Effect of Ozone and Drying Treatments on Phenolic Compounds and Antioxidant Activity in Bee Pollen

Featured Application

Bee pollen is a nutritious and functional food. However, fresh bee pollen has a high microbiological load. Treating it with ozone reduces this load and makes it easier to preserve and safer. In this study, we show that ozone treatment did not negatively affect the total amount of phenolic compounds or the antioxidant activity of this product. Therefore, this is one more reason to apply this treatment.

Abstract

Bee pollen is a food with high nutritional value and important functional properties. It is usually consumed as dried pollen, due to the need for a preservation process that controls the high microbiological load of the fresh product. Unfortunately, dry pollen is unattractive to consumers. In this sense, the use of ozone may be an alternative for preserving fresh pollen, since it reduces the microbiological load, allowing for other preservation methods, including refrigeration, and it preserves the original texture and flavors of the product, making it more palatable. However, it is important to know how ozone can affect some of the bioactive properties of pollen, such as phenolic compounds (PC), or the antioxidant properties of this food. The aim of this research was to assess the effect of ozone treatment on the above-mentioned properties and to compare it with the conventional drying treatment. For this purpose, 19 samples of fresh bee pollen were collected. From each sample, five subsamples were obtained, two of which were treated with ozone for 1 h and 2 h (O1 and O2, respectively), two were dried for 4 h and 8 h (D4 and D8, respectively), and the fifth subsample remained as the untreated control (C). The results showed that ozone treatments did not have a negative effect on phenolic content (PC) or on antioxidant activity (AA). This would be positive for the use of this decontamination method. In contrast, traditional drying treatments significantly reduced total PC while increasing AA compared to C, O1, and O2. A low correlation was found between these variables in the studied samples.

1. Introduction

The current trend in society is to consume healthy, natural, and, if possible, functional foods. Several bee products, such as honey, royal jelly, and bee pollen, have these characteristics. Of the above, pollen is the least well-known, but its consumption is growing, and there are reports predicting a progressive increase in its use of 6.1% between 2024 and 2030 [1].

Bees collect pollen from flowers and transport it to their hives, where they use it as the main source of protein for the colony. They store the surplus in the cells of the honeycomb to use in times of scarcity [2]. Forager bees are impregnated with pollen during their visits to flowers, and through an elaborate behaviour of brushing, agglomerating, and adding their own glandular secretions and nectar, they transform it into pellets, which are transported to the hives in baskets located in the third pair of legs, known as corbiculae. This is why these pollen pellets are also called corbicular pollen [3]. What we know as commercial bee pollen, used for human consumption, consists of these pellets. Beekeepers remove the corbicular pollen from the bees just as they are entering the hive by placing a pollen trap at the entrance of the hive. Pollen traps force the bees to pass tightly through 5 mm holes, causing the corbicular pollen to detach and fall into a lower drawer, from where the beekeepers collect it [4,5].

Bee pollen is considered a superfood in the human diet. Bee pollen has been shown to be useful in combating some diseases, such as prostate diseases, sugar regulation, and blood pressure, among others [6,7,8]. These benefits for human health are due to the high nutritional value and to its significant functional properties. It is rich in carbohydrates, proteins, and lipids [9,10], and it contains multiple bioactive metabolites [11], such as phenolic compounds and carotenoids [12]. Although diverse natural metabolites participate in free radical scavenging activity, it is considered that phenolic acids and flavonoids are primarily responsible for most of the antioxidant properties [13,14]. However, the antioxidant activity and polyphenolic content vary notably among different pollens due to the particularities of the source plant species and the different geographical areas [15]. Regular consumption of bee pollen and, therefore, the intake of these metabolites are known to reduce oxidative stress and inhibit macromolecular oxidation, which is related to their high antioxidant activity. Moreover, they contribute positively to reducing the risk of degenerative diseases [16,17]. In order to maintain all the properties mentioned above, it is necessary to study the preservation of bee pollen [18].

In any case, the chemical composition of bee pollen mostly depends on its botanical and geographical origin, method of pollen extraction, and pollen storage [19,20,21]. After collection, the pollen composition is changed by the action of bee enzymes. In addition, its chemical composition is influenced by the treatment processes of the fresh bee pollen, especially if these treatments involve heat [22,23].

Fresh bee pollen has a granular structure; it is very palatable and presents a very soft and fluffy texture. It has a pleasant flavour, which depends on its floral origin. Unfortunately, fresh bee pollen can have a high microbiological load, which could represent a serious risk to consumer health [24,25,26,27]. In this sense, preservation methods must be applied for commercialisation. Bee pollen is usually sold dried, with a moisture content that should not exceed 5–8%. This is achieved by drying it with heat and air flow. This preservation method prevents the growth of microorganisms that can spoil it and even pose food safety risks to consumers. At present, to preserve commercial bee pollen, the two principals techniques used are drying and freezing. Drying stops the growth of bacteria and fungi commonly found in fresh pollen [28]. Moreover, the texture of dry bee pollen and a taste reminiscent of straw or dried grass often causes rejection by consumers. For this reason, it is usually consumed dissolved in juices or dairy products. In addition, several components of bee pollen could be affected by the drying process, causing the loss of some of its nutritional qualities [29,30,31]. On the other hand, fresh frozen pollen remains a food with a very pleasant taste and texture. This method preserves the organoleptic characteristics and nutritional properties of the fresh bee pollen better than the previous one [3,5,26,32]. Furthermore, it drastically delays the growth of microorganisms; however, the initial microbiological load is not eliminated, which can cause a serious risk to human health [24,25]. This is especially dangerous if the cold chain is not appropriately maintained during the transport of the product and until it finally reaches the consumer [3,5,24,25,29,30,31,32]. In addition, gamma irradiation treatment is another preservation method, but less frequently used. Nevertheless, this method is very effective, but it often generates rejection among consumers, especially those who prefer natural foods [33,34].

Due to the above-mentioned methods, new preservation methods are being studied to reduce the initial microbiological load of the bee pollen, in order to make it a safer product that can be preserved both as frozen and refrigerated food. Among these new methods of interest is ozonization. This technique is commonly used in the decontamination of water and food and could also be an alternative method used for bee pollen. Cabello et al. [26] proposed the use of ozone to reduce the microbiological load of bee pollen, with very interesting results, achieving reductions of 3, 1.5, and 1.7 log cycles for Enterobacteriaceae, mesophilic aerobes, and molds and yeasts. Storage in the refrigerator (4–6 °C) for several weeks would be possible, making it an interesting alternative treatment. However, the method raises questions about the possible effects of ozone treatment on the components of bee pollen, especially those that contribute to the functional properties of this food, such as phenolic compounds, or the influence on the antioxidant activity of this food. In this sense, recent studies [34] have found that ozone treatment reduces the microbiological load of bee pollen without affecting its AA.

The main objective of this study is to provide new information on the effect of ozone treatment on bee pollen. Particularly, the objective is to determine how ozone affects the polyphenol content and antioxidant activity of this natural product, and to compare it with the traditional drying method.

2. Materials and Methods

2.1. Bee Pollen Samples

Nineteen samples of fresh multifloral bee pollen were acquired directly from professional beekeepers. All farms were located in the region of Andalusia (Spain). The location of these farms is shown in Figure 1. The samples were codified and transported to our laboratory while maintaining their frozen state. The samples were kept at −20 °C until use.

Figure 1. Origin of samples. Nineteen fresh pollen samples from five provinces in Andalusia (southern Spain) were used in the study. The location of the apiaries and the coding of the samples are shown.

Five 50 g subsamples were taken from each sample: (i) the first were placed in plastic bags and subjected to ozone (200 mg/h) for 1 h using a Multifunctional Ozono Care/Life^® Vida10 portable ozone generator (P.R.C.) (O1), replicating the maximum treatment reported by Cabello et al. [26]); (ii) the second was ozonated for 2 h (O2) to assess the impact of doubling the exposure time; (iii) the third was dried at 40 °C for 4 h (D4) using an airflow pollen dryer (MainoRoberto & C. S. n. c. Mod. INGEGNOSOX5, Oltrona di San Mamette, Italy); and (iv) the fourth was dried at the same temperature for 8 h (D8). The 40 °C temperature was selected as suitable for reducing pollen moisture content to 6%, the adequate level for preservation [25]. The remaining subsample was kept untreated as the control (C). All subsamples were stored at −20 °C until analysis.

2.2. Bee Pollen Sample Preparation

Before analyzing the ground bee pollen samples, an extraction process was carried out according to the method used by Campos et al. [35], with some modifications. Briefly, 10 mg of bee pollen from each sample was placed in a glass vial, and 1 mL of ethanol-water solution (50% v/v) was added. After, it was sonicated for 60 min, and the resultant extract was centrifuged at 5000 rpm for 10 min. This process was carried out in duplicate.

2.3. Total Phenolic Compounds

The total phenolic compounds were quantified using the established Folin–Ciocalteu method [36]. The procedure involved adding 50 µL of the bee pollen extract to 1.25 mL of the Folin–Ciocalteu reagent, which had been previously diluted 1:5 with distilled water. After vigorous shaking and a one-minute rest, 1 mL of a 10% w/v sodium carbonate solution was introduced, and the mixture was incubated in the dark for 30 min. The absorbance of the resultant blue complex, formed by the oxidation of phenolic compounds (PC), was then measured at 760 nm using a Beckman DU 640 spectrophotometer (Beckman, Brea, CA, USA). Quantification was achieved via a gallic acid calibration curve (0.01 to 1 g/L). The results were expressed in mg gallic acid/100 g dry bee pollen.

2.4. Antioxidant Activity

The antioxidant activity (AA) of the bee pollen extract was determined using the DPPH assay, based on the methodology established by Katalinic et al. [37], with some modifications. To perform the test, 200 µL of the sample (diluted 1:10) was added to 3 mL of a 45 mg/L DPPH solution. For the control, 200 µL of water was substituted for the sample. The absorbance of the control and the sample was measured at 517 nm in a Beckman DU 640 spectrophotometer (Beckman, USA). A calibration curve was obtained with the Trolox standard in the range of 10–200 mg Trolox/L, and the percentage inhibition was calculated according to the following formula:

$$\mathrm{Percentage}\mathrm{inhibition}={\displaystyle \frac{{\mathrm{Abs}}_{\mathrm{control}}-{\mathrm{Abs}}_{\mathrm{sample}}}{{\mathrm{Abs}}_{\mathrm{control}}}}\times 100$$

The results were expressed in mg Trolox/100 g dry bee pollen.

2.5. Determination of Botanical Origin

To determine the botanical origin of the bee pollen samples, the methodology described by Rojo et al. [38] and adapted by the RNM130-Systematic and Applied Botany research group of the Department of Botany, Ecology and Plant Physiology, University of Córdoba (Spain) was used. Each sample of the original freezer bee pollen was analyzed in three replicates. Original samples were thoroughly homogenized, and subsequently, 2 g of each sample were placed into separate vials. A colorimetric separation was then performed to obtain distinct subsamples. Samples were diluted using a vortex mixer in distilled water, applying two different volumes according to the initial mass of the solid material: (i) samples with a mass ≥ 0.1 g were diluted in 10 mL of distilled water, (ii) samples with a mass < 0.1 g were diluted in 1.5 mL of distilled water. This approach ensured comparable concentration ranges across all samples while accounting for the limited material availability in low-mass cases. From each dilution, 100 µL of sediment was taken to prepare microscope slides. The sediment was placed on a cover glass positioned on a heating plate at approximately 50 °C. Once the liquid had evaporated, a drop of fuchsin-stained glycerogelatin was added to stain the sample and facilitate the identification of pollen grains under an optical microscope (Nikon Eclipse, E400, 40× objective, East Paul Dirac Drive. The Florida State University. Tallahassee, FL, USA). The preparation was covered with a coverslip. After 24 h, excess fuchsin-stained glycerogelatin was removed, and the slide was sealed. The botanical origin of each subsample was determined by counting 100 pollen grains per subsample. The pollen spectra of the samples were calculated considering both the weight of each subsample and its botanical origin, and the results were expressed as percentages.

Pollen types were identified and classified according to their morphological characteristics, following standard palynological criteria. For pollen type identification and nomenclature, the reference melissopalynotheque developed by the RNM130 research group was used [39].

2.6. Data Analysis

Data were statistically processed using SPSS (Statistical Package for the Social Sciences) Statistics software for Windows, IBM Corp., 2016. Version 24.0. IBM Corp., Armonk, NY, USA. All the variables available were tested to check whether the data violated the assumptions for regular parametric tests to report valid results. Parametric statistics were applied when possible. When the data results were non-normally distributed or there was no variance homogeneity (heteroscedasticity), non-parametric statistics were used. The tests are specified in the results.

3. Results

The average of the total content of PC, expressed as mg of gallic acid in 100 g of dry bee pollen, and AA, expressed as mg Trolox in 100 g dry bee pollen for all samples, is shown in Table 1 (more information about the descriptive statistics is shown in the Supplementary Material, Table S1). Since the data did not follow a normal distribution for either variable (Shapiro–Wilk test, p ≤ 0.05), non-parametric statistics were applied. The fresh bee pollen samples and those treated with ozone for 1 h and 2 h (O1 and O2) showed a significantly higher total amount of PC than the samples dried for 4 h and 8 h (D4 and D8). In contrast, bee pollen samples D4 and D8 showed significantly higher AA than fresh pollen samples and those treated with ozone O1 and O2 (Kruskal–Wallis test, p ≤ 0.05; Mann–Whitney U test, p ≤ 0.05). When all the samples and treatments were considered together, a significant but low correlation was found between the total amount of PC and the AA recorded in the samples (Kendall’s tau_b, p = 0.000, r = 0.182).

Table 1. Average of total phenolic compounds (mg gallic acid/100 g dry bee pollen) and antioxidant activity (mg Trolox/100 g dry bee pollen) obtained from bee pollen subjected to different preservation treatments.

Samples	Code	Total Phenolic Compounds	Antioxidant Activity
Fresh bee pollen (Control)	C	1121 ± 351 a	2.82 ± 0.82 a
Drying 4 h	D4	1000 ± 361 b	3.28 ± 1.12 b
Drying 8 h	D8	1011 ± 351 b	3.30 ± 1.15 b
Ozone 1 h	O1	1119 ± 367 a	2.81 ± 0.89 a
Ozone 2 h	O2	1216 ± 392 a	2.79 ± 0.90 a

(n = 76, Mean ± Standard Deviation). Different superscript letters in each of the variables indicate significant differences between treatments (Mann–Whitney U test, p ≤ 0.05).

Uneven behaviour was found when each sample was considered independently. Each bee pollen sample responded differently when the treatments were applied. Multiple significant differences were found within each sample for both PC and AA (see Figure 2 and Figure 3 and Table S2 in the Supplementary Materials).

Figure 2. Quantification of the total PC obtained from the nineteen bee pollen samples subjected to different preservation treatments: fresh bee pollen (control, C), drying 4 h (D4), drying 8 h (D8), ozone 1 h (O1), and ozone 2 h (O2). The results are shown as Mean ± Standard Deviation (n = 4). More information is shown in Table S2.

Figure 3. AA obtained from the nineteen bee pollen samples subjected to different preservation treatments: fresh bee pollen (control, C), drying 4 h (D4), drying 8 h (D8), ozone 1 h (O1), and ozone 2 h (O2). The results are shown as Mean ± Standard Deviation (n = 4). More information is shown in Table S2.

On the other hand, the effect sizes (Cohen’s f test) of the two factors (treatments and samples) considered in this study on the PC and AA variables were calculated (Table 2). The variable “treatments” showed a significant effect size on both PC and AA, although this effect was not high. In contrast, the samples factor did show a high effect size on PC and AA (detailed information about Cohen’s f test is shown in Supplementary Material “Meaning and calculation of the Cohen’s f test”).

Table 2. Effect sizes (Cohen’s f test) of the factors on the variability recorded in the variables. Values up to 0.1 mean the effect is considered negligible. Values between 0.1 and 0.25 are considered indicative of a low effect. Values between 0.25 and 0.4 are indicative of a moderate effect. Finally, greater than 0.4 indicates a strong effect [40].

	Treatment	Sample
mg gallic acid/100 g dry bee pollen	0.22	2.65
mg Trolox/100 g dry bee pollen	0.24	0.99

When initial samples of fresh bee pollen were considered, a slight significant positive correlation between the two variables can be observed (Kendall’s tau_b, p = 0.044, r = 0.158). On the contrary, when each sample was considered independently, a wide diversity in terms of total polyphenol content and antioxidant activity was found. No relationship between the variables was observed. A higher amount of total PC does not necessarily imply greater AA (see Figure 4).

Figure 4. Average of PC (mg gallic acid/100 g dry bee pollen) and AA (mg Trolox/100 g dry bee pollen) obtained from fresh bee pollen samples. No visual relationship was found between variables. Results are shown as Mean ± Standard Deviation (n = 4).

This is supported by the fact that no significant correlation was found between the two variables for each of the samples (Kendall’s tau b, p ≤ 0.05), except for sample G2, in which a high significant correlation was found (Kendall’s tau b, p ≤ 0.000, r = 0.800).

Because a significant influence on the PC content and AA could be due to the floral origin of the samples [41], each fresh bee pollen sample was considered separately. A wide variety of floral origins of pollen samples was found for both the set of samples and for each sample. The number of floral pollen types in the samples ranged from 4 to 31. Cistaceae (Cistus) pollen was the most represented pollen group in the sample set, accounting for 33.85% of the total pollen, followed by Boraginaceae (Echium, 15.54%), Cistaceae (Helianthemum, 10.96%), Fagaceae evergreen Quercus pollen type (Q. coccifera and Q. ilex, 9.99%) and Rosaceae type Rubus (Eriobotrya japonica, Geum, Pyracantha, Rubus, 3.23%). Boraginaceae (Cynoglossum), Brassicaceae; Fagaceae deciduous Quercus pollen type (Q. faginea, Q. humilis, Q. petrea, Q. robur, including the exception of Q. suber), and Oleaceae (Olea europaea) were represented between 2.00% and 3.00%. The remaining pollen types were represented in an amount less than 2.00%. Some of the pollen types were present in very small proportions and only in some samples (more detailed information is shown in Table S3).

Sample S1 contained the highest total amount of PC, and up to 12 different types of pollen were detected in it. Most of them corresponded to the Fagaceae (Q. coccifera, Q. ilex, 64.00% of the pollen), followed by Cistaceae (Cistus, 13.00%) and Rosaceae (Rosa), Cistaceae (Helianthemum), Boraginaceae (Echium), Myrtaceae (Eucalyptus), and Ranunculaceae (Clematis), with 7.00%, 6.50%, 4.67%, 2.00%, and 1.00%, respectively. The remaining pollen types were represented by less than 1%. In contrast, the highest AA was recorded in sample Co1. Up to 17 pollen types were recorded in this sample, the most represented being Rosaceae type Rubus (Eriobotrya japonica, Geum, Pyracantha, Rubus, 30.14%), Fagaceae (Q. coccifera and Q. ilex, 23.44%), Cistaceae (Cistus, 19.62%), Boraginaceae (Cynoglossum, 9.09%), and Boraginaceae (Borago officinalis, 5.26%). The remaining pollen types were found in smaller quantities.

Sample G2 was the only one in which a high correlation between the total amount of PC and AA was found. This sample was dominated by the pollen types Cistaceae (Helianthemum, 36.95%), Fagaceae (Q. coccifera, and Q. ilex, 24.63%), Cistaceae (Cistus., 15.76%), and Tamaricaceae (Tamarix, 5.91%). The remaining pollen types, up to a total of seven more, were represented in smaller quantities. In any case, this sample did not show exceptionally high total PC content or AA (see Figure 2 and Figure 3).

On the other hand, the correlation between the different botanical families found in the bee pollen samples and the PC and AA was analyzed (Kendall’s tau_b, p ≤ 0.05). When the correlation coefficient was high (r ≥ 0.5) [42], only the family Fagaceae evergreen Quercus pollen type Q. coccifera, Q. ile showed a significant positive correlation with total phenolic compounds for all treatments, with correlation coefficients (r) of 0.587, 0.600, 0.584, 0.593, and 0.648 for C, O1, O2, D4, and D8, respectively. No other family showed a significant high correlation with total phenolic compounds, either in the original fresh pollen samples or after being subjected to the different treatments. Moreover, a significant high correlation was found with respect to AA for the botanical families Arecaceae type Phoenix (Chamaerops humilis, Phoenix) (r = 0.505 and 0.522 for O1 and O2, respectively), Boraginaceae Echium (r = 0.600 and 0.632, for O1 and O2, respectively), and Liliaceae Muscari (r = 0.503 for O2). In addition, a significant high correlation was found between dehydration treatments and botanical families Boraginaceae Borago officinalis (r = 0.582 and 0.585 for D4 Y D8, respectively) and Cistaceae Helianthemum (r = 0.517 for D8). No significant large negative correlation was found between botanical families and PC and AA for any of the treatments, except for the botanical family Cistaceae Cistus (r = −0.516 for D8). This last botanical family also showed a significant low or medium negative correlation with total PC and AA in the control treatment (C), with AA in treatments O1 and O2, with PC and AA in treatment D4, and with PC in treatment D8. Multiple significant correlations with medium and low r values were also found for other botanical families and PC and AA for all treatments. Detailed information is shown in Table S4.

4. Discussion

In recent years, several studies have investigated the bioactive phenolic components of bee pollen. The total content of PC found in the fresh pollen samples (C) studied in this research from Andalusia (Spain) samples (684.59 ± 14.84 to 2075.56 ± 43.12 mg gallic acid/100 g dry bee pollen) (see Table S2) were similar to those found in samples from Romania (ranged from 440 ± 10 to 1640 ± 30 mg gallic acid/100 g dry bee pollen) [43]. The PC values in the samples studied are consistent with the results reported by Ilie et al. [44] for Romanian bee pollen (values in samples vary between 1080 ± 2 and 1620 ± 2 mg gallic acid/100 g dry bee pollen). In addition, in the study of the Andalusian samples, higher levels of total PC were obtained when compared with the findings of Aylanc et al. [18] and Domínguez-Valhondo et al. [20] (which ranged from 562 ± 18 to 857 ± 33 mg gallic acid/100 g dry bee pollen). Compared to other studies involving bee pollen samples from Spain (the range was between 772 and 2639 mg gallic acid/100 g dry bee pollen), Argelian (ranged from 772 ± 29 to 2350 ± 148 mg gallic acid/100 g dry bee pollen), and Morocco (ranged from 807 ± 104 to 3239 ± 15 mg gallic acid/100 g dry bee pollen) [38,41,45], the extracts of the samples studied exhibited lower contents of total phenolic compounds.

The results of the AA of the fresh bee pollen samples (C) ranged from 1.51 ± 0.05 to 5.56 ± 0.30 mg Trolox/100 g dry bee pollen (see Table S2). The variation could be attributed to the botanical origins and the presence of various antioxidants, such as flavonoids, phenolic acids, and vitamins C and E [46]. These results for AA were closer to those reported by Harbane et al. [47] and Zaidi et al. [48] for Algerian products. Compared to other studies involving bee pollen samples from Spain [29] (ranging from 3.64 ± 0.23 to 7.32 ± 0.08 mg Trolox/100 g dry bee pollen), the extracts of the studied samples exhibited lower values of AA.

Recent studies have found a significant relationship between PC and AA in bee pollen [36,38]. In contrast, several studies have concluded that antioxidant activity is not clearly associated with its total phenolic content [15,16,43]. According to the latter, the results showed a weak relationship between the amount of PC and AA. It is possible that the presence of specific phenolic compounds determines the increase in antioxidant activity [49]. This suggests that the identity of the phenolic compounds, rather than their concentration, may be a more critical factor. Furthermore, the antioxidant activity is influenced by synergistic and antagonistic effects among phenolic compounds and their interactions with other phytochemicals (such as Vitamin C) within the extract [50]. The polyphenolic profile is variable in bee pollen, and the antioxidant activity of polyphenols depends on the number and location of the hydroxyl groups in their chemical structure [51]. Hence, it is important to relate the botanical origin with the individual phenolic compounds, because these relationships can contribute to the discrimination of the antioxidant capacity of some pollen samples based on their floral origin.

The results point out that the contribution of PC to AA in the samples studied is relatively low, as shown by the low correlation between the variables, both for the original fresh pollen samples (a visual assessment can be seen in Figure 4) and when considering all samples and treatments together. This could be explained because multiple causes can affect the characteristics of bee pollen. Factors such as botanical origin, the region of origin, or the soil, among others, can significantly influence the composition of bee pollen and its AA [18,19,20,21]. In addition, other various components of bee pollen, different from PC, such as carotenoids, vitamins B, C, and E, or phenolamines, are also involved in AA [12,52,53]. Furthermore, based on our results, these components probably have a greater impact in this regard than PC. This is a topic that should be studied in the future.

Chemical composition of the bee pollen is influenced by the treatment processes applied to the fresh pollen, especially if these treatments involve heat [22,23]. In response to the objective of this research, which was to study how different pollen preservation treatments could affect its polyphenol content and AA, it also obtained results that were not as expected. Fresh pollen (C) and pollen treated for 1 and 2 h with ozone (O1 and O2) showed a higher polyphenol content than pollen dehydrated for 4 and 8 h (D4 and D8). This indicates that the total phenolic content was negatively affected by the drying process as a probable consequence of the activity of polyphenol oxidase and peroxidases, which can be attributed to the enzyme release following the freeze–thawing process preceding oven drying [54]. Similar results concerning to degradation of phenolic content were found by other authors [29,55].

However, ozone treatment even increased the total PC in some samples (Table S2), especially when ozone was applied for two hours, with significant differences for many of the samples (see Figure 2 and Table S2). These findings could be due to the changes in enzyme concentration and activity caused by ozone treatments [56], which are consistent with previous studies that reported similar results on bee pollen [34] and in fruit and vegetables [56]. However, since the effect sizes of the treatments on PC and AA were not high (see Table 2), further studies would be necessary to confirm this fact more reliably. Given the antioxidant role that polyphenols play in the cells, most of them are biosynthesized in both vines and grapes as a response to biotic and abiotic stresses due to the activation of the phenylpropanoid pathway [57]. However, the effect of ozonation on polyphenol content is not always clear; some papers describe a positive effect of ozone in terms of polyphenol accumulation [58].

Contrary to the above, as can be observed in Figure 3, AA was significantly higher in the D4 and D8 samples. In any case, the results show that ozone treatment did not negatively affect either the total amount of PC or AA compared to the fresh pollen samples. This fact could be a positive aspect in favour of using ozone as a method of preserving bee pollen compared to traditional methods. On the other hand, the increase in AA in heat-treated samples, compared to the control bee pollen, could be explained by the fact that during the drying process, there is a concentration of these other compounds with AA. However, the design of this research does not allow us to support this hypothesis. On the other hand, the increase in AA value with temperature, even though the phenolic content decreases, may be due to the increase in the concentration of specific polyphenols that contribute greatly to that antioxidant activity value, or affect other compounds that positively influence the value of this activity, such as vitamin C or carotenes. Similar results have been found by other authors [59].

The results of this study also show high effect sizes of the samples on the variability of AA and even higher on the variability in PC (see Table 2). This is probably due to the difference in pollen composition, which is strongly influenced by its floral origin. When the botanical origin of the samples was considered, a wide variety was found in the sample set and in each sample. The different botanical origins of the pollen probably influenced the composition, including the composition of PC and other substances with AA. Rojo et al. [38] point out that the nutraceutical properties of bee pollen would be enhanced by increasing the diversity of pollen in the samples. Up to 31 botanical families in the sample set were found. Cistaceae, Boraginaceae, Fagaceae, and Rosaceae were the most frequent. However, they were found in many samples, regardless of the amount of PC and AA they exhibited. For this reason, the correlation between the presence of each botanical family and the PC and AA recorded in the samples after applying each treatment was studied. The family Fagaceae Evergreen Quercus pollen type (Q. coccifera, Q. ilex) was the only one that showed a large, significant (positive) correlation with polyphenol content for all treatments. In contrast, it did not show a significantly high correlation with antioxidant activity. This would reinforce the hypothesis that polyphenols may not be a fundamental component in explaining AA in bee pollen samples. This is also reinforced by the fact that only the botanical families Arecaceae type Phoenix (Chamaerops humilis, Phoenix), Boraginaceae Borago officinalis, Boraginaceae Echium, Ericaceae Erica, Liliaceae Muscari, and Cistaceae Helianthemum, showed a large, significant correlation with AA, and only in some of the treatments, and none of them showed it with PC (See Table S3).

On the other hand, the high-positive significant correlations could not be explained by the greater presence of pollen in the samples, as the other botanical families: Arecaceae type Phoenix (Chamaerops humilis, Phoenix), Boraginaceae Borago officinalis, Cistaceae Helianthemum, Ericaceae Erica, and Liliaceae Muscari, were only detected in 2, 3, 8, 2, and 3 samples, respectively. Only Boraginaceae Echium was represented in a high number of samples (n = 17). It is important to emphasize that the family Cistaceae Cistus only showed negative correlation values with respect to PC and AA in the different treatments, although they were mainly medium or low. This fact is important, since this botanical family was present in all samples and accounted for 33% of the bee pollen in the samples as a whole.

5. Conclusions

Ozone treatments did not negatively affect the total amount of phenolic compounds and antioxidant activity in the treated samples. In contrast, drying treatments reduced the total phenolic content while increasing the antioxidant activity of bee pollen.

Based on the results obtained, no evidence was found to contraindicate the use of ozone as a preservation method for bee pollen. However, this study did not include other bee pollen compounds that might be susceptible to this treatment, such as carotenoids, vitamins B, C, and E, or phenolamines, among others. Therefore, subsequent investigations should complement these findings and ascertain that these compounds remain unaffected by this treatment. On the other hand, a significant but low correlation was found between the total concentration of phenolic compounds and the antioxidant activity of the samples. Probably, because other components of bee pollen also had an important influence on the antioxidant activity displayed by this natural product. Furthermore, the difficulty in achieving conclusive results during the evaluation of treatments on PC and AA was partly attributable to the variability of botanical origin. Given the heterogeneity of the different samples, future research should focus on utilizing monofloral samples to better isolate treatment effects. Finally, other studies could be conducted to optimize ozone application methods for bee pollen preservation.