Dynamic Changes in Phenolics and Antioxidant Capacity during Pecan (Carya illinoinensis) Kernel Ripening and Its Phenolics Profiles

Pecan (Carya illinoinensis) kernels have a high phenolics content and a high antioxidant capacity compared to other nuts—traits that have attracted great interest of late. Changes in the total phenolic content (TPC), condensed tannins (CT), total flavonoid content (TFC), five individual phenolics, and antioxidant capacity of five pecan cultivars were investigated during the process of kernel ripening. Ultra-performance liquid chromatography coupled with quadruple time-of-flight mass (UPLC-Q/TOF-MS) was also used to analyze the phenolics profiles in mixed pecan kernels. TPC, CT, TFC, individual phenolics, and antioxidant capacity were changed in similar patterns, with values highest at the water or milk stages, lowest at milk or dough stages, and slightly varied at kernel stages. Forty phenolics were tentatively identified in pecan kernels, of which two were first reported in the genus Carya, six were first reported in Carya illinoinensis, and one was first reported in its kernel. The findings on these new phenolic compounds provide proof of the high antioxidant capacity of pecan kernels.


Introduction
The pecan [Carya illinoinensis (Wangenh.) K. Koch], originally from North America, is one of the most important nut crops worldwide. It was introduced into China over 100 years ago and has attracted great attention this decade as a nut full of nutritional value [1,2]. Recently, more and more attention has been given to the high antioxidant capacity of pecan kernels. Wu et al. [3] screened over 100 commonly consumed foods in the USA and found that pecan kernels have the highest TPC and antioxidant capacity among the nut group, higher than many other fruits and vegetables rich in phenolics. Another study on proanthocyanidins (which also belong to the phenolics family) content in 98 common foods showed that pecans had the second highest content [4]. These results highlight the possibility of considering pecans as a nutritional food that has high phenolic content and good antioxidant capacity.
The strong antioxidant capacity of pecan kernel mainly comes from the phenolic compounds [5]. Phenolics found in pecan kernels were mainly flavan-3-ols [6], anthocyanidins [7], proanthocyanidins [4], phenolic acid [8], and their sugar-containing glycosides [9] or polymeric tannins with degrees up to 10 [4]. Phenolics were reported to have strong antioxidant activities, Figure 1. Changes in total phenolic content (a), condensed tannins (b), and total flavonoid content (c) during the ripening of pecan kernels. Each value is a mean ± standard deviation of triplicate analysis results of different samples. Means with different letters in each line were significantly different according to the multiple-range Tukey's test at p < 0.05.

Antioxidant Capacity
Antioxidant capacity was valued using two in vitro assays, including the 2, 2-diphenyl-1-picrylhydrazyl-based assay (DPPH) and the 2, 2′-azino-bis (3-ethylbenzo-thiazoline-6-sulphonic acid) diammonium salt-based assay (ABTS) (Figure 2). The highest antioxidant capacity levels were observed at the water stages in all five cultivars of both assays, and declined significantly (p < 0.05). When mature (Pawnee, Wichita, and Jinhua at S8; Stuart and Shaoxing at S9), antioxidant capacities of pecan cultivars were in the descending order Pawnee ≈ Jinhua > Stuart > Wichita > Shaoxing in DPPH assay and Pawnee > Jinhua ≈ Stuart > Wichita > Shaoxing in ABTS assay. The antioxidant capacity was affected significantly by cultivars. Among the five cultivars, Pawnee had the highest DPPH of 461.80 μmol TE g −1 (S2) and Wichita Figure 1. Changes in total phenolic content (a); condensed tannins (b); and total flavonoid content (c) during the ripening of pecan kernels. Each value is a mean ± standard deviation of triplicate analysis results of different samples. Means with different letters in each line were significantly different according to the multiple-range Tukey's test at p < 0.05.
had the highest ABTS of 201.49 μmol TE g −1 (S2), while Stuart had the lowest DPPH of 37.24 μmol TE g −1 and ABTS of 6.74 μmol TE g −1 (S7). Over-ripeness increased the antioxidant capacity in all cultivars (except Jinhua). Although the high antioxidant capacity of pecan kernel was reported before [3,4,10], no report of their changes was available during kernel ripening. As shown in Figure 2, the antioxidant capacity followed a similar trend to TPC, CT, and TFC during pecan nut ripening and the antioxidant capacity of immature samples was higher than that of mature samples. Since dynamic changes of pecan kernel antioxidant capacity had not been investigated before, we can only Although the high antioxidant capacity of pecan kernel was reported before [3,4,10], no report of their changes was available during kernel ripening. As shown in Figure 2, the antioxidant capacity followed a similar trend to TPC, CT, and TFC during pecan nut ripening and the antioxidant capacity of immature samples was higher than that of mature samples. Since dynamic changes of pecan kernel antioxidant capacity had not been investigated before, we can only compare our results of mature kernels with previous reports on pecans. Our results for mature kernels were similar to those reported by Villarreal-Lozoya et al. [6] and de la Rosa et al. [8].

Identification of Phenolics in Pecan Kernels Using UPLC-Q/TOF-MS
UPLC-Q/TOF-MS was used to identify the phenolics in pecan kernels. Cultivar Pawnee was chosen due to its relatively high antioxidant capacity among the five cultivars. Pecan kernels of Pawnee of 10 developmental stages were pooled and analyzed using UPLC-Q/TOF-MS. Both the positive and the negative mode were used to identify phenolic compounds, while results showed that the negative mode was more suitable for phenolics. So, subsequent analyses were carried out using negative mode.
A total of 40 phenolics were tentatively identified in pecan kernels, of which two were first reported in the genus Carya, six were first reported in species Carya illinoinensis and one was first reported in pecan kernels. The retention time (RT), molecular ion, fragment ions, molecular formula, and mass measurement errors (∆m) are shown in Table 1. More than half of the phenolics identified in pecan kernel were ellagic acid derivatives. The free ellagic acid (peak 20) was detected at RT of 6.36 min with a [M − H] − of m/z 300.9961 and confirmed by comparing with commercial standard. Ellagic acid hexoside (Peak 19) and ellagic acid pentose (peak 21) can easily be found by the characteristic fragment ion at m/z 301 along with the neutral losses of 162 amu (the loss of hexoside) and 132 amu (the loss of pentose), respectively. Peak 29 was tentatively identified as ellagic acid rhamnoside because of the same fragment ion at m/z 301 and a neutral loss of 146 amu [36]. Peak 25 was assigned as ellagic acid rutinoside for a neutral loss of 308 amu, which corresponded to the loss of a rhamnose-glucose structure. To the best of our knowledge, both ellagic acid rhamnoside and ellagic acid rutinoside were identified in the Carya genus for the first time.
Ellagic acid is formed from the condensation reaction of two gallic acids. Galloyl-linked ellagic acid derivatives can also be found in pecan kernels. Similar methyl ellagic acid derivatives can be identified with the characteristic fragment ion at m/z 315, including methyl ellagic acid (peak 30), methyl ellagic acid hexoside (peak 24), methyl ellagic acid pentose (peak 31), and methyl ellagic acid galloyl pentose (peak 36, 37 and 38). Dimethyl ellagic acid derivatives added an extra 15 amu (another methyl group) to form the main ion at m/z 328. Several peaks that contained this kind of structure were assigned, including dimethyl ellagic acid (peak 39 and 40) and dimethyl ellagic acid hexoside (peak 27).
Another ellagic-acid-related derivative family was the ellagitannins (ETs), which contain the hexahydroxydiphenoyl (HHDP) group ( More than half of the phenolics identified in pecan kernel were ellagic acid derivatives. The free ellagic acid (peak 20) was detected at RT of 6.36 min with a [M − H] − of m/z 300.9961 and confirmed by comparing with commercial standard. Ellagic acid hexoside (Peak 19) and ellagic acid pentose (peak 21) can easily be found by the characteristic fragment ion at m/z 301 along with the neutral losses of 162 amu (the loss of hexoside) and 132 amu (the loss of pentose), respectively. Peak 29 was tentatively identified as ellagic acid rhamnoside because of the same fragment ion at m/z 301 and a neutral loss of 146 amu [36]. Peak 25 was assigned as ellagic acid rutinoside for a neutral loss of 308 amu, which corresponded to the loss of a rhamnose-glucose structure. To the best of our knowledge, both ellagic acid rhamnoside and ellagic acid rutinoside were identified in the Carya genus for the first time.
Ellagic acid is formed from the condensation reaction of two gallic acids. Galloyl-linked ellagic acid derivatives can also be found in pecan kernels. Similar methyl ellagic acid derivatives can be identified with the characteristic fragment ion at m/z 315, including methyl ellagic acid (peak 30), methyl ellagic acid hexoside (peak 24), methyl ellagic acid pentose (peak 31), and methyl ellagic acid galloyl pentose (peak 36, 37 and 38). Dimethyl ellagic acid derivatives added an extra 15 amu (another methyl group) to form the main ion at m/z 328. Several peaks that contained this kind of structure were assigned, including dimethyl ellagic acid (peak 39 and 40) and dimethyl ellagic acid hexoside (peak 27).
Galloyl-HHDP-glucose (strictinin/isostrictinin) was identified at Another dominant class of compounds in pecan kernels was catechin and its derivatives. Gallic acid derivatives can also be found in pecan kernels. The free gallic acid (peak 4) can be identified at RT of 1.15 min and confirmed by comparing with a commercial standard. Several other gallic acid derivatives were also identified, including gallic acid hexoside (m/z 331, peak 2), methyl gallate (m/z 183, peak 12), and digalloyl-glucose (m/z 483, peak 14). Methyl gallate was previously found in pecan leaves [31], while this is the first time it has been reported in pecan kernels.
Two dicarboxylic acid derivatives were detected in pecans for the first time. Glansreginin A and B were identified by the characteristic [M − H] − ions at m/z 592 (peak 32) and 565 (peak 22) and fragment ions at m/z 403/343/241, which were identical to previous reports [33]. These two compounds were first isolated and elucidated in walnut kernels [38]; they have also been found in hazelnut kernels [39].

Quantification of Phenolics in Pecan Kernels Using HPLC
Although UPLC-Q/TOF-MS is very sensitive and suitable for identification of compounds, due to its relatively low peak resolution within a short procedure, HPLC was used to quantify the dynamic changes of several representative phenolics.
The contents of gallic acid, (+)-catechin, (−)-epicatechin, EGCG, and ellagic acid were analyzed in five cultivars of 10 developmental stages (Figure 4). The contents of five individual phenolics were all highest at water stages, lowest at dough stages, and varied at the kernel stages, similar to the change patterns of TPC, CT, and TFC. Similar results can be found in research on Anacardium occidentale L., while phenolic contents (like gallic acid and EGCG) decreased significantly from unripe to medium-ripe and ripe [40]. This fall may be attributable to their conversion to other downstream forms such as esterified or glycoside forms, along with the nut development. It may also attribute to the high proportion of testa at immature stages. In immature pecan kernels, the proportion of testa was high, and decreased gradually with the nut development. Senter et al. reported firstly that pecan testa might contain powerful antioxidant phenolics and act as a natural screen to prevent infection by viruses, bacteria, and molds [41]. They speculated that phenolics existed mainly in the outer testa or pellicle of the pecan kernels [42]. Research showed that contents of (+)-catechin and (−)-epicatechin in the testa of cashew nut were 20 times and five times higher than in green tea and dark chocolate, respectively [43,44]. Further experiments are needed to obtain more accurate data on phenolics in different organs of pecan kernels.
Molecules 2018, 23, x FOR PEER REVIEW 9 of 17 unripe to medium-ripe and ripe [40]. This fall may be attributable to their conversion to other downstream forms such as esterified or glycoside forms, along with the nut development. It may also attribute to the high proportion of testa at immature stages. In immature pecan kernels, the proportion of testa was high, and decreased gradually with the nut development. Senter et al. reported firstly that pecan testa might contain powerful antioxidant phenolics and act as a natural screen to prevent infection by viruses, bacteria, and molds [41]. They speculated that phenolics existed mainly in the outer testa or pellicle of the pecan kernels [42]. Research showed that contents of (+)-catechin and (−)-epicatechin in the testa of cashew nut were 20 times and five times higher than in green tea and dark chocolate, respectively [43,44]. Further experiments are needed to obtain more accurate data on phenolics in different organs of pecan kernels. In immature pecan kernels, the most abundant phenolic was (+)-catechin, followed by (−)-epicatechin, while EGCG had the lowest content. The content of (+)-catechin was much higher, e.g., over 10 times (17.7100 mg g −1 at S1) higher than ellagic acid (1.4982 mg g −1 at S1) in Shaoxing, than other phenolics during immature stages. This was also probably due to the fact that they were used as building blocks to form proanthocyanidins. Two or three catechin units linked together formed proanthocyanidins (dimers and trimers), while three or more catechin units linked together formed oligomeric proanthocyanidins [45]. Both proanthocyanidins and oligomeric proanthocyanidins are abundant in pecan kernels. Two proanthocyanidins were identified in this In immature pecan kernels, the most abundant phenolic was (+)-catechin, followed by (−)-epicatechin, while EGCG had the lowest content. The content of (+)-catechin was much higher, e.g., over 10 times (17.7100 mg g −1 at S1) higher than ellagic acid (1.4982 mg g −1 at S1) in Shaoxing, than other phenolics during immature stages. This was also probably due to the fact that they were used as building blocks to form proanthocyanidins. Two or three catechin units linked together formed proanthocyanidins (dimers and trimers), while three or more catechin units linked together formed oligomeric proanthocyanidins [45]. Both proanthocyanidins and oligomeric proanthocyanidins are abundant in pecan kernels. Two proanthocyanidins were identified in this experiment. So, catechin was very likely to be stored at a high concentration at the beginning of kernel formation for future use.
In mature kernels (Pawnee, Wichita, and Jinhua at S8; Stuart and Shaoxing at S9), the contents of ellagic acid were highest in Pawnee, Stuart, Jinhua, and Shaoxing, but in Wichita the content of (+)-catechin was the highest. This result was consistent with previous reports on pecans, in which most of them reported the most abundant free phenolic was ellagic acid [9,19], while some reported it to be (+)-catechin [46]. Phenolics are secondary metabolites, so environmental factors such as temperature and sun irradiation could affect their concentration [47]. For example, a walnut husk cultivated at higher altitude contained higher total phenolic content than one cultivated at a lower altitude [48]. The temperature and sunshine duration during pecan kernel mature were recorded; however, the effects of these environmental factors on phenolic contents await further study.

The Correlation of Antioxidant Capacity and Phenolics
A correlation was determined between antioxidant capacity and individual phenolics ( Table 2). TPC, CT, TFC, and individual phenolics were all significantly correlated with antioxidant capacity, especially of (+)-catechin, (−)-epicatechin, and TPC. (+)-Catechin had the highest correlation coefficient with antioxidant capacity, followed by (−)-epicatechin among individual phenolics. Coefficients between TPC and antioxidant capacity were higher than those of CT and TFC. This result was consistent with previous reports on mature pecans [8,19,49,50]. (+)-Catechin and (−)-epicatechin might be the main components responsible for the high antioxidant capacity of pecan kernels. It is reported that phenolics have high scavenging abilities with DPPH free radicals [51]. The value of DPPH in our experiment was higher than that of ABTS ( Figure 2); similar results can be found in previous pecan reports [8,50]. ABTS was measured at the near-infrared area of 734 nm, which cut off a lot of interference from other absorbing components such as sugars [51]. This might lead to a more precise result and might be the reason for its higher correlation coefficient with phenolics compared to DPPH (Table 2).  Phenolics identified in pecan kernels mainly included derivatives of ellagic acid, gallic acid, and catechin. Seven ETs and two proanthocyanidins were identified in this experiment. Both ETs and proanthocyanidins were reported to have a high antioxidant capacity [52]. However, due to the lack of quantification of other phenolics identified by UPLC-Q/TOF-MS, the correlations of antioxidant capacity with these phenolics were not analyzed. Further experiments will focus on that aspect to elucidate the relations between the complex phenolics and antioxidant capacity.

Pecan Samples
The experimental trees were planted in the scientific orchards of the Institute of Botany, Jiangsu Province and Chinese Academy of Sciences, under normal management of water and fertilizer. Nuts of five pecan cultivars, including Pawnee, Stuart, Wichita (introduced cultivars), Jinhua, and Shaoxing (local seedling selection cultivars), were sampled at 10 developmental stages from 85 to 175 days after full blossoming at 10-day intervals in 2016 ( Table 3). The daily mean temperatures of sampling interval were 18.27-32.09 • C and the daily mean sunshine durations of sampling interval were 1.00-10.35 h. The total sampling times of five cultivars varied due to the differences in fruit development. Cultivars Pawnee, Wichita, and Jinhua were fully mature at S8, while cultivars Stuart and Shaoxing were at S9. Fruits were sampled from as soon as the kernel appeared until it was mature. Additionally, fruits were allowed to stay on the tree for 10 days after maturity to evaluate the effect of over-ripeness. Nine healthy pecan trees (10 years old) of each cultivar were selected as sample trees. Two healthy nuts were hand-picked from the four directions, east, south, west, and north, which gave a total of eight nuts from each tree. Nuts from every three trees were combined together to form three biological replicates, which contained 24 nuts in each replicate and 72 nuts for each stage of a single cultivar. The samples were immediately placed in an ice box and transported back to the laboratory. The kernels were separated manually and stored at −70 • C for analysis. Before use, samples were powdered and homogenized.  3 Daily mean temperatures and sunshine durations of sampling interval were calculated using data of 10 days between the intervals, for example, data of 85 days was calculated using data of 74-85 days, and so on.

Sample Extraction
Pecan kernels were dried and then defatted with hexane. Phenolic compounds were extracted according to the methods of de la Rosa et al. [8] and Villarreal-Lozoya et al. [6], with slight modifications. After volatilization of the solvent residue, defatted pecan kernels (1 g) were placed in a 50-mL centrifuge tube with a screw cap; 20 mL of 80% acetone was added, and the mixture was rested overnight, ultrasonic extracted for 2 h, and centrifuged at 6000 rpm for 10 min at 4 • C (Hettich, Andreas Hettich GmbH & Co. KG, Tuttlingen, Germany). The kernel residues were ultrasonically extracted again with another 20 mL of 80% acetone, and the supernatants were combined. Solvents were removed by nitrogen blowing (Anpel nitrogen evaporator, Anpel Laboratory Technologies Inc., Shanghai, China) at 50 • C and lyophilized (Songyuan Huaxing LGJ-12, Beijing Songyuan Huaxing Technology Develop Co., Ltd, Beijing, China). Dried extracts were dissolved in methanol and stored at −20 • C until use.

Total Phenolic Content
TPCs were determined according to the method of de la Rosa et al. [8] and Robbins et al. [19]. Methanol extract (10 µL) was transferred to a 10-mL centrifuge tube and 2 mL of 7.5% (w/v) sodium carbonate and 2.5 mL of 10% (v/v) Folin-Ciocalteu regent were added and the mixture reacted in a water bath at 50 • C for 15 min in the dark. After cooling to room temperature, the absorbance was measured at 760 nm using UV spectrophotometer (Shimadzu UV-2100, Shimadzu Corporation, Kyoto, Japan). Ellagic acid was used as a standard reference, and the results were expressed as milligrams of ellagic acid equivalent (EAE) per gram of defatted kernel weight (mg EAE g −1 ). All blanks used in this article were made using pure methanol without sample extracts, and treated along with samples using the same method at the same time. All analyses in this article were performed in triplicate with three biological replicates.

Condensed Tannins
CT assays were performed using a vanillin assay [8]. The methanol extract (200 µL) was transferred to a 10-mL centrifuge tube, and then 2.5 mL vanillin regent (5 g of reagent and 1 L of 4% HCl methanol, v/v) was added and reacted at room temperature for 20 min in the dark. Absorbance was measured at 500 nm using a UV spectrophotometer. (+)-Catechin was used as a standard reference, and the results were expressed as milligrams of (+)-catechin equivalents (CE) per gram of defatted kernel weight (mg CE g −1 ).

Total Flavonoid Content
TFCs were measured according to the method of de la Rosa et al. [20] with some modifications. Briefly, methanol extract (100 µL) was transferred to a 10-mL centrifuge tube, 1 mL of 5% sodium nitrite (w/v) was added, and the mixture was incubated for 5 min in the dark. Then, 1 mL of 10% aluminum chloride (w/v) was added and incubated for 3 min. After adding 5 mL of 4% NaOH (w/v) and incubating in the dark for 30 min, absorbance was measured at 510 nm using UV spectrophotometer. (+)-Catechin was used as a standard reference, and the results were expressed as milligrams of (+)-catechin equivalents per gram of defatted kernel weight (mg CE g −1 ).

Antioxidant Capacity
Firstly, the antioxidant capacity of pecan kernels was measured using a DPPH assay according to a previous report [53], and modified as described by Prado et al. [50] and Villarreal-Lozoya et al. [6]. A DPPH radical solution was prepared by dissolving 39.43 mg DPPH in 1 L methanol, and storing at 4 • C before use. Methanol extract (10 µL) was transferred into 15-mL centrifuge tube and 4 mL of DPPH radical solution was added. The centrifuge tubes were kept in the dark for 30 min for a scavenging reaction. Then, absorbance was measured at 515 nm with a UV spectrophotometer. Methanol without sample was used as a blank and treated with the same method as the samples. Absorbance of blank was subtracted from each sample. Trolox was used as a standard reference, and the results were expressed as µmol trolox equivalents (TE) per gram of defatted kernel weight (µmol TE g −1 ).
The ABTS assay was carried out according to previous reports [21,50]. Briefly, an ABTS + solution (7.0 mM) was prepared by dissolving 38.36 mg of ABTS in 10 mL deionized water, mixed with a potassium persulfate solution (2.45 mM) at a ratio of 1:1 (v/v); it was kept in the dark for at least 16 h to form radicals. Before use, the ABTS + solution was diluted with ethanol to an absorbance of 0.70 ± 0.05 at 734 nm with UV spectrophotometer. Then, 40-µL extracts were mixed with 2 mL ABTS + solution, and the absorbance was measured at 734 nm after 6 min. Methanol without sample was used as a blank and treated with the same method as the samples. Absorbance of the blank was subtracted from each sample. Trolox was used as a standard reference, and the results were expressed as µmol trolox equivalents per gram of defatted kernel weight (µmol TE g −1 ).

UPLC-Q/TOF-MS
Methanol extracts of pecan kernels (Pawnee) at 10 developmental stages were mixed together. Mixed samples were analyzed using a Waters ACQUITY UPLC system (Waters, Milford, MA, USA) coupled with a Waters ACQUITY Q-TOF mass spectrometer (Waters, Milford, MA). An ACQUITY UPLC BEH C 18 column (2.1 mm × 50 mm, 1.7 um, Waters, Milford, MA, USA) was used at 35 • C. The mobile phase was composed of acetonitrile (A) and water containing 0.1% formic acid (B, v/v) at a flow rate of 0.4 mL min −1 . The solvent gradient was as follows: 5-10% A at 0-4 min, 10-50% A at 4-13 min, and 50-100% A at 13-20 min. Both negative and positive ion modes were used in this experiment, with a full mass scan at m/z 100-1500. The parameters of MS were set as follows: capillary voltage, 3 kV; source temperature, 120 • C; desolvation temperature, 350 • C; cone voltage, 50 V; cone gas flow, 50 L h −1 ; desolvation gas flow, 600 L h −1 . Rutin was used as the lock mass.
The identification of compounds were carried out first by comparing the retention time, [M − H] − ion, MS/MS fragments, and UV/Vis spectra data with previous reports on pecans to find known compounds. Then, available commercial standards including gallic acid, (+)-catechin, (−)-epicatechin, EGCG, and ellagic acid were used to further confirm these compounds. Compounds not found in the reports on pecans were searched for in the reports on the genus Carya, and then the family Juglandaceae, together with the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. Peaks found in all three biological replicates with mass measurement errors (∆m) lower than 10 ppm were recorded in Table 1.

HPLC
HPLC were performed on Agilent 1100 HPLC (Agilent Technologies, San Diego, CA, USA) with a C 18 column (Gemini, 250 mm × 4.6 mm, 5 µm particle size, Phenomenex, Torrance, CA, USA) conducted at 35 • C. The mobile phase was composed of acetonitrile (A) and water containing 2% of glacial acetic acid (B, v/v) a flow rate of 1 mL min −1 . The solvent gradient was as follows: 5-37% A at 0-40 min. Detection wavelengths were 250 nm (ellagic acid) and 280 nm (other four phenolics). Standard compounds were analyzed under the same conditions and standard curves were drawn; then contents of phenolic compounds were calculated using regression equations.

Statistical Analysis
The data were analyzed using Excel and SPSS (version 18.0) software. MS data was processed using Mass Lynx (version 4.1, Waters MS Technologies, Manchester, UK) software. All analyses were performed in triplicate. One-way analysis of variation (ANOVA) was performed and the mean values were compared by Turkey's test; differences at p < 0.05 were considered to be significant. Correlations were also obtained using SPSS software.

Conclusions
Dynamic changes in TPC, CT, TFC, phenolics, and antioxidant capacity of five widely cultivated Chinese-grown pecan cultivars (Pawnee, Stuart, Wichita, Jinhua, and Shaoxing) were analyzed. Similar dynamic change patterns were found: values were highest at the water or milk stages, lowest at milk or dough stages, and slightly varied at kernel stages. Phenolics profiles were also performed using pecan kernels. Using UPLC-Q/TOF-MS, 40 phenolics were tentatively identified in pecan mixed kernels, of which ellagic acid rhamnoside and ellagic acid rutinoside were first reported in the genus Carya; six phenolics, i.e., bis-HHDP-glucose (pedunculagin/casuariin isomer, two isomers), HHDP-valoneoyl-glucose (praecoxin A/platycariin isomer), valoneoyl-glucose, glansreginin A, and glansreginin B were first reported in pecans; and methyl gallate was first reported in pecan kernels. (+)-Catechin was found to be the most abundant phenolic in immature kernels and may convert to proanthocyanidins and oligomeric proanthocyanidins, which may be the major antioxidant constituents at the kernel stages. This experiment only carried on for one ripening cycle; further research is needed using samples of more cycles to characterize more precisely the tendency for phenolic compositions. Further research is also needed to investigate the accumulation of these newly found phenolics in pecan kernels and testa, and their relationship with the antioxidant capacity of pecan kernels.