Because honey is a saturated solution of sugar obtained from nectar which undergoes dehydration of up to 15 to 18% humidity, it is considered a naturally safe food against microorganisms. The high concentration of carbohydrates, as an antimicrobial factor, limits the amount and activity of water to microorganisms, causing a low reduction potential and high viscosity that limits the entry of oxygen into honey. These factors contribute to the stability of honey, which prevents microorganisms from developing and multiplying, resulting in senescence until their mortality (Molen et al., 1997; Wen et al., 2017). Therefore, only a limited number of microorganisms are to expected survive in honey. Yeasts, fungi, and such spore-forming bacteria as
However, many countries have their own regulations, which generate some divergences in these parameters (Silva et al., 2016; Pascual-Maté et al., 2018). Although the microbiological analysis of honey is not necessary as routine, it improves the product's food safety and helps to diagnose possible hygienic-sanitary flaws in colony management and product handling. Thus, importance of Brazil's honey production and increased exports, its commercial honey is governed by the laws of the Southern Common Market (MERCOSUR), which accepts a maximum of 100 CFU/g for fungi and yeast and does not allow the presence of
The western region of Paraná in the year 2016 produced 705 tons, which represents 15.6% of state production (IBGE, 2018). Because beekeeping is important in this region, the beekeepers working together with Beekeeping Cooperative of the West Coast of Paraná, in partnership with the University State of the West of Paraná, ITAIPU-Binational, have been awarded the Indication of Origin (IP) seal, a type of Geographical Indication (GI), with the National Institute of Intellectual Property (INPI, 2017). These institutions and others that have joined the group have been working to obtain, another GI, the origin denomination quality label for honey produced in this region through research that proves its quality and relate the characteristics of the product to its phytogeographic origin (Camargo et al., 2014; Jacobus et al., 2019). In this study, we characterize the microbiological situation of the honey samples to assess whether beekeepers are applying and following food safety rules. Then, Non-metric Multidimensional Scaling analysis was applied to establish multifactorial relationships and observe the behavior of sample clusters from the region.
Sixty-seven (n = 67) honey samples from
Location map of the study area: (right) map showing the position of the State of Paraná in Brazil, (left) map showing the position of western region of Paraná, with emphasis on sampling municipalities. 1 - Corbélia, 2 - Diamante do Oeste, 3 - Entre Rios do Oeste, 4 - Francisco Alves, 5 - Itaipulândia, 6 -Marechal Cândido Rondon, 7 - Matelândia, 8 - Missal, 9 - Palotina, 10 - Ramilândia, 11 - Santa Helena, 12 - São José das Palmeiras, 13 – Toledo, 14 - Terra Roxa, as well as the Paraná River Basin III, which borders Paraguay (15).
The sample extraction process was carried out either at the producers’ own honey houses or at the cooperative extraction unit, both following international standards due to the export process and the honey origin denomination seal. All samples separate according to producer and period were collected using sterile 500 mL vials. Each evaluated honey sample represents a sample composed of an apiary. The samples were kept at room temperature without be pasteurized, filtrated or mixed with those of different apiaries and later were delivered to the cooperative. They were collected weekly, sent to the laboratory and stored at 6 to 8ºC, because of the tropical climate, until the time of analysis.
Twenty-five grams of each honey sample was homogenized with peptone water 0,1% and decimal dilutions were made 10−1, 10−2, and 10−3. All evaluations were performed according to Silva, Junqueira, & Si (1997) and Sereia et al. (2017). The coliforms were analyzed at 35°C in broth bright green bile lactose 2%, incubated at 35°C for 24 to 48 hours, of fecal coliforms (45°C) and
The values low and high, median, percentiles, quartiles, the mean and standard deviation for the microbiological communities were calculated. Variable data were analyzed by multivariate factor analysis. The Non-metric Multidimensional Scaling (NMDS) was used to test the differences in the parameters among the different honey samples. Subsequently, the Euclidean distance was used with the normalized data and the command “metaMDS” to generate random and interactive processes and to find the best possible solution. The value of the measured NMDS adjustment was evaluated by the stress value. The cophenetic correlation coefficient was calculated to estimate the fit between the dissimilarity matrix and the generated dendrogram (Estevinho et al., 2016). Statistical analyses were performed with statistical software R, version 3.0.2.
Microbiological analyses showed that all samples presented values within the range determined by the RDC-12 technical regulation of 2001. The microbial counts in the samples are shown in Tab. 1. In general, the mean incidence of aerobic mesophilic bacteria, yeasts, and fungi was lower than that reported by other studies (Tab. 1). The total count of aerobic mesophilic had a mean value of 2.52 log CFU/g (Tab. 1), ranging from 0 to 4.4 log CFU/g among samples from different locations (Tab. 2).
Microbial analyses of the 67 samples of honey from the west State of Paraná, Southern Brazil
Parameters | % | Low | High | Median | Q1a | Q3b | Mean ±SDc |
---|---|---|---|---|---|---|---|
Aerobic mesophiles d | 70 | 0 | 4.48 | 2.70 | 0 | 4.40 | 2.52 (±1.85) |
77 | 0 | 3.40 | 1.30 | 1.00 | 2.60 | 1.46 (±1.05) | |
Total coliforms d | 60 | 0 | 2.32 | 0.96 | 0 | 1.30 | 0.78 (±0.66) |
Fecal coliformsd | 62 | 0 | 1.81 | 0.48 | 0 | 0.85 | 0.50 (±0.47) |
Total yeasts d | 34 | 0 | 2.15 | 0 | 0 | 1.00 | 0.46 (±0.68) |
Total molds d | 83 | 0 | 9.00 | 2.85 | 1.00 | 4.62 | 2.95 (±2.44) |
% = percentage of incidence of microorganisms in 67 honey samples evaluated.
Q1: 25% of the samples are between medians and quartiles1.
Q3: 75% of the samples are between the median and quartiles3.
Mean and standard deviation (SD).
Colony forming unit per gram of honey (log CFU/g).
The mean values of microbial analyses of the 67 samples of honey from the beekeepers in 14 municipalities in the west State of Paraná, Southern Brazil
Location | Aerobic mesophiles | Total coliforms | Fecal coliforms | Total yeasts | Total molds | |
---|---|---|---|---|---|---|
Santa Helena (n=27) | 2.6 | 1.5 | 0.5 | 3.1 | 1.0 | 0.6 |
Missal (n=7) | 1.6 | 1.5 | 0.3 | 4.5 | 1.0 | 0.5 |
Terra Roxa (7) | 3.4 | 1.3 | 0.5 | 1.7 | 1.0 | 0.1 |
Marechal Cândido Rondon (n=7) | 3.2 | 1.7 | 0.3 | 0.7 | 0.3 | 0.4 |
Diamante do Oeste (n=4) | 0.5 | 2.2 | 1.2 | 3.5 | 0.4 | 0.7 |
Entre Rios do Oeste (n=3) | 4.4 | 1.1 | 0.7 | 5.7 | 0.9 | 0.7 |
Matelândia (n=3) | 2.3 | 1.5 | 0.3 | 2.5 | 0.4 | 1.1 |
Corbélia (n=2) | 1.2 | 1.2 | 1.1 | 1.0 | 0.0 | 0.4 |
Toledo (n=2) | 1.4 | 1.2 | 0.0 | 5.6 | 0.6 | 0.4 |
Francisco Alves (n=1) | 4.4 | 1.0 | 0.0 | 1.0 | 0.0 | 0.0 |
Itaipulândia (n=1) | 4.4 | 0.0 | 0.0 | 2.0 | 1.4 | 0.0 |
Palotina (n=1) | 0.0 | 1.0 | 0.0 | 8.9 | 1.2 | 0.0 |
Ramilândia (n=1) | 4.4 | 1.3 | 0.0 | 4.8 | 1.2 | 0.0 |
São José das palmeiras (n=1) | 0.0 | 1.7 | 1.7 | 1.0 | 0.0 | 0.8 |
Colony forming unit per gram of honey (log CFU/g).
The total yeast and mold counts, present mean values of 0.46 log CFU/g and 2.95 log CFU/g suggest adequate management by the great majority of beekeepers (Tab. 1). However, the Entre Rios do Oeste and Itaipulândia samples had a mean between 7.1 and 8.9 log CFU/g for molds. Regarding the sanitary quality of the samples, was a reduced incidence of
Table 3. Shows the genera of isolated fungi. The genus that presented the highest incidence was
Genera of molds identified in honey samples from the west of the State of Paraná, Southern Brazil
Parameters | % | Low | High | Median | Q1a | Q3b | Mean ±SDc |
---|---|---|---|---|---|---|---|
22 | 0 | 5.00 | 0 | 0 | 0 | 0.35 (±0.83) | |
31 | 0 | 3.30 | 0 | 0 | 1.00 | 0.52 (±0.88) | |
77 | 0 | 5.30 | 1.00 | 1.00 | 1.57 | 1.20 (±1.03) | |
49 | 0 | 3.30 | 0 | 0 | 1 | 0.71(±0.87) | |
13 | 0 | 3.24 | 0 | 0 | 0 | 0.16 (±0.50) |
% = percentage of occurrence of molds in 67 samples evaluated.
Q1: 25% of the samples are between medians and quartiles1.
Q3: 75% of the samples are between the median and quartiles3.
Mean and standard deviation (SD).
Colony forming unit per gram of honey (log CFU/g).
NMDS analysis visualized the formation of a general structure of the bacterial groups in the honey samples and indicated a similarity between them (Fig. 2). A e stress value of 0.154 indicated that the distance between the points in the ordering chart is representative of the degree of similarity between the samples and the bacterial communities. Samples from the upper left quadrant had higher values for molds, yeasts,
Shows Non-metric Multidimensional Scaling (NMDS) for the similarity of the microorganisms of 67 honey samples. Each number refers to a sample and the acronyms are the different microbial communities which generates a chord distance matrix with a stress value of 0.154.
However, NMDS was not sufficient to distinguish whether the separation of the cluster was due to some microbial communities showing a degree of overlap. Thus, the application of the method of the width of the silhouette determines the optimal point for the formation of two groups (Fig. 3). Occurring a cophenetic correlation coefficient (CCC) of 0.76 between the original similarity matrix and the resulting matrix of similarity. Estevinho et al. (2016) believes this CPCC value is reasonable as a representative factor of the hierarchy.
The bar chart shows the average width of the silhouette for k = 2 groups for 67 samples.
After the ordination of the NMDS graph and the application of the silhouette-width method, the formation of two groups was determined. Together with the clustering analyses (UPGMA), using the Euclidean distance, the difference of the groups between the samples was determined more accurately.
The new NMDS projection (Fig. 4) demonstrated that despite the large projection area, there was a similarity between the microbiological parameters. Although the profiles were visually separable, the samples were separated into two groups. The largest group presented a group of sixty-four samples covering all the microbiological parameters except for total coliforms, which were grouped in the group within the circle, since these three samples presented the highest values of total coliforms with the other parameters.
UPGMA clustering of a string distance matrix between two groups on an NMDS ordering chart for microbiological data of 67 honey samples (CC = 0.76).
Honey is a saturated solution of sugar, which limits the quantity and activity of water for the microorganism, whose development and multiplication is prevented its, leading to its senescence until its mortality (WEN et al., 2017). Other antimicrobial factors of honey are low protein content, acidity, and reduced oxidation-reduction potential together with the high viscosity that limits the entry of oxygen into the honey (SILVA et al., 2016). This research showed that the honey produced by beekeepers was very similar with some exceptions, which are being corrected with technical monitoring. The observed values of the microbial population of aerobic mesophiles from this study are below those found in the literature. The samples of honey from Benin ranged from 2.25 to 2.46 log CFU/g (Azonwade et al., 2018) and those of organic honey from Portugal around 1.3×102 ± 75.5×101 CFU/g (Estevinho et al., 2012). These values demonstrate that the incidence of these microorganisms occurs even in regions with high sanitary control.
The microorganisms found in this study for aerobic mesophiles, molds, and yeasts are probably related to the primary contamination by the intestinal contents of the bees, the environment of the colony and the plants foraged by them (Anderson et al., 2013; Wen et al., 2017). These same authors affirmed that a microbial flow occurs inside the colony, because they observed the same microorganisms in the flowers, the bee bread, the environment of the hive and the intestine of the bees.
This study indicated the incidence of
The incidence of
The low yeast value demonstrated limited growth of these microorganisms, indicating the good hygienic-sanitary quality of the samples because yeasts are transferred to honey through primary and secondary contamination (Snowdon & Cliver, 1996). Furthermore, physicochemical analyses indicated low moisture content (<20%) of the collected honey, which makes bacterial growth difficult because the above 21% moisture is known to favor the proliferation of these microorganisms which survive in acidic conditions, cause fermentation and the formation of carbon dioxide and alcohol and reduce the quality of honey (Snowdon & Cliver 1996; White, 2010). Sereia et al. (2010) evaluated the hygienic-sanitary quality of organic and non-organic honey samples from the Paraná River islands and observed that organic honey presented lower quality than conventional honey to hygienic-sanitary quality and that the humidity of the samples was the main factor for the reduced product quality.
The incidence of these microorganisms indicated secondary contamination for the sanitary quality (total coliforms and fecal coliforms) and safety (
Total coliforms and fecal coliforms in the literature (Pucciarelli et al., 2014) were present in a greatest variety in honey samples in Argentina, Italy (Sinacori et al., 2014) and Mexico (Vázquez-Quiñones et al., 2018) and absent in the samples evaluated from Spain (Sanz et al., 1995), Argentina (Iurlina & Fritz, 2005), Portugal Estevinho et al. (2012) and Mexico Fernández et al. (2017). Therefore, this is first study on honey in the process of designation of origin in Brazil and provides information for future work to reduce these microorganisms in honey. Increased microbiological values in the locations furthest from the cooperative (Santa Helena municipalities) such as Terra Roxa, Marechal Cândido Rondon, Entre Rios do Oeste, and Francisco Alves may be due to less technical assistance and training of beekeepers. These skills are important as good production practices improve the quality of the honey produced. Evaluation microorganisms at different points in honey processing (honeycombs, honeycomb uncapping, honey extractor, honey sump, and honey drum), demonstrated that in eight honey extraction units, when good manufacturing standards were not followed observed contamination increased per coliforms at all points of evaluation (Fernández et al., 2017). Sereia et al. (2011) concluded the same in the evaluation of organic honey production systems and reported a reduction of molds and yeasts on samples of honey processed adequately to those whose beekeepers did not follow good hygiene practices, by Administrative Rule No. 367 of September 4, 1997 (Brasil, 1997).
The genera of molds identified in this study corroborate the data of the literature, since the genus
Wen et al. (2017) evaluated the microbial load in the first fifteen days of honey maturation and observed that while the ripening process and reduction of the moisture content of honey occurred there was a reduction in the quantity and richness of microorganism species. Besides, these authors pointed out that, after five days of maturity, there was a reduction of the genera
Thus, it was concluded that the majority of the samples were within the parameters accepted by the legislation. The largest microbial communities were of mesophilic aerobes and molds. The analyses showed that honey samples are similar to one another for most microbial communities. These results indicate that honey produced in the western region of Paraná are very similar. The results reported in this study show a sampling of the situation of the bee chain in the region, and with this information it is possible to set up an activity plan to improve technology transfer to beekeepers.