Quality Assessment of Bee Pollen : A Cross Sectional Survey in Bulgaria

Background: Due to its nutrient content, a variety of spoilage microorganisms can grow in bee pollen, especially when handling practices are not appropriate. So, this survey was designed to assess the physicochemical and microbiological properties of bee pollen collected from Bulgaria. Methods: In June 2014, 13 fresh and 19 dried bee pollen samples were collected from Bulgaria. Water activity (aw), pH, Total Titratable Acidity (TTA), and bacteriological and fungal counts were evaluated. Data were analyzed using Statistica ver. 10. Results: Mean aw value for fresh and dried bee pollen samples were 0.717 and 0.359, respectively showing significant (p<0.01) difference. The mean pH values of fresh bee pollen samples (4.23) was significantly (p<0.01) lower than that of dried ones (5.21). Also, there is statistically significant difference (p<0.01) between mean TTA value of fresh bee pollen (3.69 g/100 g) and mean TTA value of dried samples (2.09 g/100 g). The mean total viable count in fresh bee pollen samples was 182153.8 Colony Forming Unit (CFU)/g which was significantly (p<0.01) higher than that of dried samples (30352.6 CFU/g); whereas statistical analysis demonstrated that there were no significant differences (p>0.01) among the fresh and dried pollen for Enterobacteriaceae, lactic acid bacteria, and Staphylococcus spp. counts. The fungal colony count in the fresh bee pollen samples varied from 560 to 37000 CFU/g which were significantly (p<0.01) higher than that in the dried ones. Conclusion: The fresh bee pollen analyzed in this survey had poor microbial quality compared to the dried bee pollen. There is also need to adopt appropriate manufacturing practices to prevent possible contamination by equipment or handling of Bulgarian bee


Introduction
Pollen grain is a tiny body with various structures, produced in the male apparatus of seed-bearing plants and is an important source of macro-and micro-nutrients (Bastos et al., 2004;Isidorov et al., 2009), and substantial amounts of polyphenols and primarily flavonoids (Carpes et al., 2009;Leja et al., 2007).
Due to its high nutritional composition, particularly in terms of its energy supply, bee pollen is considered as a food source for man (Atanassova and Lazarova, 2010;Block et al., 1997;Campos et al., 2008).Besides being a food source, pollen is also be used for therapeutic purposes (Kacaniova et al., 2011).However, due to its nutrient content, a variety of spoilage microorganisms can grow in bee pollen, especially when the collection, drying, storage, and marketing practices are not appropriate (Gonzalez et al., 2005;Medina et al., 2004).Also, contamination can be attributed to various factors and sources, weather, plant materials, insects, agricultural devices, etc. (Hani et al., 2012).In this context, humans may possibly be exposed to the negative influence of different fungal species, fungal spores, other invasive microorganisms or toxic fungal metabolites such as mycotoxins, directly by consumption of contaminated foodstuffs.Number of studies have revealed significant contamination levels of bee pollen by pathogenic microorganisms (Belhadj et al., 2014;Brindza et al., 2010), which compromise its safety for consumers.Bulgaria and other countries of the world, including Brazil, Poland, and Switzerland have established national standards for bee pollen (Campos et al., 2008).So, this survey was designed to assess the physicochemical as well as microbiological properties of bee pollen collected from different regions of Bulgaria.

Sample collection and preparation
In June 2014, a total of 32 bee pollen samples (weighting 45-59 g per sample) consisting of 13 fresh and 19 dried ones were collected.They represented mixed pollen loads from different plants of eight different regions in Bulgaria from beekeepers as follow: Strandzha (n=4 fresh and 4 dried), Shumen (n=6 fresh and 6 dried), Sliven (n=1 fresh and 1 dried), Stara Zagora (n=2 dried), Karlovo (n=2 dried), Vratsa (n=2 fresh), Veliko Tamovo (n=3 dried), and Lovech (n=l dried).The samples were placed in sterile plastic zip-lock bags, vacuumed, packed, and stored at 4 °C until analysis.All the samples were handled aseptically to prevent secondary contamination.

Physicochemical analysis
The Water activity (a w ) of the pollen samples was determined using a dew-point a w meter (Wert-Messer Meter, Germany) with an auto-temperature control system.Five g of each sample was placed in the sample holder and a w value was recorded at 20±0.2 ºC.The pH measurements were done according to the modified method of Obadina et al. (2008).Briefly, 10 g of each sample was dissolved in 90 ml deionized water, homogenized, and filtered through a Whatman no. 1 filter paper.The pH level of each sample was then determined using a pH meter (Eutech pH 510 meter, Eutech Instruments, Singapore).
Measurement of Total Titratable Acidity (TTA) was performed following the method described by Xu et al. (2012) with little modification.One g of each sample was dissolved in 20 ml deionized water, homogenized, and filtered through a Whatman no. 1 filter paper.Two-drop phenolphthalein indicator was added to the filtrate and titrated against 0.1 M NaOH to a pH 8.2 (Eutech pH 510 meter, Eutech Instruments).The TTA (g/100g) was calculated as follows: Where titre is volume of NaOH consumed during titration (ml), N is the Normality of base (mol/l), E w is milli-equivalent weight of tartaric acid (0.075), and S w is the sample weight.

Microbiological analysis
One g of each pollen sample was added to nine ml sterile ringer solution and vortexed.Serially dilutions to 10 -4 were done.One ml of each aliquot concentration was inoculated onto McConkey agar (Sigma, USA), plate count agar (Hi Media Laboratories, Mumbai, India), mannitol salt agar (Hi Media Laboratories, Mumbai, India), de Man, Rogosa and Sharpe agar (Biolab SA, Switzerland), as well as Salmonella-Shigella agar (Merck-Darmstadt, Germany) using spread plate technique as described by Adebayo et al. (2013) with some modifications.The plates were incubated at 37 °C for 24 h, and then colonies were counted and recorded as Colony Forming Unit (CFU) per g of sample.
Using the method of Njobeh et al. (2009) with little modifications, 1 g from each pollen sample was weighed into a sterile test tube and diluted by 9 ml sterile ringer solution and vortexed, then serially diluted down to 10 -4 .Then, an aliquot of one ml from each sample was inoculated onto potato dextrose agar (Biolab SA, Switzerland) following the pour plate technique.The plates were incubated at 25 °C in the dark for seven days, and then colonies were counted and recorded in CFU/g.Thereafter, single colonies were sub-cultured on potato dextrose agar to obtain a pure fungal culture and incubated at 25 °C for 4-7 days followed by macromorphological characterization and identification.The distinct fungal species were done based on macro-and Downloaded from jfqhc.ssu.ac.ir at 4:22 IRST on Saturday December 1st 2018 Journal website: http://www.jfqhc.commicro-morphologic identification keys of Pitt and Hocking (2009).

Statistical analysis
Data were analyzed using Statistica ver. 10 (StatSoft, Inc., USA).All analytical measurements were carried out in two replicates.Significant levels were considered as p<0.01.

Results
Physicochemical characteristics of 32 Bulgarian bee pollen samples are shown in Table 1.Mean a w value for fresh and dried bee pollen samples were 0.717 and 0.359, respectively showing significant (p<0.01)difference.The mean pH values of fresh bee pollen samples (4.23) was significantly (p<0.01)lower than that of dried ones (5.21).Also, there is statistically significant difference (p<0.01) between mean TTA value of fresh bee pollen (3.69 g/100 g) and mean TTA value of dried samples (2.09 g/100 g).
As described in Table 2, the mean total viable count in fresh bee pollen samples (182153.8CFU/g) was significantly (p<0.01)higher than that of dried samples (30352.6CFU/g); whereas statistical analysis demonstrated that there were no significant differences (p>0.01)among the fresh and dried pollen for Enterobacteriaceae, lactic acid bacteria, as well as Staphylococcus spp.counts.
The fungal colony count in the fresh bee pollen samples varied from 560 to 37000 CFU/g with a mean of 10512.3CFU/g which were significantly (p<0.01)higher than fungal colony count in the dried ones varying from 150 to 11000 CFU/g with a mean of 2418.4CFU/g.Overall, 136 fungal isolates representing 49 fungal species were detected.Broadly, the genus Penicillium was the most frequently occurring that consisted of 49 isolates which among them P. brevicompactum was the most dominant species isolated from 14 different samples; also, Fusarium (5 isolates) was the least fungal contaminant (Table 3).

Discussion
In the present survey, we found that some Bulgarian bee pollen samples had not perfect quality characteristics, especially from view point of microbial quality.Pollen grains are highly sterile reproductive gametophytes from flowering plants, therefore any form of contamination is secondary (Shevtsova et al., 2014).Investigations revealed that some plant diseases may be transmitted by pollen (Card et al., 2007;Flores et al., 2005).Contamination of pollen could result from bees activities (foraging and transport), handling by humans, or from its natural habitat (Gilliam, 1979;Serra and Escola, 1997).Other extrinsic factors such as wind, rain-splash, dew or fog drip, and sprinkler irrigation splash can also contribute to pollen contamination (Lacey and West, 2007).Pollen, bacterial, fungal, and several other contaminants all co-exist in the aerospace (Lacey and West, 2007).Bees dampen the pollen with nectar and lug them in corbiculaes on their rearmost legs during transport (Atanassova and Lazarova, 2010); thus, bee pollen is more vulnerable to microbial contamination.Sanitary conditions around the hive, human handling operations such as harvesting, drying, packaging, and storage also have strong implications on pollen contamination (Serra and Escola, 1997).
The a w is an important factor in food safety and stability (Mathlouthi, 2001;Rahman, 2010).Higher a w tend to promote the proliferation of microorganisms (Abbas et al., 2009;Mathlouthi, 2001) such as pathogenic bacteria grown at a w >0.85, whereas fungi grown well at a w >0.6 (Rahman, 2010).The a w values seen in our study were too close to the levels that are favour for both bacterial and fungal activities.This agrees with observations by Feas et al. (2012) who obtained considerable microbial load in bee pollen at a w and pH range of 0.21-0.37 and 4.3-5.2,respectively.Although high a w affects other aspects of food quality/stability such as enzymatic activities and moisture sorption, more aseptic techniques should be employed in the dehydration and handling of pollen since this might reduce the risk of human-induced contaminations.The pH and TTA values are also suggestive of dynamic microbial activity (Jay et al., 2005).Gram-positive bacteria are known to ferment foods resulting in higher pH and TTA values.However, mixture of endogenous enzymes and other microorganisms such as yeast and mould from the native microflora of food substances are also known to participate in the fermentation processes (Jay et al., 2005).
High contamination of the pollen samples by microorganisms in this study is similar to observations by Shevtsova et al. (2014) on pollen samples from Slovakia.The presence of coliforms in food generally indicates faecal contamination, which causes various illnesses in humans (Jay et al., 2005).It is well known that some Enterobacteriaceae members are opportunistic pathogens rarely cause disease in healthy individuals (Sanders and Sanders, 1997).It is also important that in all dry bee pollen samples, the total aerobic plate count possessed less than 100000 CFU/g (Campos et al., 2008).Lactic acid bacteria are known to cause food spoilage such as milk fermentation (Zavisic et al., 2012).The growth of Gram-positive, facultative anaerobes such as Lactobacilli in the pollen samples of the current survey is similar with the findings of Belhadj et al. (2014) who isolated Lactobacillus plantarum, L. fermentum, Lactococcus lactis, L. ingluviei, Pediococcus pentosaceus, L. acidipiscis, and Weissella cibaria from raw bee pollen grains collected in some Algerian areas.
The present survey revealed biodiversity in fungal species in the bee pollen samples such as Aspergillus spp., Fusarium spp., and Penicillium spp.known to be major producers of mycotoxins which have adverse health implications on humans and animals alike (Beev et al., 2013;Jay et al., 2005).In this regard, Serra and Escola (1997) detected a high number of molds in bee pollen collected from Spain.Also, Kacaniova et al. (2011) found some mycotoxins in bee pollen sampled from Slovakia.Similar to our results, Gonzalez et al. (2005) established that Penicillium spp. is the predominant fungi in ready-to-eat bee pollen samples.In a similar work carried out in Italy, Nardoni et al. (2016) also identified variety of fungi in bee pollen, including Downloaded from jfqhc.ssu.ac.ir at 4:22 IRST on Saturday December 1st 2018 Journal website: http://www.jfqhc.comCladosporium spp., Alternaria spp., Humicola spp., Mucoraceae spp., Acremonium spp., Penicillium spp., and Aspergillus spp.The results of Petrović et al. (2014) on Serbian bee pollen samples showed presence of Acremonium, Alternaria, Aspergillus, Cladosporium, Epiccocum, Fusarium, Mucor, Penicillium, and Rhizopus ranging from 10 3 to 10 5 CFU/g which is comparable to our results.

Conclusion
It can be concluded that the fresh bee pollen analysed in this survey had poor microbial quality compared to the dried bee pollen.There is also need to adopt appropriate manufacturing practices to prevent possible contamination by equipment or handling of Bulgarian bee pollen.It would be imperative in the future to do additional investigations for precise identification of isolated Enterobacteriaceae from dried bee pollen.Also, effective monitoring and control is necessary for the safety of bee pollen intended for consumption in Bulgaria.

Table 1 :
Physicochemical characteristics of Bulgarian fresh and dried bee pollen samples

Table 2 :
Bacterial contamination of Bulgarian fresh and dried bee pollen samples

Table 3 :
Fungal contamination in Bulgarian fresh and dried bee pollen