In total, 40 samples (fresh and/or boiled depending commonly consumed method of each plant) of 29 indigenous plants were collected three times from two conservation areas in Kanchanaburi province (representative of the western region) and Amnatchareon province (representative of the northeastern region) of Thailand. They were identified and collected with only the edible part being selected for analysis. Each edible plant part was prepared for consumption using local preparation methods, i.e., fresh, blanching, boiling. The samples were categorized into six groups based upon the plant part that is commonly consumed, namely, young leaves, flower, fruit, pod, tuber, and young stem (Table 1). The samples were transported under cooled conditions in an ice-box and sent to the analytical ISO 17025 laboratory at the Institute of Nutrition, Mahidol University (INMU). Due to growing conditions, the number of samples for each plant over the four-year survey and collection period (2013-2015 and 2019) was limited. However, each sample was analyzed in triplicate.

Each indigenous plant sample was cleaned twice with tap water and once with deionized water. In each source of sample, the edible part of each sample was collected, weighed, and divided into two portions. The first portion was prepared as a homogeneous sample for moisture analysis only. The other portion entailed cooked samples that were blanched or boiled depending on the commonly used preparation method. Thereafter, both raw and cooked samples were homogenized using a food processor (Mara®, Thailand) and dried using a freeze-dried system (Heto®, Power Dry PL9000 Freeze Dryer, Thermo Fisher Scientific). They were then weighed, ground, vacuum-packed in laminated aluminum foil bags, and stored at 20˚C until analysis.

Moisture content was analyzed by drying according to the Official Analytical Chemists, Method 927.05 [16] [17]. The homogenized sample was weighed into an aluminum container and dried in a hot air oven at 100˚C ± 1˚C until constant weight. The results were expressed as gram per 100 g of fresh weight (FW).

Organic acids were extracted from 1 g of finely ground-dried sample with 50 mL 2M HCl at temperature 80˚C for 15 minutes. They were centrifuged at 3000 rpm for 15 minutes. Supernatant was passed through a filter of 0.45 µm cellulose acetate membrane. Organic acid contents were determined by the HPLC method. A 20 µL of filtrated sample was injected into the HPLC system (Waters chromatography system, Milford, USA) with UV detector (Waters® 486), set at 210 nm. The separation of each organic acid was carried out on a 300 × 7.8 mm Biorad Aminex ion exclusion column (HPX-87H), using an isocratic elution at 0.5 mL/min with 0.0125M sulphuric acid as a mobile phase. Figure 1 shows a chromatogram corresponding to a six-component mixture of standards and an indigenous sample. One can observe the good resolution and separation of the identified organic acids in a real sample. All samples were analyzed in triplicate. Organic acid contents were calculated against their standard curves (calibration of mix cocktail standards) [18] [19].

Ascorbic acid or vitamin C was determined using the HPLC method [20]. In brief, each sample was extracted in 3% metaphosphoric acid and the homogenate was filtered. Ascorbic acid was separated by reversed-phase HPLC with UV detection at 248 nm (Agilent® 1100 series, USA) and quantified against external standards.

Total acidity and pH were determined by the potentiometric method using a

glass electrode according to the AOAC Method 942.15B [16] [17]. Each sample was dissolved with deionized water, stirred, measured for pH value, and then titrated to the range of pH 8.10 ± 0.2. Total acidity was calculated, as percentage, based on each type of dominant organic acid in the sample.

For precision in the organic acid analysis, Cha-moung leaves were dried at 50˚C ± 1˚C until dryness then ground into fine particles. They were then used as in-house quality control (QC) samples for representative organic acids, since these leaves are known to contain a significant amount of oxalic acid and several other organic acids. The assigned values of oxalic were developed from duplicate analysis of 10 single samples of the in-house QC samples on 10 different days. The QC sample was then analyzed along with the unknown samples in each run of the organic acid measurement over time. Due to a lack of an available certified reference material, spike standard organic acids in a QC sample were performed (n = 10) to demonstrate accuracy. Limit of detection (LOD) and limit of quantitation (LOQ) were studied using spiked lowest concentration of mixed standard into the QC sample and then calculated as 3 and 10 times of standard deviation from 10 times analysis, respectively.

The Cha-moung (Garcinia cowa Roxb.) dried leaf, as the in-house QC sample, contained a high amount of oxalic acid (1120 ± 103 mg/100g of fresh weight, FW). The organic acid content in each analytical set stayed within the acceptable range (mean ± 2 standard deviation of the assigned value). Recoveries of oxalic acid added (10 mg/g of sample) prior to sample extraction were performed in both the in-house QC sample and the dried food samples. Percentages of recovery of oxalic acids ranged from 90.6% to 99.4% (Table 2) which stayed in the acceptable recovery range (90% - 107%) [17]. LOD of oxalic, citric, malic, succinic, formic, and acetic acids were 0.9, 2.2, 2.0, 0.2, 0.1, and 3.0 mg/100g, respectively. LOQ of oxalic, citric, malic, succinic, formic, and acetic acids were 3.0, 7.3, 12.0, 0.5, 0.5, and 9.0 mg/100g respectively.

The major component in the studied plants was water, ranging from 69.9 g/100g to 97.4 g/100g FW (Tables 3-6). In general, organic acids can be divided into two groups. The first group entails organic acids that have been reported to have negative health effects, such as oxalic acid. The second group comprises those that provide health benefits, such as citric, malic, succinic, and acetic acids. Results and discussion for each organic acid are presented in Tables 3-6.

¹Three individual samples were analyzed in duplicate and presented as mean + standard deviation; ²ND = Not detectable (less than LOD).

¹Three individual samples were analyzed in triplicate and presented as mean ± standard deviation; ²ND = Not detectable (less than LOD).

¹Three individual samples were analyzed in triplicate and presented as mean ± standard deviation; ²ND = Not detectable (less than LOD).

¹Three individual samples were analyzed in triplicate and presented as mean ± standard deviation; ²ND = Not detectable (less than LOD).

The highest ascorbic acid content was found in fruits of Ma-karm-pom (Phyllanthus emblica L.) (648 ± 189 mg/100g FW), which agrees well with a previous study [29] (575 ± 452 mg/100g FW). A high level of ascorbic acid was found in leaves of Som-mong (Garcinia cowa Roxb.ex DC.), both fresh and blanched (340 ± 59 and 356 ± 78 mg/100g FW respectively), and leaves of fresh Som-lom (Aganonerion polymorphum Pierre ex Spire) (329 ± 58 mg/100g FW). Ascorbic acid in fresh fruits of Ma-kok-pa (Spondias pinnata (L.f.) Kurz) and Ma-mao (Antidesma velutinosum Blume) were 37 ± 10 and 2 ± 1 mg/100g FW, respectively, which were similar to the results of a previous study [29]. Young leaves in the family of Guttiferae, which has a strong sour taste, such as Tew-deang (Cratoxylum formosum subsp. Pruniflorum [Kurz] Gogel, Tew-kaw (Cratoxylum formosum [Jack] Dyer), and Tew-mon (Cratoxylum cochinchinense [Lour.] Blume), also contained high amounts of ascorbic acid (142 ± 35, 136 ± 34, and 102 ± 36 mg/100 g FW, respectively).

High total acidity and low pH (a measure of the power of an acid in sample) measurements are used in sour taste indicators. Som-mong (Garcinia cowa Roxb.ex DC.), both blanched and fresh leaves, provided the highest total acidity content (16.27% ± 0.03% and 16.07% ± 0.10% FW, respectively) but they had very low pH (3.81 ± 0.10 and 3.80 ± 0.12, respectively). Fruits of Ma-mao (Antidesma velutinosum Blume), Ma-karm-pom (Phyllanthus emblica L.), and Ma-kok-pa (Spondias pinnata (L.f.) Kurz) had high total acidity (13.74% ± 0.30%, 13.45% ± 0.14% and 13.16% ± 0.30% FW, respectively); whereas they had low pH (3.75 ± 0.33, 3.48 ± 0.24 and 3.24 ± 0.07, respectively). Leaves of Tew-deang (Cratoxylum formosum subsp. Pruniflorum [Kurz] Gogel) and fresh flowers of Kra-don (Careya sphaerica Roxb) also contained moderate total acidity (6.09% ± 0.71% and 4.29% ± 0.31% FW, respectively) and slightly acid with a low pH (5.50 ± 0.13 and 5.49 ± 0.30). The other plant parts had total acidity less than 4.0% and pH more than 5.3.

Organic acids have been used as a component in calculating total energy apart from protein, fat, and carbohydrate, by providing 3 kcal/g [15]. Only some fermented foods and foods containing high amounts of total organic acids could affect the energy levels of those foods. Limited information exists in the different food composition databases around the world concerning energy from organic acids as part of total energy. From the studied plants, seven samples out of forty plants, namely fruits of Ma-karm-pom (Phyllanthus emblica L.), and Ma-kok-pa (Spondias pinnata (L.f.) Kurz), fresh young leaves of Tew-deang (Cratoxylum formosum subsp. Pruniflorum [Kurz] Gogel), fresh young fruits of Ma-mao (Antidesma velutinosum Blume), fresh flower of Kra-don (Careya sphaerica Roxb), fresh ripe fruits of Ma-mao (Antidesma velutinosum Blume), and fresh young leaves of Tew-kaw (Cratoxylum formosum [Jack] Dyer) had high energy contributions in descending order of 36, 22, 16, 16, 13, 10, and 10 kcal/100g FW, respectively. Organic acids in other parts of the above plants, and other plants, contributed energy of less than 10 kcal/100g fresh weight. Consequently, energy from the sum of organic acids could be included for only some plants, especially for those with higher amounts.

According to traditional belief, sour-tasting food is probably high in vitamin C or ascorbic acid. However, this belief may not always be true, because organic acids in a plant may also play an important role in causing a sour taste. Pearson’s coefficient correlations (r) were used here due to the common and complex interactions between all parameters related with the sour taste (Table 7). In general, it is well-known that lower pH and high total acidity could indicate a sour taste in food. Results from this study demonstrated that the sum of organic acids in the studied plants presented significant and inverse correlations with pH (r = −0.680) and the total acidity positive correlations (r = 0.672) but were not significantly correlated with ascorbic acid (r = 0.536). For all individual organic acids, only oxalic acid had a significantly positive correlation with ascorbic acid (r = 0.670), whereas a study by Suarez et al. [30] reported a negative correlation (r = −0.156). The pH is strongly inversely correlated with total acidity (r = −0.913) which agrees with the trend in the Suarez et al. [30] study in different cultivars of tomatoes (r = −0.384). The pH also had a notable negative correlation with the sum of organic, citric, and formic acids by order of r = −0.680, −0.660, and −0.595, respectively, but was not correlated with ascorbic acid (r = −0.494). Total

^a Pearson’s coefficient correlation; ^b Correlation is significant at the 0.01 level (2-tailed).

acidity presented a significantly positive correlation with the sum of organic, citric, formic, and ascorbic acids (r = 0.672, 0.648, 0.539, and 0.570, respectively). In terms of the sour taste in plants, as indicated by low pH and high total acidity, this finding suggests that this taste could come from a sum of organic, citric, and formic acids, but not from ascorbic acid.