (4°C) and filtered with filter paper (3hw, 110 mm, 65 g/m2; Boeco, Germany). With this extract obtained from each sample, the contents of total polyphenols, total flavonoids, DPPH, FRAP and methylxanthine contents were analyzed in triplicate. The determination of the total polyphenol content was carried out using the Folin-Ciocalteu colorimetric method [26]. Eighteen μL of the extract, 124.5 μL of deionized water, 37.5 μL of Folin-Ciocalteu reagent and 120 μL of 7.1% anhydrous sodium carbonate (Na2CO3) were taken. It was left to react for 60 min in the dark at room temperature, after which the absorbance was read at 760 nm. Gallic acid was used as a standard. Results were expressed as mg gallic acid equivalent (mg GAE)/g dried cocoa bean. Total flavonoid content was determined by reaction with aluminum chloride (AlCl3) according to the methodology proposed by Zhishen et al. [27] with slight modifications. The reaction mixture consisted of 120 μL of deionized water, then 30 μL of the extract was added, followed by 9 μL of 5% sodium nitrite (NaNO2) (waited 5 minutes), 9 μL of 10% aluminum chloride (AlCl3) (waited 5 minutes), then 60 μL of 1M sodium hydroxide (NaOH) (waited 15 minutes) and finally 72 μL of deionized water. It was left to react in the dark at room temperature for 30 minutes and the absorbance was read at 510 nm. The (+)-catechin was used as a standard for the quantification of total flavonoids. Results were expressed as mg catechin equivalent (mgCE)/g dried cocoa bean. Quantification of methylxanthines and epicatechin were developed on an Ultimate 3000 HPLC, equipped with an auto-injection system and UV-VIS detector, analytical reverse phase column (Zorbax Eclipse XDB 150mm × 2.1mm) with particle size of 5μm, at 25°C. All compounds were detected at a wavelength of 273 nm. The mobile phase was water/acetic acid (99.7/0.3 v/v) (solvent A) and methanol (Solvent B), flow rate 0.5 mL/min. The gradient was as follows: 0–10 min, 15% linear B; 10.1–18 min, 25% linear B; 18.1–25 min, 30% linear B; 25.1–30 min, 100% linear B; 30.1–35 min, 0% linear B, followed by 5 min of column re-equilibration before a new injection with an injection volume of 5μL. All analytes were identified and quantified by the external standard method using calibration curves of the standard substances [28]. 2.3.4. Antioxidant activity. The 1,1-diphenyl-2-picrylhydrazil (DPPH) radical scavenging activity was used according to the method of Brand-Williams et al. [29] with slight modifications. A stock solution of DPPH (20 mg/L) was prepared in absolute methanol; the absorbance of the radical was adjusted to 0.3 absorbance units with methanol at 4°C, then 3 μL of the extract and 297 μL of the adjusted DPPH solution were taken. It was allowed to react in the dark for 30 min at room temperature and the absorbance was read at a wavelength of 517 nm. The results were expressed as Trolox equivalent antioxidant capacity (TEAC) values in μmol of Trolox (μmol Trolox)/g of dry cocoa beans, constructing a reference curve using Trolox as an antioxidant. The reducing capacity FRAP (ferric reducing antioxidant power) evaluates the antioxidant capacity of a sample according to its ability to reduce ferric iron (Fe+3) present in a complex with 2,4,6-tri(2-pyridyl)-s-triazine (TPTZ), to the ferrous form (Fe+3) [30]. The assay was carried out in a pH 3.6 acetic acid-sodium acetate buffer containing TPTZ and FeCl3. 15 μL of the extract, 15 μL of buffer and 270 μL of FRAP solution were used as samples. It was allowed to react in the dark for 30 min at room temperature and the absorbance was read at a wavelength of 590 nm. The FRAP values were expressed as mg ascorbic acid (mg AA/g of dry cocoa bean), based on a reference curve of ascorbic acid as the primary standard. 2.3.5. Sensory analysis. The cocoa paste samples (cocoa bean roasted and milled) were analyzed by a sensory panel following the process described by ICONTEC [22]. The roasting curve was constructed according to bean size and moisture percentage; the roasted beans were also passed through a grinder and the husk was separated from the nibs. Finally, the nibs were processed and refined by a melangeur for the manufacture of cocoa paste/liquor. Sensory evaluation was carried out with the help of five trained panelists, and the main qualities were determined: cocoa, acidity, astringency, bitterness, and notes of fresh fruit, brown, floral, wood, spice and nut, as well as atypical flavors in the samples. The order of perception, intensity, residual flavor and persistence were established as scores given by the panel. The international cocoa evaluation scale of excellence [31] was used, which scores from 0 to 10 points, where 0 indicates the complete absence of the evaluated attribute and 10 indicates a very high intensity. 2.3.1. Physical analysis. At the end of the fermentation and drying process, a cutting test was carried out according to the NTC 1252 2021 edition [22] to determine the colorations and defects present. For this purpose, 100 cocoa beans were randomly selected from each fermented cocoa pod at each collection center, cut longitudinally with the help of a CocoaT Bean knife (Cocoatown, Alpharetta, Georgia, USA) and the number of beans that were completely fermented, purple, moldy, damaged by insects, and slaty were quantified. 2.3.2. Bromatological composition. Total fat content was determined by the Randall method with petroleum ether (AOAC Method No.991.36) [23], using the semi-automatic solvent extractor (SER 148 VELP Scientific, Italy) whose results were expressed as percentage of fat. Ash content was determined by incinerating the organic matter in a muffle furnace (1,100°C, 22.9A Fisher Scientific, Spain) at 550 ± 5°C until the sample was free of carbon, cooled in a desiccator and the amount of ash was calculated (AOAC Method No. 930.30) [24], the results were expressed as percentage of ash. Moisture content was determined using the methodology proposed by Nuñez et al. [25] with modifications. A crucible was brought to constant weight, 10 g of dry cocoa beans were weighed into the crucible and left in the oven (Heratherm OMH400, Fisher Scientific, Spain) at 105 ± 5°C for 4 hours. The results were expressed as percent of moisture. Titratable acidity was measured according to AOAC method 939.05 [24]. Extracts of dried cocoa beans (5 g) were homogenized in 50 mL of deionized water, subsequently filtered through filter paper (3hw, 110 mm, 65 g/m2; Boeco, Germany) and titrated with a standard 0.1 N sodium hydroxide (NaOH) solution to a pH of 8.3. The results were expressed as percent of acidity. 2.3.3. Polyphenolic compounds. First, the methanolic extract of defatted cocoa beans was obtained by weighing 3 g and depositing them in falcon tubes, adding 15 mL of methanolic solution (25% HPLC grade methanol, 24% deionized water and 1% HPLC grade formic acid). The samples were homogenized for 25 minutes using vortex and ultrasound for 20 minutes. Subsequently, they were left in darkness for 24 h and centrifuged for 15 min at 4,500 rpm (4°C) and filtered with filter paper (3hw, 110 mm, 65 g/m2; Boeco, Germany). With this extract obtained from each sample, the contents of total polyphenols, total flavonoids, DPPH, FRAP and methylxanthine contents were analyzed in triplicate. The determination of the total polyphenol content was carried out using the Folin-Ciocalteu colorimetric method [26]. Eighteen μL of the extract, 124.5 μL of deionized water, 37.5 μL of Folin-Ciocalteu reagent and 120 μL of 7.1% anhydrous sodium carbonate (Na2CO3) were taken. It was left to react for 60 min in the dark at room temperature, after which the absorbance was read at 760 nm. Gallic acid was used as a standard. Results were expressed as mg gallic acid equivalent (mg GAE)/g dried cocoa bean. Total flavonoid content was determined by reaction with aluminum chloride (AlCl3) according to the methodology proposed by Zhishen et al. [27] with slight modifications. The reaction mixture consisted of 120 μL of deionized water, then 30 μL of the extract was added, followed by 9 μL of 5% sodium nitrite (NaNO2) (waited 5 minutes), 9 μL of 10% aluminum chloride (AlCl3) (waited 5 minutes), then 60 μL of 1M sodium hydroxide (NaOH) (waited 15 minutes) and finally 72 μL of deionized water. It was left to react in the dark at room temperature for 30 minutes and the absorbance was read at 510 nm. The (+)-catechin was used as a standard for the quantification of total flavonoids. Results were expressed as mg catechin equivalent (mgCE)/g dried cocoa bean. Quantification of methylxanthines and epicatechin were developed on an Ultimate 3000 HPLC, equipped with an auto-injection system and UV-VIS detector, analytical reverse phase column (Zorbax Eclipse XDB 150mm × 2.1mm) with particle size of 5μm, at 25°C. All compounds were detected at a wavelength of 273 nm. The mobile phase was water/acetic acid (99.7/0.3 v/v) (solvent A) and methanol (Solvent B), flow rate 0.5 mL/min. The gradient was as follows: 0–10 min, 15% linear B; 10.1–18 min, 25% linear B; 18.1–25 min, 30% linear B; 25.1–30 min, 100% linear B; 30.1–35 min, 0% linear B, followed by 5 min of column re-equilibration before a new injection with an injection volume of 5μL. All analytes were identified and quantified by the external standard method using calibration curves of the standard substances [28]. 2.3.4. Antioxidant activity. The 1,1-diphenyl-2-picrylhydrazil (DPPH) radical scavenging activity was used according to the method of Brand-Williams et al. [29] with slight modifications. A stock solution of DPPH (20 mg/L) was prepared in absolute methanol; the absorbance of the radical was adjusted to 0.3 absorbance units with methanol at 4°C, then 3 μL of the extract and 297 μL of the adjusted DPPH solution were taken. It was allowed to react in the dark for 30 min at room temperature and the absorbance was read at a wavelength of 517 nm. The results were expressed as Trolox equivalent antioxidant capacity (TEAC) values in μmol of Trolox (μmol Trolox)/g of dry cocoa beans, constructing a reference curve using Trolox as an antioxidant. The reducing capacity FRAP (ferric reducing antioxidant power) evaluates the antioxidant capacity of a sample according to its ability to reduce ferric iron (Fe+3) present in a complex with 2,4,6-tri(2-pyridyl)-s-triazine (TPTZ), to the ferrous form (Fe+3) [30]. The assay was carried out in a pH 3.6 acetic acid-sodium acetate buffer containing TPTZ and FeCl3. 15 μL of the extract, 15 μL of buffer and 270 μL of FRAP solution were used as samples. It was allowed to react in the dark for 30 min at room temperature and the absorbance was read at a wavelength of 590 nm. The FRAP values were expressed as mg ascorbic acid (mg AA/g of dry cocoa bean), based on a reference curve of ascorbic acid as the primary standard. 2.3.5. Sensory analysis. The cocoa paste samples (cocoa bean roasted and milled) were analyzed by a sensory panel following the process described by ICONTEC [22]. The roasting curve was constructed according to bean size and moisture percentage; the roasted beans were also passed through a grinder and the husk was separated from the nibs. Finally, the nibs were processed and refined by a melangeur for the manufacture of cocoa paste/liquor. Sensory evaluation was carried out with the help of five trained panelists, and the main qualities were determined: cocoa, acidity, astringency, bitterness, and notes of fresh fruit, brown, floral, wood, spice and nut, as well as atypical flavors in the samples. The order of perception, intensity, residual flavor and persistence were established as scores given by the panel. The international cocoa evaluation scale of excellence [31] was used, which scores from 0 to 10 points, where 0 indicates the complete absence of the evaluated attribute and 10 indicates a very high intensity. 2.4. Data analysis and experimental design A completely randomized plot design with factorial arrangement was used (i. cocoa bean processing plants called "Sites" corresponding to ASOACASAN ASA, COMICACAO CMI and COMCAP COC; and ii. fermentation time, corresponding to samples of cocoa beans obtained at 4 and 7 days after the start of the fermentation process) using three repetitions with the objective of analyzing the impact on the physics of the grain, bromatological composition, polyphenolic compound content and antioxidant activity as sensory attributes. Each repetition corresponded to a mass of cocoa bean with mucilage carried out independently in each fermentation box at each site from which a sample was obtained for analysis 4 and 7 days after the start of the fermentation process. The data were analyzed by fitting a linear mixed model (LMM) where in the fixed factor the sites (ASA, CMI and COC) and the fermentation time (days 4 and 7) were adjusted, and within the random factor the repetition (mass of cocoa bean) within the sites. Assumptions of normality and homogeneity of variance were examined through an exploratory analysis of residuals. In addition, the LSD Fisher test (P< 0.05) was performed to determine if differences existed between fixed factors. Line graphs were also made to show the behavior of temperature and pH during fermentation. Subsequently, radar plots were made to visualize the sensory attributes of the sites by days of fermentation. In addition, Pearson’s correlation using the “corr” package [32] was employed to establish correlations between antioxidant activity, polyphenolic compounds and methylxanthines. Finally, a principal component analysis (PCA) was performed to determine the multivariate relationships between the variables evaluated and the study sites. The models were applied using the statistical software InfoStat [33], PCA and Pearson correlation were performed using the packages ade4, ggplot2 [34], factoextra [35], FactoMineR [36] in the R language software, using the RStudio interface [37]. 3. Results 3.1. Temperature and pH of fermented cocoa beans During the fermentation period, the temperature increased until 4 days, reaching an average of 45°C, after which it stabilized (Fig 1A). The pH of the pulp did not vary significantly during the fermentation period, however, there was a small variation at 2 and 3 days between sites (Fig 1B). The pH in the cotyledon at the beginning of the fermentation process ranged from 5.6 to 6.3 with a steady decrease to 4.5 (Fig 1C). Download: PPT PowerPoint slide PNG larger image TIFF original image Fig 1. Daily behavior of three variables during cocoa fermentation in three processing plants in the department of Caquetá. ASA: ASOACASAN, CMI: COMICACAO, COC: COMCAP: a) mass temperature, b) pulp pH, c) cotyledon pH. Values correspond to means and standard errors (n = 3). https://doi.org/10.1371/journal.pone.0306680.g001 3.2. Physical properties of dry cocoa beans The fermentation process between sites varied (Table 1, P<0.05), a situation that resulted in the number of properly fermented grains; however, no difference was found between fermentation hours. On the contrary, differences between sampling hours were found in the number of violet and partially fermented grains in ASA and totally fermented grains in COC. Between sites, the main difference was in the amount of violet and partially fermented grains (Table 1, P<0.05). Download: PPT PowerPoint slide PNG larger image TIFF original image Table 1. Results (%) of the shear test on fermented and dried cocoa beans at two fermentation times. https://doi.org/10.1371/journal.pone.0306680.t001 3.3. Component bromatological In the different variables of the bromatological component, significant differences were found at the level of fermentation time as well as at the sites (Table 2, P<0.05). For example, fat content and moisture content varied with fermentation time, being significantly higher at seven days; the opposite was true for ash (Table 2, P<0.05). As for the titratable acidity level, no differences were found between fermentation days. Download: PPT PowerPoint slide PNG larger image TIFF original image Table 2. Bromatological composition of fermented and dried cocoa beans at two fermentation times. https://doi.org/10.1371/journal.pone.0306680.t002 3.4. Polyphenolic compounds At the site level, TPC, TF, DPPH and FRAP contents were statistically different, with COC being different from the other sites (P<0.05, Fig 2). Total polyphenol content was higher at the COC site (507.05 mg GAE/g Cocoa Fig 2A) with respect to the other sites (< 360 mg GAE/g Cocoa). The TF content followed a similar behavior to TPC, with significant differences between sites (P<0.05) and differences between fermentation times for ASA. The TF was higher in COC (309.1 mg catechin/g cocoa, Fig 2B) with respect to CMI (215.59 mg catechin/g cocoa) and ASA (185.7 mg catechin/g cocoa). Values in DPPH (Fig 2C) ranged from 5869.3 to 7781.8 μmol Trolox/g cocoa and for the FRAP assay ranged from 369.8 to 606.7 mg AA/g cocoa among the sites (Fig 2D). At the methylxanthine level, theobromine was different between sites and only different at the level of hours of fermentation in COC (P<0.05, Fig 2E). In the case of caffeine and epicatechin, no differences were found between sites (Fig 2F and 2G); however, only for caffeine there were differences between the fermentation hours for the three sites, while for epicatechin only COC showed differences in fermentation hours. Download: PPT PowerPoint slide PNG larger image TIFF original image Fig 2. Content of polyphenols, methylxanthines and antioxidant activity in fermented cocoa beans. Different letters indicate statistically significant differences by LSD Fisher means test (P< 0.05). Letters a, b, c means significant and non-significant ns differences between the time of fermentation at each site and A, B, C: between the different sites. ASA: ASOACASAN, CMI: COMICACAO, COC: COMCAP. https://doi.org/10.1371/journal.pone.0306680.g002 3.5. Sensory profile In general, the presence of complementary attributes such as nutty, spicy, woody, floral and fresh fruit was found at all sites (Fig 3). It was evident that COC presented higher acidity (5.5) with respect to ASA (3.67) and CMI (3.33) on the fourth day of fermentation; however, on day 7 acidity decreased in COC (4.5) but increased in CMI (5.0). Astringency and bitterness increased with days of fermentation in ASA and COC, while astringency and bitterness decreased with the course of fermentation in CMI. Download: PPT PowerPoint slide PNG larger image TIFF original image Fig 3. Sensory profile of cocoa liquor in three processing plants: a. ASA: ASOACASAN, b. CMI: COMICACAO, c. COC: COMCAP. https://doi.org/10.1371/journal.pone.0306680.g003 Fig 4 shows the different positive and negative correlations between the different variables analyzed. A positive correlation was found between polyphenolic compounds and antioxidants with theobromine. Likewise, titratable acidity had a positive correlation with total polyphenols, total flavonoids and variables related to antioxidant capacity (FRAP and DPPH). On the other hand, caffeine was negatively correlated with DPPH, theobromine and epicatechin (Fig 4). Download: PPT PowerPoint slide PNG larger image TIFF original image Fig 4. Correlation analysis between the different variables analyzed. The numbers shown presented significant correlations (P<0.05) in the color gradient from red to blue showing negative and positive correlation, respectively. https://doi.org/10.1371/journal.pone.0306680.g004 Principal component analysis explained 43.8% of the variance (PCA, Fig 5), with CP1 explaining 27.2% and opposing the ASA and COC sites by variables such as moisture, titratable acidity as well as phenolic compound content and antioxidant activity. CP2 explained 16.6%, opposing different sensory attributes such as cocoa and bitterness. According to the Monte-Carlo test, the sites explained 24.9% of the variance (Fig 5). Download: PPT PowerPoint slide PNG larger image TIFF original image Fig 5. PCA projection of variables related to bromatological composition, polyphenol content, methylxanthines and antioxidant activity in cocoa beans fermented in different cocoa bean processing plants in the department of Caquetá, Colombia. ASA: ASOACASAN, CMI: COMICACAO, COC: COMCAP. Contribution of each of the variables evaluated in the PC1/PC2 principal components of PCA, the gradient from green to red means from greater to lesser contribution. https://doi.org/10.1371/journal.pone.0306680.g005 3.1. Temperature and pH of fermented cocoa beans During the fermentation period, the temperature increased until 4 days, reaching an average of 45°C, after which it stabilized (Fig 1A). The pH of the pulp did not vary significantly during the fermentation period, however, there was a small variation at 2 and 3 days between sites (Fig 1B). The pH in the cotyledon at the beginning of the fermentation process ranged from 5.6 to 6.3 with a steady decrease to 4.5 (Fig 1C). Download: PPT PowerPoint slide PNG larger image TIFF original image Fig 1. Daily behavior of three variables during cocoa fermentation in three processing plants in the department of Caquetá. ASA: ASOACASAN, CMI: COMICACAO, COC: COMCAP: a) mass temperature, b) pulp pH, c) cotyledon pH. Values correspond to means and standard errors (n = 3). https://doi.org/10.1371/journal.pone.0306680.g001 3.2. Physical properties of dry cocoa beans The fermentation process between sites varied (Table 1, P<0.05), a situation that resulted in the number of properly fermented grains; however, no difference was found between fermentation hours. On the contrary, differences between sampling hours were found in the number of violet and partially fermented grains in ASA and totally fermented grains in COC. Between sites, the main difference was in the amount of violet and partially fermented grains (Table 1, P<0.05). Download: PPT PowerPoint slide PNG larger image TIFF original image Table 1. Results (%) of the shear test on fermented and dried cocoa beans at two fermentation times. https://doi.org/10.1371/journal.pone.0306680.t001 3.3. Component bromatological In the different variables of the bromatological component, significant differences were found at the level of fermentation time as well as at the sites (Table 2, P<0.05). For example, fat content and moisture content varied with fermentation time, being significantly higher at seven days; the opposite was true for ash (Table 2, P<0.05). As for the titratable acidity level, no differences were found between fermentation days. Download: PPT PowerPoint slide PNG larger image TIFF original image Table 2. Bromatological composition of fermented and dried cocoa beans at two fermentation times. https://doi.org/10.1371/journal.pone.0306680.t002 3.4. Polyphenolic compounds At the site level, TPC, TF, DPPH and FRAP contents were statistically different, with COC being different from the other sites (P<0.05, Fig 2). Total polyphenol content was higher at the COC site (507.05 mg GAE/g Cocoa Fig 2A) with respect to the other sites (< 360 mg GAE/g Cocoa). The TF content followed a similar behavior to TPC, with significant differences between sites (P<0.05) and differences between fermentation times for ASA. The TF was higher in COC (309.1 mg catechin/g cocoa, Fig 2B) with respect to CMI (215.59 mg catechin/g cocoa) and ASA (185.7 mg catechin/g cocoa). Values in DPPH (Fig 2C) ranged from 5869.3 to 7781.8 μmol Trolox/g cocoa and for the FRAP assay ranged from 369.8 to 606.7 mg AA/g cocoa among the sites (Fig 2D). At the methylxanthine level, theobromine was different between sites and only different at the level of hours of fermentation in COC (P<0.05, Fig 2E). In the case of caffeine and epicatechin, no differences were found between sites (Fig 2F and 2G); however, only for caffeine there were differences between the fermentation hours for the three sites, while for epicatechin only COC showed differences in fermentation hours. Download: PPT PowerPoint slide PNG larger image TIFF original image Fig 2. Content of polyphenols, methylxanthines and antioxidant activity in fermented cocoa beans. Different letters indicate statistically significant differences by LSD Fisher means test (P< 0.05). Letters a, b, c means significant and non-significant ns differences between the time of fermentation at each site and A, B, C: between the different sites. ASA: ASOACASAN, CMI: COMICACAO, COC: COMCAP. https://doi.org/10.1371/journal.pone.0306680.g002 3.5. Sensory profile In general, the presence of complementary attributes such as nutty, spicy, woody, floral and fresh fruit was found at all sites (Fig 3). It was evident that COC presented higher acidity (5.5) with respect to ASA (3.67) and CMI (3.33) on the fourth day of fermentation; however, on day 7 acidity decreased in COC (4.5) but increased in CMI (5.0). Astringency and bitterness increased with days of fermentation in ASA and COC, while astringency and bitterness decreased with the course of fermentation in CMI. Download: PPT PowerPoint slide PNG larger image TIFF original image Fig 3. Sensory profile of cocoa liquor in three processing plants: a. ASA: ASOACASAN, b. CMI: COMICACAO, c. COC: COMCAP. https://doi.org/10.1371/journal.pone.0306680.g003 Fig 4 shows the different positive and negative correlations between the different variables analyzed. A positive correlation was found between polyphenolic compounds and antioxidants with theobromine. Likewise, titratable acidity had a positive correlation with total polyphenols, total flavonoids and variables related to antioxidant capacity (FRAP and DPPH). On the other hand, caffeine was negatively correlated with DPPH, theobromine and epicatechin (Fig 4). Download: PPT PowerPoint slide PNG larger image TIFF original image Fig 4. Correlation analysis between the different variables analyzed. The numbers shown presented significant correlations (P<0.05) in the color gradient from red to blue showing negative and positive correlation, respectively. https://doi.org/10.1371/journal.pone.0306680.g004 Principal component analysis explained 43.8% of the variance (PCA, Fig 5), with CP1 explaining 27.2% and opposing the ASA and COC sites by variables such as moisture, titratable acidity as well as phenolic compound content and antioxidant activity. CP2 explained 16.6%, opposing different sensory attributes such as cocoa and bitterness. According to the Monte-Carlo test, the sites explained 24.9% of the variance (Fig 5). Download: PPT PowerPoint slide PNG larger image TIFF original image Fig 5. PCA projection of variables related to bromatological composition, polyphenol content, methylxanthines and antioxidant activity in cocoa beans fermented in different cocoa bean processing plants in the department of Caquetá, Colombia. ASA: ASOACASAN, CMI: COMICACAO, COC: COMCAP. Contribution of each of the variables evaluated in the PC1/PC2 principal components of PCA, the gradient from green to red means from greater to lesser contribution. https://doi.org/10.1371/journal.pone.0306680.g005 4. Discussion 4.1. Temperature and pH of fermented cocoa beans An increase in temperature was found in the fermentation mass due to the activity carried out by the yeasts involved in the metabolization of sugars in the cocoa pulp (exothermic process), to produce ethanol and other metabolities (other alcohols, esters, aldehydes and ketones), a process that releases heat and increases the fermentation temperature [38]. It was observed that from the fifth day of fermentation the temperature did not vary, reaching a temperature between 40–50°C, a desirable range for good fermentation [39]. Although the environmental temperature was not a condition that varied, a significant increase in the temperature of the cocoa fermentation mass was found in the first three days in ASA, this is possibly due to a greater amount of cocoa bean in mucilage that was used to ferment which influenced maintaining the temperature resulting from the exothermic reaction during the anaerobic phase in the fermentation of the cocoa mass [38]. Situation that has been evidenced in studies where the behavior of temperature is analyzed in different places simultaneously in the departments of Huila, Santander and Antioquia, Colombia which were different between sites [16]. On the other hand, from the third day of fermentation the mass recorded an increase in pH, a situation attributed to the decrease by leaching of citric acid contained in the pulp [40] and the decrease of volatile organic acids leading to an increase in pH in the fermentation mass [41] as to the metabolization of citric acid by the yeasts [42]. This decrease in pH within the first few days at all sites is attributed mainly to the diffusion of organic acids in the bean produced by lactic acid bacteria [43] and together with the high temperatures, which lead to embryo death, triggering a series of biochemical changes that greatly impact the development of bean flavor and color. Also, this behavior was considered [44] to be because of turning the cocoa mass after 48 hours, because it favors aeration and the growth of acid-acetic bacteria [45]. Finally, the pH of the cotyledon at the end of fermentation in the three sites was very similar ASA (4.33), CMI (4.59) and COC (4.63), these results are very important since some studies mention that pH above 5.5 at the end of fermentation indicates lower quality fermented beans [46], however, Calvo et al. (2021) [16] mentions that pH ranges in cotyledon between 4.8 and 5.2 indicates a good fermentation process. In this study, fermentation at CMI and COC were the closest to the conditions. Recent studies [47, 48], in addition to showing the decrease in grain pH during the fermentation process, showed changes in this variable between locations. For example, the pH of the cotyledon decreased consistently throughout the fermentation, reaching final values of 5.1 ± 0.4 in location A and 5.6 ± 0.4 in location F, with no significant difference in the variations after 72 h of fermentation. 4.2. Physical properties of dry cocoa beans Variation was found in the number of fermented grains according to the category. For example, the increase in the percentage of fermented and completely fermented beans (brown) are mainly associated with anthocyanin degradation [49] because of gradually increasing temperature in the fermentation box [50]. Purple cocoa beans are characteristic of an inadequately fermented bean and brown beans are typical of optimal fermentation [51], a situation that was present in the CMI and COC sites. The percentage of partially fermented beans presented significant differences between sites, with higher values in COC. The presence of partially fermented beans is produced by some factors such as: reduction in the amount of turning in the cocoa mass during fermentation or due to the location of the beans in the fermentation box [52]. For this reason, several studies mention the importance and impact of turning on the physical characteristics of cocoa beans [53]. The effect of the site on the quality of fermentation has been documented by different studies [47, 48, 54, 55] which mention an impact on the number of beans with complete fermentation, for example Murcia-Artunduaga et al. [55] they mention that the municipality of Tarqui (Huila, Colombia) presented the best results of physical characteristics of the cocoa beans, while Oporapa (Huila, Colombia) presented a high tendency to have “pasilla” beans. In the different locations (Manabí and Los Ríos provinces in Ecuador [48]) in which the incidence of fermentation on the physical quality of cocoa beans was evaluated, a significant variation was found in relation to a greater proportion of well-fermented beans (24.7 ± 13.4%) and slightly fermented grains (58.3 ± 6.7%) between the sites, the above product of temperature variations. 4.3. Component bromatological An increase in fat content was evidenced with increasing days of fermentation at all sites (P<0.05), the data obtained agree with the study of Millena et al. [56] who found an increase in fat content from the fifth day of fermentation. The increase in fat contents during fermentation days can be attributed to the interaction between acidity and temperature during fermentation, leading to the alteration of lipid bonds [57]. On the other hand, the values in ash percentage were below 3%, values that coincide with those reported by Calvo et al. [16], likewise, this decrease in ash content during fermentation was reported by Afoakwa et al. [58]. This behavior is explained by the loss of water-soluble minerals drained during the fermentation process [59]. Moisture contents varied significantly in all sites, being ASA with the highest values (3.98%) with respect to CMI and COC, it is believed that this behavior is due to the geographical location and precipitation of the site, because ASA is located in the Andean-Amazonian transition with higher precipitation than the other sites included in the study; according to Bomdzele and Molua [60] high precipitation can slow grain drying and increase moisture content. The decrease in grain moisture content during fermentation at the COC site is explained by the increase in temperature allowing moisture to diffuse outward and by the effect of turning the fermentable mass [61]. The increase in acidity percentage is associated with the increase of lactic acid and acetic acid during the fermentation days [16]. These acids enter the cotyledon generating biochemical reactions and causing embryo death [62]. 4.4. Polyphenolic compounds Total polyphenol content was higher at the COC site (507.05 mg GAE/g cocoa) compared to the other sites (< 360 mg GAE/g cocoa), the highest COC contents are mainly attributed to beans from hybrid cocoa plantations found in the study area. According to Jonfia-Essien et al. [63] hybrid cocoa beans have higher polyphenol levels and antioxidant activity than clones, due to their genetic variability. Regarding total polyphenol values, he results of the present study were much higher than those reported by Ramón et al. (2022) [4] in fermented and roasted beans (210.2 mg GAE/g cocoa), with the difference that in our study, the cocoa beans evaluated were not roasted, only fermented and dried. According to Urbańska et al. [64] the polyphenol content may decrease due to the oxidation of these compounds by the effect of high temperatures in the roasting process and interaction with Maillard reaction products [65, 66]. In addition to the above, a decrease in total polyphenol content was observed at all sites with increasing fermentation time, which is attributed to the effect of increasing temperature in the fermentation box [16], although this decrease has also been reported due to the influence of polyphenol oxidase enzyme activity during fermentation and drying [67]. In general terms, we posit that grains that have undergone adequate fermentation have lower polyphenol contents. This is desirable to achieve pleasant sensory profiles, because higher polyphenols content is associated with astringency [68], while anthocyanins are linked to the violet color of unfermented grains [69]. Both aspects are considered undesirable in chocolate production [70, 71]. However, their decrease has a proportional effect on antioxidant activity, a desirable feature from the functional point of view of chocolate to decrease the risk of oxidative stress-mediated diseases [72]. Total flavonoids content (TF) followed a similar behavior that TPC, with significant differences between sites (P<0.05) and differences between fermentation times for ASA. The TF was higher in COC (309.1 mg catechin/g cocoa) with respect to CMI (215.6 mg catechin/g cocoa) and ASA (185.7 mg catechin/g cocoa). Our results showed changes in TF contents because of fermentation days, this explains that flavonoids also decreased during fermentation as did polyphenols, corroborating what was found by Hu et al. [73] in their research on chemical properties in roasted and unroasted cocoa beans, in which they mention that flavonoids are more sensitive to high temperatures than polyphenols. The values found in TF in our study were higher than those reported by da Silva Oliveira et al. [74] and like those found by Cuéllar-Álvarez et al. [75] in Copoazú (Theobroma grandiflorum) beans with different fermentation days. The literature primarily reports the total flavonoid content extracted from fermented and roasted grains; hence, our values are higher than those reported by many authors, because in our study the grains used for physical and chemical properties analysis did not go through the roasting process. Zaman et al. [76] mentions that roasting causes the contents of flavonoids, flavanols, flavones and anthocyanins to decrease with increasing time and temperature during roasting. While flavonoids are important bioactive compounds that have been associated with their anti-inflammatory properties, they have a negative effect on the sensory attributes of the bean; however, chocolate manufacturers today pay great attention to declaring their products as functional foods [77]. This opens the door to discussions regarding the manufacture of chocolates with fermented/roasted beans, fermented/unroasted beans or fermented/roasted beans enriched with encapsulated polyphenols. 4.5. Sensory profile In general, complementary attributes such as nutty, spicy, woody, floral and fresh fruit were present in all sites, which are categorized as fine cocoa aroma and flavor [78]. The sensory evaluation revealed that, in all sites and days of fermentation, cocoa notes obtained intensity scores higher than 6, however, our results contrast with those reported by Calvo et al. [16] who found values below 5 in cocoa samples from three different regions in Colombia. Also, Romero and Pabón [79] mentions that beans with high cocoa flavor intensity indicate a successful processing (fermentation and drying), although the main attributes are mostly influenced by genotype [80]. In fermentation and drying, pyrazine contents increase because they are compounds of thermal origin and increase with increasing temperature [81], some studies mention that pyrazines are related to cocoa flavor [8]. It was evidenced that COC presented higher acidity (5.5) with respect to ASA (3.7) and CMI (3.3) on the fourth day of fermentation, however, on day 7 acidity decreased in COC (4.5), but increased in CMI (5.0), this decrease in acidity is mainly due to acetic acid volatilization during fermentation and drying because of temperature [82]. The increase in acidity during fermentation at CMI was mainly attributed to the presence of volatile acids, generated as a product of over fermentation and poor drying [58, 83]. Acidity is an important attribute, as high levels of acidity in cocoa samples can be detrimental to bean quality [84]. Astringency and bitterness increased with fermentation days in ASA and COC, this behavior was associated with the effect of a possible over fermentation, we also believe that the increase in astringency is due to a deficiency in the turnovers of the fermenting mass [85, 86] behavior that contrasts with a good fermentation [87] while astringency and bitterness decreased with the course of fermentation at CMI, this decrease is characteristic of a good fermentation process [88]. This flavor is acquired mainly in the roasting process, because it has a direct effect on the development of flavors in the cocoa beans through various reactions, including Maillard reactions, which contribute to the presence of nutty flavor and aroma [89]. Rodriguez-Campos et al. [84] concluded that 6 days of fermentation are sufficient to produce volatile compounds with desirable flavor notes in cocoa, conclusions that contrast with the results of this study, given that the profiles obtained here had flavor notes characteristic of over fermentation in ASA, although favorable notes in COC and CMI with 7 days of fermentation. In addition, it is known that cocoa flavor and complementary attributes are influenced by the genetic potential of cocoa, but also by postharvest practices (fermentation and drying) [90]. 4.1. Temperature and pH of fermented cocoa beans An increase in temperature was found in the fermentation mass due to the activity carried out by the yeasts involved in the metabolization of sugars in the cocoa pulp (exothermic process), to produce ethanol and other metabolities (other alcohols, esters, aldehydes and ketones), a process that releases heat and increases the fermentation temperature [38]. It was observed that from the fifth day of fermentation the temperature did not vary, reaching a temperature between 40–50°C, a desirable range for good fermentation [39]. Although the environmental temperature was not a condition that varied, a significant increase in the temperature of the cocoa fermentation mass was found in the first three days in ASA, this is possibly due to a greater amount of cocoa bean in mucilage that was used to ferment which influenced maintaining the temperature resulting from the exothermic reaction during the anaerobic phase in the fermentation of the cocoa mass [38]. Situation that has been evidenced in studies where the behavior of temperature is analyzed in different places simultaneously in the departments of Huila, Santander and Antioquia, Colombia which were different between sites [16]. On the other hand, from the third day of fermentation the mass recorded an increase in pH, a situation attributed to the decrease by leaching of citric acid contained in the pulp [40] and the decrease of volatile organic acids leading to an increase in pH in the fermentation mass [41] as to the metabolization of citric acid by the yeasts [42]. This decrease in pH within the first few days at all sites is attributed mainly to the diffusion of organic acids in the bean produced by lactic acid bacteria [43] and together with the high temperatures, which lead to embryo death, triggering a series of biochemical changes that greatly impact the development of bean flavor and color. Also, this behavior was considered [44] to be because of turning the cocoa mass after 48 hours, because it favors aeration and the growth of acid-acetic bacteria [45]. Finally, the pH of the cotyledon at the end of fermentation in the three sites was very similar ASA (4.33), CMI (4.59) and COC (4.63), these results are very important since some studies mention that pH above 5.5 at the end of fermentation indicates lower quality fermented beans [46], however, Calvo et al. (2021) [16] mentions that pH ranges in cotyledon between 4.8 and 5.2 indicates a good fermentation process. In this study, fermentation at CMI and COC were the closest to the conditions. Recent studies [47, 48], in addition to showing the decrease in grain pH during the fermentation process, showed changes in this variable between locations. For example, the pH of the cotyledon decreased consistently throughout the fermentation, reaching final values of 5.1 ± 0.4 in location A and 5.6 ± 0.4 in location F, with no significant difference in the variations after 72 h of fermentation. 4.2. Physical properties of dry cocoa beans Variation was found in the number of fermented grains according to the category. For example, the increase in the percentage of fermented and completely fermented beans (brown) are mainly associated with anthocyanin degradation [49] because of gradually increasing temperature in the fermentation box [50]. Purple cocoa beans are characteristic of an inadequately fermented bean and brown beans are typical of optimal fermentation [51], a situation that was present in the CMI and COC sites. The percentage of partially fermented beans presented significant differences between sites, with higher values in COC. The presence of partially fermented beans is produced by some factors such as: reduction in the amount of turning in the cocoa mass during fermentation or due to the location of the beans in the fermentation box [52]. For this reason, several studies mention the importance and impact of turning on the physical characteristics of cocoa beans [53]. The effect of the site on the quality of fermentation has been documented by different studies [47, 48, 54, 55] which mention an impact on the number of beans with complete fermentation, for example Murcia-Artunduaga et al. [55] they mention that the municipality of Tarqui (Huila, Colombia) presented the best results of physical characteristics of the cocoa beans, while Oporapa (Huila, Colombia) presented a high tendency to have “pasilla” beans. In the different locations (Manabí and Los Ríos provinces in Ecuador [48]) in which the incidence of fermentation on the physical quality of cocoa beans was evaluated, a significant variation was found in relation to a greater proportion of well-fermented beans (24.7 ± 13.4%) and slightly fermented grains (58.3 ± 6.7%) between the sites, the above product of temperature variations. 4.3. Component bromatological An increase in fat content was evidenced with increasing days of fermentation at all sites (P<0.05), the data obtained agree with the study of Millena et al. [56] who found an increase in fat content from the fifth day of fermentation. The increase in fat contents during fermentation days can be attributed to the interaction between acidity and temperature during fermentation, leading to the alteration of lipid bonds [57]. On the other hand, the values in ash percentage were below 3%, values that coincide with those reported by Calvo et al. [16], likewise, this decrease in ash content during fermentation was reported by Afoakwa et al. [58]. This behavior is explained by the loss of water-soluble minerals drained during the fermentation process [59]. Moisture contents varied significantly in all sites, being ASA with the highest values (3.98%) with respect to CMI and COC, it is believed that this behavior is due to the geographical location and precipitation of the site, because ASA is located in the Andean-Amazonian transition with higher precipitation than the other sites included in the study; according to Bomdzele and Molua [60] high precipitation can slow grain drying and increase moisture content. The decrease in grain moisture content during fermentation at the COC site is explained by the increase in temperature allowing moisture to diffuse outward and by the effect of turning the fermentable mass [61]. The increase in acidity percentage is associated with the increase of lactic acid and acetic acid during the fermentation days [16]. These acids enter the cotyledon generating biochemical reactions and causing embryo death [62]. 4.4. Polyphenolic compounds Total polyphenol content was higher at the COC site (507.05 mg GAE/g cocoa) compared to the other sites (< 360 mg GAE/g cocoa), the highest COC contents are mainly attributed to beans from hybrid cocoa plantations found in the study area. According to Jonfia-Essien et al. [63] hybrid cocoa beans have higher polyphenol levels and antioxidant activity than clones, due to their genetic variability. Regarding total polyphenol values, he results of the present study were much higher than those reported by Ramón et al. (2022) [4] in fermented and roasted beans (210.2 mg GAE/g cocoa), with the difference that in our study, the cocoa beans evaluated were not roasted, only fermented and dried. According to Urbańska et al. [64] the polyphenol content may decrease due to the oxidation of these compounds by the effect of high temperatures in the roasting process and interaction with Maillard reaction products [65, 66]. In addition to the above, a decrease in total polyphenol content was observed at all sites with increasing fermentation time, which is attributed to the effect of increasing temperature in the fermentation box [16], although this decrease has also been reported due to the influence of polyphenol oxidase enzyme activity during fermentation and drying [67]. In general terms, we posit that grains that have undergone adequate fermentation have lower polyphenol contents. This is desirable to achieve pleasant sensory profiles, because higher polyphenols content is associated with astringency [68], while anthocyanins are linked to the violet color of unfermented grains [69]. Both aspects are considered undesirable in chocolate production [70, 71]. However, their decrease has a proportional effect on antioxidant activity, a desirable feature from the functional point of view of chocolate to decrease the risk of oxidative stress-mediated diseases [72]. Total flavonoids content (TF) followed a similar behavior that TPC, with significant differences between sites (P<0.05) and differences between fermentation times for ASA. The TF was higher in COC (309.1 mg catechin/g cocoa) with respect to CMI (215.6 mg catechin/g cocoa) and ASA (185.7 mg catechin/g cocoa). Our results showed changes in TF contents because of fermentation days, this explains that flavonoids also decreased during fermentation as did polyphenols, corroborating what was found by Hu et al. [73] in their research on chemical properties in roasted and unroasted cocoa beans, in which they mention that flavonoids are more sensitive to high temperatures than polyphenols. The values found in TF in our study were higher than those reported by da Silva Oliveira et al. [74] and like those found by Cuéllar-Álvarez et al. [75] in Copoazú (Theobroma grandiflorum) beans with different fermentation days. The literature primarily reports the total flavonoid content extracted from fermented and roasted grains; hence, our values are higher than those reported by many authors, because in our study the grains used for physical and chemical properties analysis did not go through the roasting process. Zaman et al. [76] mentions that roasting causes the contents of flavonoids, flavanols, flavones and anthocyanins to decrease with increasing time and temperature during roasting. While flavonoids are important bioactive compounds that have been associated with their anti-inflammatory properties, they have a negative effect on the sensory attributes of the bean; however, chocolate manufacturers today pay great attention to declaring their products as functional foods [77]. This opens the door to discussions regarding the manufacture of chocolates with fermented/roasted beans, fermented/unroasted beans or fermented/roasted beans enriched with encapsulated polyphenols. 4.5. Sensory profile In general, complementary attributes such as nutty, spicy, woody, floral and fresh fruit were present in all sites, which are categorized as fine cocoa aroma and flavor [78]. The sensory evaluation revealed that, in all sites and days of fermentation, cocoa notes obtained intensity scores higher than 6, however, our results contrast with those reported by Calvo et al. [16] who found values below 5 in cocoa samples from three different regions in Colombia. Also, Romero and Pabón [79] mentions that beans with high cocoa flavor intensity indicate a successful processing (fermentation and drying), although the main attributes are mostly influenced by genotype [80]. In fermentation and drying, pyrazine contents increase because they are compounds of thermal origin and increase with increasing temperature [81], some studies mention that pyrazines are related to cocoa flavor [8]. It was evidenced that COC presented higher acidity (5.5) with respect to ASA (3.7) and CMI (3.3) on the fourth day of fermentation, however, on day 7 acidity decreased in COC (4.5), but increased in CMI (5.0), this decrease in acidity is mainly due to acetic acid volatilization during fermentation and drying because of temperature [82]. The increase in acidity during fermentation at CMI was mainly attributed to the presence of volatile acids, generated as a product of over fermentation and poor drying [58, 83]. Acidity is an important attribute, as high levels of acidity in cocoa samples can be detrimental to bean quality [84]. Astringency and bitterness increased with fermentation days in ASA and COC, this behavior was associated with the effect of a possible over fermentation, we also believe that the increase in astringency is due to a deficiency in the turnovers of the fermenting mass [85, 86] behavior that contrasts with a good fermentation [87] while astringency and bitterness decreased with the course of fermentation at CMI, this decrease is characteristic of a good fermentation process [88]. This flavor is acquired mainly in the roasting process, because it has a direct effect on the development of flavors in the cocoa beans through various reactions, including Maillard reactions, which contribute to the presence of nutty flavor and aroma [89]. Rodriguez-Campos et al. [84] concluded that 6 days of fermentation are sufficient to produce volatile compounds with desirable flavor notes in cocoa, conclusions that contrast with the results of this study, given that the profiles obtained here had flavor notes characteristic of over fermentation in ASA, although favorable notes in COC and CMI with 7 days of fermentation. In addition, it is known that cocoa flavor and complementary attributes are influenced by the genetic potential of cocoa, but also by postharvest practices (fermentation and drying) [90]. 5. Conclusions The management of the fermentation process has a significant impact on the different characteristics (biochemical, physical and sensory) of the cocoa beans. A similar trend of fermentation mass variables was found at all sites where cotyledon pH decreased during the fermentation process and fat and moisture content varied with fermentation time. At the site level, COC was different from the other sites specifically for total polyphenol content (TPC), total flavonoids (TF), DPPH and FRAP. The TPC was higher in the COC site (507 mg GAE/g Cocoa) with respect to the other sites (< 360 mg GAE/g Cocoa). The TF content followed a similar behavior to TPC, with significant differences between sites and differences between fermentation times for ASA. The TF was higher in COC (309.1 mg catechin/g) with respect to CMI (215.6 mg catechin/g) and ASA (185.7 mg catechin/g). Values in DPPH ranged from 5869.3 to 7781.8 μmol Trolox/g cocoa and for the FRAP assay ranged from 369.8 to 606.7 mg AA/g cocoa among sites. Complementary attributes such as nutty, spicy, woody, woody, floral and fresh fruit were observed at all sites. Astringency and bitterness increased with days of fermentation at ASA and COC, while astringency and bitterness decreased with the course of fermentation at CMI. Supporting information S1 File. Variables physicochemical and sensory attributes of cocoa beans. https://doi.org/10.1371/journal.pone.0306680.s001 (XLSX) Acknowledgments We also thank the following collaborators: Asociación Orgánica Agrícola de San José del Fragua (ASOACASAN), Comité de Cultivadores de Cacao en Sistemas Agroforestales del municipio de San Vicente del Caguán (COMICACAO) and Comité de Cacaoteros de los municipios del Paujil y el Doncello (COMCAP) for providing the benefit facilities to carry out this study. Finally, we would like to thank Corporacion Econexus Colombia (INSITU) for their support in the tasting of the samples.
TI - Fermentation and its effect on the physicochemical and sensory attributes of cocoa beans in the Colombian Amazon
JF - PLoS ONE
DO - 10.1371/journal.pone.0306680
DA - 2024-10-03
UR - https://www.deepdyve.com/lp/public-library-of-science-plos-journal/fermentation-and-its-effect-on-the-physicochemical-and-sensory-XLIhIyX2RQ
SP - e0306680
VL - 19
IS - 10
DP - DeepDyve
ER -