**4. Pollen microscopical analysis**

market unbalances. So, Brazil has the best potential to produce honey in the world with low

**Figure 1.** Brazilian honey production (tons). Data compiled for ABEMEL with information from: aliceweb.gov.br.

**Figure 2.** Brazilian honey production and exportation. Data compiled for ABEMEL with information from: aliceweb.

**Figure 3.** Brazilian honey exportation (tons). Data compiled for ABEMEL with information from: aliceweb.gov.br.

**Figure 4.** Brazilian honey per capita consumption (g)/year. Data compiled for ABEMEL with information from: aliceweb.

gov.br.

146 Honey Analysis

gov.br.

risk. And can produce a very good and quality honey with organic certification.

Microscopical analysis of pollen from bee products can offer several interesting information, as geographical source of the material (honey and propolis), botanical origin [25] and also can help about investigations involving contamination, yeast content (fermentation), dust, microscopic particles and others. In this last case, i.e. when the analysis is more complex and involve contamination investigation, this analysis is called palynological analysis [26].

Geographical origin and botanical source usually can be determined when pollen has not been completely removed by a technological process by filtration. Besides, in several countries pollen determination is not a requirement of quality; in Brazil, this point is requested by Normative Instruction no. 11, 2000 [7], and European Community is using a lot of morphological or DNA analysis in order to validate botanical or geographical source, besides OGM material (DNA analysis for this last one). Although this point is not a quality requirement for several countries, it can be used to confirm the geographical and botanical source, especially when some doubts appear. The pollen identification can be carried out using very simple and classical methods as microscopical morphological identification or using more advanced technologies as "DNA barcoding" [27]. The micromorphological analysis is very useful and the analysis can involve identification, as far as possible, of all pollen grains in the sediment, after properly preparation of the sample. The results can be expressed as an (i) estimate value, (ii) determination of frequency classes, and the (iii) count expressed in percentage. For the first case, it is necessary to count around 100 grains and elements correspondent, in the second, around 200–300 pollen grains, in this case, if the pollen is of only a few species, around 200 pollen grains is enough, and finally, in the last case, the presentation of the frequencies as percentage is possible counting around 1200 pollen grains, with two slides counted [25].

When the honey is classified according to plant source, the common name or botanical name is written with word "honey" (CODEX STAN 12‐1981) [8]. The MAPA use classical methods as the reference and the results are compared with the literature. The São Paulo's state government has a databank with more than 17,000 slides, but the access it is only in loco (http://botanica. sp.gov.br/palinologia/palinologia‐colecao‐cientifica‐palinoteca/). Nevertheless, there is electronics databank available, as picture bank of Universidade de São Paulo (http://www.lea.esalq.usp. br/polen/) [28].

The pollen analysis also used to classify the honey as monofloral or unifloral, when the dominance of pollen of a single plant species, the bifloral dominance of pollen of only two plant species and plurifloral or heterofloral with no dominance of pollen of any plant species. Dominant pollen occurs in honey sediments above 45%, at least 300 pollen grains counted. This kind of classification is commercially important because monofloral honey is the most valued since it keeps the same physicochemical and organoleptic characteristics [26].

Despite the facility of preparation of slides in the traditional method, the interpretations of results and time involved with pollen grain counting sometimes is a challenge, in this way molecular tools could be applied. The "DNA barcoding" could be used to identify source plants in the honey. In this method, a short sequence of the DNA of the standardized portions of the genome is used and the results are compared with a reference database, as the GenBank [27]. DNA markers, such as nuclear 18S rDNA, the plastid trnL gene, plasmid coding regions rbcL and matK, trnH‐psbA and ITS2, were used to test their ability to identify plant traces from different honey samples, and [27] suggested that the rbcL region and the trnH‐psbA spacer could be considered to establish the origin, quality, and safety of honey with DNA barcoding, since besides more studies are necessary the stakeholder was established. In order to exemplify the microscopically analysis of pollen in Brazilian honey samples, our group evaluated five samples, including two samples of orange honey, one sample of plurifloral honey, one of "cipó‐uva" honey, and a sample identified by beekeepers as "coffee" honey, that in fact is a plurifloral one, since only a very few amount of coffee pollen was found in the sample. **Figure 5** shows some pollen identified in the honey samples evaluated, and **Figure 6** shows the microscopical image of the pollen obtained in two increases 20 and 40×, usual way to count pollen grain on honey samples (for sample preparation, see [25]).

**Figure 5.** (A, B) Orange pollen, *Citrus* sp. (Rutaceae); (C) *Mimosa* sp. (Mimosaceae); (D) Coffee pollen, *Coffea arabica* (Rubiaceae); (E) *Alternanthera* sp. (Amaranthaceae); (F) "Cipó‐uva" pollen, *Serjania* sp. (Sapindaceae); (G) "Vassourinha‐ do‐campo pollen," *Baccharis* sp. (Asteraceae), and (H) Melastomataceae. All slides were viewed with a Carl Zeiss (Jena, Germany) microscope using the 100× magnification oil immersion objective. Phase contrast brightfield was taken with an AxioCam camera (Carl Zeiss). Images were processed using the AxioVision software version 3.1 and saved as TIFF files. Photographs were taken by Nathália U. Ferreira and Thaila F. dos Reis.

of the genome is used and the results are compared with a reference database, as the GenBank [27]. DNA markers, such as nuclear 18S rDNA, the plastid trnL gene, plasmid coding regions rbcL and matK, trnH‐psbA and ITS2, were used to test their ability to identify plant traces from different honey samples, and [27] suggested that the rbcL region and the trnH‐psbA spacer could be considered to establish the origin, quality, and safety of honey with DNA barcoding, since besides more studies are necessary the stakeholder was established. In order to exemplify the microscopically analysis of pollen in Brazilian honey samples, our group evaluated five samples, including two samples of orange honey, one sample of plurifloral honey, one of "cipó‐uva" honey, and a sample identified by beekeepers as "coffee" honey, that in fact is a plurifloral one, since only a very few amount of coffee pollen was found in the sample. **Figure 5** shows some pollen identified in the honey samples evaluated, and **Figure 6** shows the microscopical image of the pollen obtained in two increases 20 and 40×, usual way

**Figure 5.** (A, B) Orange pollen, *Citrus* sp. (Rutaceae); (C) *Mimosa* sp. (Mimosaceae); (D) Coffee pollen, *Coffea arabica* (Rubiaceae); (E) *Alternanthera* sp. (Amaranthaceae); (F) "Cipó‐uva" pollen, *Serjania* sp. (Sapindaceae); (G) "Vassourinha‐ do‐campo pollen," *Baccharis* sp. (Asteraceae), and (H) Melastomataceae. All slides were viewed with a Carl Zeiss (Jena, Germany) microscope using the 100× magnification oil immersion objective. Phase contrast brightfield was taken with an AxioCam camera (Carl Zeiss). Images were processed using the AxioVision software version 3.1 and saved as TIFF

files. Photographs were taken by Nathália U. Ferreira and Thaila F. dos Reis.

to count pollen grain on honey samples (for sample preparation, see [25]).

148 Honey Analysis

**Figure 6.** Microscopical analysis of honey samples obtained from different geographic and botanical areas. (A, B) Plurifloral honey, (C, D) orange (*Citrus* spp.) honey, (E, F) coffee (*Coffea arabica*) honey, and (G, H) "Cipó‐uva" honey (*Serjania* spp.) (20 and 40× increase, respectively). All slides were viewed with a Carl Zeiss (Jena, Germany) microscope using the 40× magnification oil immersion objective or 20× lens. Phase contrast bright field was taken with an AxioCam camera (Carl Zeiss). Images were processed using the AxioVision software version 3.1 and saved as TIFF files. Photographs were taken by Thaila F. dos Reis.
