Journal of Apicultural Research
ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/tjar20
Standard methods for pollen research
Maria G. Campos, Ofélia Anjos, Manuel Chica, Pascual Campoy, Janka
Nozkova, Norma Almaraz-Abarca, Lidia M. R. C. Barreto, João Carlos
Nordi, Leticia M. Estevinho, Ananias Pascoal, Vanessa Branco Paula, Altino
Chopina, Luis G. Dias, Živoslav L. j. Tešić, Mirjana D. Mosić, Aleksandar
Ž. Kostić, Mirjana B. Pešić, Dušanka M. Milojković-Opsenica, Wiebke
Sickel, Markus J. Ankenbrand, Gudrun Grimmer, Ingolf Steffan-Dewenter,
Alexander Keller, Frank Förster, Chrysoula H. Tananaki, Vasilios Liolios,
Dimitrios Kanelis, Maria-Anna Rodopoulou, Andreas Thrasyvoulou, Luísa
Paulo, Christina Kast, Matteo A. Lucchetti, Gaëtan Glauser, Olena Lokutova,
Ligia Bicudo de Almeida-Muradian, Teresa Szczęsna & Norman L. Carreck
To cite this article: Maria G. Campos, Ofélia Anjos, Manuel Chica, Pascual Campoy, Janka
Nozkova, Norma Almaraz-Abarca, Lidia M. R. C. Barreto, João Carlos Nordi, Leticia M. Estevinho,
Ananias Pascoal, Vanessa Branco Paula, Altino Chopina, Luis G. Dias, Živoslav L. j. Tešić,
Mirjana D. Mosić, Aleksandar Ž. Kostić, Mirjana B. Pešić, Dušanka M. Milojković-Opsenica,
Wiebke Sickel, Markus J. Ankenbrand, Gudrun Grimmer, Ingolf Steffan-Dewenter, Alexander
Keller, Frank Förster, Chrysoula H. Tananaki, Vasilios Liolios, Dimitrios Kanelis, Maria-Anna
Rodopoulou, Andreas Thrasyvoulou, Luísa Paulo, Christina Kast, Matteo A. Lucchetti, Gaëtan
Glauser, Olena Lokutova, Ligia Bicudo de Almeida-Muradian, Teresa Szczęsna & Norman L.
Carreck (2021) Standard methods for pollen research, Journal of Apicultural Research, 60:4, 1-109,
DOI: 10.1080/00218839.2021.1948240
To link to this article:  https://doi.org/10.1080/00218839.2021.1948240
Published online: 07 Sep 2021.
Submit your article to this journal 
Article views: 799
View related articles 
View Crossmark data
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=tjar20
Citing articles: 1 View citing articles 
REVIEW ARTICLE
Standard methods for pollen research
Maria G. Camposa
 , Of
elia Anjosb,c , Manuel Chicad , Pascual Campoye , Janka Nozkovaf,
Norma Almaraz-Abarcag , Lidia M. R. C. Barretoh, Jo~ao Carlos Nordii, Leticia M. Estevinhoi , Ananias Pascoali ,
Vanessa Branco Paulai, Altino Chopinai, Luis G. Diasi , Zivoslav L. j. Tesi
cj , Mirjana D. Mosi
cj,
Aleksandar Z. Kosti
ck , Mirjana B. Pesi
ck , Dusanka M. Milojkovi
c-Opsenicaj , Wiebke Sickell ,
Markus J. Ankenbrandm , Gudrun Grimmern, Ingolf Steffan-Dewenterm , Alexander Kellero , Frank F€orsterp
, Chrysoula H. Tananakiq , Vasilios Lioliosp , Dimitrios Kanelisq , Maria-Anna Rodopoulouq , Andreas
Thrasyvoulouq , Lu
ısa Pauloq, Christina Kasts , Matteo A. Lucchettis, Ga€etan Glausert, Olena Lokutovau,
Ligia Bicudo de Almeida-Muradianv , Teresa SzczeRsnaw and Norman L. Carreckx,y
aCoimbra Chemistry Centre (CQC, FCT Unit 313) (FCTUC) & Observatory of Drug-herb Interactions, University of Coimbra, Coimbra,
Portugal; bIPCB– Instituto Polit
ecnico de Castelo Branco, Castelo Branco, Portugal; cCEF, Instituto Superior de Agronomia, Universidade de
Lisboa, Tapada da Ajuda, 1349-017, Lisboa, Portugal; dAndalusian Research Institute DaSCI “Data Science and Computational
Intelligence”, University of Granada, Granada, Spain; eAutomatics and Robotics Center, Universidad Polit
ecnica de Madrid, Madrid, Spain;
fCentro Interdisciplinario de Investigaci
on para el Desarrollo Integral Regional unidad Durango, Instituto Polit
ecnico, Nacional, M
exico;
gInterdisciplinary Research Center for the Integral Regional Development-Durango, National Polytechnic Institute, Mexico City, Mexico;
hApicultural Center of Research of Taubat
e University, UNITAU - Universidade de Taubat
e-SP, Taubat
e, Brazil; iCentro de Investigac¸~ao de
Montanha (CIMO), Instituto Polit
ecnico de Braganc¸a, Braganc¸a, Portugal; jFaculty of Chemistry, Department of Analytical Chemistry,
University of Belgrade, Belgrade, Serbia; kFaculty of Agriculture, Department of Chemistry and Biochemistry, University of Belgrade,
Belgrade, Serbia; lTh€unen Institute, Institute of Biodiversity, Braunschweig, Germany; mDZHI, Universit€atsklinikum W€urzburg, W€urzburg,
Germany; nDepartment of Animal Ecology and Tropical Biology, Biocenter, University of W€urzburg, W€urzburg, Germany; oCenter for
Computational and Theoretical Biology & Department of Bioinformatics, Biocenter, University of W€urzburg, W€urzburg, Germany;
pBioinformatics & Systems Biology, Heinrich-Buff-Ring 58, Justus-Liebig-Universit€at, Gießen, Germany; qLaboratory of Apiculture-Sericulture,
School of Agriculture, Aristotle University of Thessaloniki, Thessaloniki, Greece; rCATAA- Associac¸~ao Centro de Apoio Tecnol
ogico Agro-
Alimentar de Castelo Branco, Castelo Branco, Portugal; sAgroscope, Swiss Bee Research Centre, Bern, Switzerland; tNeucha^tel Platform of
Analytical Chemistry, University of Neucha^tel, Neucha^tel, Switzerland; uThe National Institute of Horticultural Research Konstytucji 3 Maja
1/3, 96 - 100, Skierniewice, Poland; vPharmaceutical Science School, University of S~ao Paulo, S~ao Paulo, Brazil; wThe National Institute of
Horticultural Research Konstytucji 3 Maja 1/3, 96 - 100, Skierniewice, Poland; xCarreck Consultancy Ltd. Woodside Cottage, Shipley, West
Sussex, UK; yLaboratory of Apiculture and Social Insects, School of Life Sciences, University of Sussex, Brighton, East Sussex, UK
(Received 20 March 2020; accepted 22 June 2021)
“Bee pollen” is pollen collected from flowers by honey bees. It is used by the bees to nourish themselves, mainly by provid-
ing royal jelly and brood food, but it is also used for human nutrition. For the latter purpose, it is collected at the hive
entrance as pellets that the bees bring to the hive. Bee pollen has diverse bioactivities, and thus has been used as a health
food, and even as medication in some countries. In this paper, we provide standard methods for carrying out research on
bee pollen. First, we introduce a method for the production and storage of bee pollen which assures quality of the product.
Routine methods are then provided for the identification of the pollen’s floral sources, and determination of the more
important quality criteria such as water content and content of proteins, carbohydrates, fatty acids, vitamins, alkaloids, phen-
olic and polyphenolic compounds. Finally, methods are described for the determination of some important bioactivities of
bee pollen such as its antioxidant, anti-inflammatory, antimicrobial and antimutagenic properties.
M
etodos est
andar Para la investigaci
on del polen
El "polen de abeja" es el polen recogido de las flores por las abejas mel
ıferas. El polen de abeja es utilizado para nutrir
a las propias abejas, principalmente para proporcionar jalea real y alimento para las cr
ıas, pero tambi
en se utiliza para la
nutrici
on humana. Para este 
ultimo fin, se recoge en la entrada de la colmena en forma de gr
anulos que las abejas llevan
a la colmena. El polen de abeja tiene diversas bioactividades, por lo que se hautilizado como alimento para la salud, e
incluso como medicamento en algunos pa
ıses. En este art
ıculo, proporcionamos m
etodos est
andar para llevar a cabo
investigaciones sobre el polen de abeja. En primer lugar, presentamos un m
etodo de producci
on y almacenamiento de
polen de abeja que garantiza la calidad del producto. A continuaci
on, se ofrecen m
etodos de rutina para la identificaci
on
de las fuentes florales del polen y la determinaci
on de los criterios de calidad m
as importantes, como el contenido de

Corresponding author. Email: mgcampos@ff.uc.pt and ofelia@ipcb.pt
Please refer to this paper as: Campos, M.G., Anjos, O., Chica, M., Campoy, P., Nozkova, J., Almaraz-Abarca, N., Barreto, L.M.R.C.,
Nordi, J.C., Estevinho, L.M., Pascoal, A., Paula, V.B., Chopina, A., Dias, L., Tesi
c, Z.L., Mosi
c, M.D., Kosti
c, A.Z., Pesi
c, M.B.,
Milojkovi
c-Opsenica, D.M., Sickel, W., Ankenbrand, M.J., Grimmer, G., Steffan-Dewenter, I., Keller, A., F€orster, F., Tananaki, C.H.,
Liolios, V., Kanelis, D., Rodopoulou, M.A., Thrasyvoulou, A., Paulo, L., Kast, C., Lucchetti, M.A., Glauser, G., Lokutova, O., Bicudo
de Almeida-Muradian, L., SzczeR sna, T., Carreck, N.L. (2021) Standard methods for pollen research. In V. Dietemann, P. Neumann,
N. L. Carreck, & J. D. Ellis (Eds.), The COLOSS BEEBOOK - Volume III: Standard methods for Apis mellifera hive product research.
Journal of Apicultural Research, 60(4). https://doi.org/10.1080/00218839.2021.1948240

 2021 International Bee Research Association
Journal of Apicultural Research, 2021
Vol. 60, No. 4, 1–109, https://doi.org/10.1080/00218839.2021.1948240
agua y de prote
ınas, carbohidratos, 
acidos grasos, vitaminas, alcaloides y compuestos fen
olicos y polifen
olicos. Por

ultimo, se describen m
etodos para la determinaci
on de algunas bioactividades importantes del polen de abeja, como sus
propiedades antioxidantes, antiinflamatorias, antimicrobianas y antimutag
enicas.
花粉研究的标准方法
“蜂花粉”是蜜蜂从花中采集的花粉。首先, 它用于蜜蜂本身的营养, 主要是提供蜂王浆和哺育
幼虫的食物, 其次也用于人类营养。对于后一目的, 花粉作为蜜蜂带到蜂巢的颗粒, 在蜂巢入口
处被收集。蜂花粉具有多种生物活性, 因此被用作健康食品, 甚至在一些国家被用作药物。本
文提供了研究蜂花粉的标准方法。首先, 我们介绍了一种生产和储存蜂花粉的方法, 以保证产
品的质量。然后提供常规鉴定方法来鉴定花粉源及更重要的质量标准, 如水含量和蛋白质、
碳水化合物、脂肪酸、维生素、生物碱、酚类和多酚类化合物的含量等。最后, 描述了测定
蜂花粉的一些重要生物活性的方法, 如其抗氧化、抗炎、抗菌和抗诱变特性。
Keywords: pollen; harvest; storage; multi-classification; computer; vision; corbicula; proteins; lipids; vitamins;
antioxidant; minerals; toxicity; alkaloids; UHPLC-HRMS; phenolic; flavonoids; ABTS; FTIR-ATR; antioxidant; free radical
scavenging activity; ORAC; DPPH
; antimicrobial; antimutagenic; COLOSS; BEEBOOK
Table of contents
Page
No.
1. Introduction 8
2. Producing bee pollen from
harvest to storage
9
2.1. Production and handling of bee
pollen before processing
10
2.1.1. Quality control in the apiary 10
2.1.1.1. Smoke 10
2.1.1.2. bee hives 11
2.1.1.3. Clothing 11
2.1.1.4. Tools 11
2.1.1.5. Inputs 11
2.1.1.6. Handling 11
2.1.2. Types of pollen traps 11
2.1.3. Chosen area of apiary production 12
2.1.4. Quality control in the collection 12
2.1.5. Quality control in the transport of
raw material
12
2.2. Handling of bee pollen in the
processing sector
12
2.2.1. Flowchart from production to processing
of pollen
13
2.2.2. Processing bee pollen 13
Table of contents
Page
No.
2.3. Materials and equipment involved
in the process
13
3. Inexpensive pollen load
authenticationbymeansof computer
vision and classification algorithms
14
3.1. General overview of the method 15
3.2. Description of the computational
techniques
15
3.2.1. Segmentation algorithm 15
3.2.2. Homogenizing the pollen loads of the
image by mean shift filtering
15
3.2.3. One-class multi-classification algorithm based
on distance
17
3.2.4. Multi-classification confidence and
ambiguity discovery process
18
3.3. Experimentation 18
3.3.1. Experimental setup: equipment and data 18
3.3.2. Image processing results 19
3.3.3. Numerical performance of the
multi-classification algorithms
20
3.3.4. Validation of the complete prototype 20
3.4. Concluding remarks 21
4. Evaluation and classification of
corbicular pollen morphology
21
4.1. Characteristics of corbicular
pollen samples
22
4.2. Methodology for assessing the
morphometric characters of
corbicular pollen
22
2 M. G. Campos et al.
Table of contents
Page
No.
4.2.1. Methodological procedures for evaluating
the quantitative characteristics
23
4.2.2. Methodological procedures for evaluating
the qualitative characteristics
24
4.2.3. List of descriptors and classification 24
4.2.3.1. General structure of descriptors list 24
4.2.3.2. Passports descriptors 25
4.2.3.3. General structure of characterization
and evaluation descriptors
26
4.2.3.4. Methods and techniques for creating
and modifying the descriptors
26
5. Identification of the floral origin
by microscopy
28
5.1. Structure of the pollen grain 30
5.2. Floral origin 32
5.3. Features of pollen grains 32
5.3.1. Size 32
5.3.2. Shape 32
5.3.3. Apertures 34
5.3.4. Exine sculpture (exine ornamentation)
and surface
34
5.3.5. Other features 35
5.3.6. Irregular grains 36
5.3.7. Pollen “types” 36
5.4. Colors of pollen 37
5.5. Methods of pollinic pollen analysis 37
5.5.1. Types of microscopy examination 37
5.5.1.1. Light microscope 37
5.5.1.2. Transmission electron
microscope (TEM)
37
5.5.1.3. Scanning electron microscope (SEM)
(three-dimensional image)
37
Table of contents
Page
No.
5.6. Microscopic techniques 38
5.6.1. Preparation of pollen grain slides from
pollen loads for microscopic research
38
5.6.2. Identification of pollen under microscope 38
5.6.3. Description of pollen grains 38
5.7. Palinoindication of environmental
pollution
38
5.8. Concluding remarks 38
6. Phenolic and polyphenolic
profiles as fingerprints of floral
origin of bee pollen
39
6.1. Materials and methods 40
6.1.1. Sample management 40
6.1.2. Sample collection 40
6.1.2.1. Pollen hand collected – reference
samples
40
6.1.2.2. Bee pollen pellets/loads 40
6.1.3. Methods 40
6.1.3.1. HPLC/DAD analyses 40
6.1.3.1.1. Sample preparation (hand collected
pollen and/or pollen from herbarium)
40
6.1.3.1.2. Sample preparation (bee pollen) 40
6.1.3.1.3. HPLC/DAD analysis 40
6.2. Theoretical rules for structural
determination of the main phenolic
acids and flavonoids
41
6.3. Quantification of the major floral
specie (floral source) in a mixture
43
7. Determination of phenolic profile
of bee pollen extracts by LC and
LC/MS
44
7.1. Polyphenolics: importance,
structure and classification
44
7.2. Preparation of pollen extracts for
determination of phenolic compounds
44
7.2.1. Extraction of phenolic compounds from
pollen – general procedure
47
7.2.2. Preparation of enzymatic hydrolysates
from pollen
47
Standard methods for pollen research 3
Table of contents
Page
No.
7.2.2.1. Preparation of PEP 47
7.2.2.2. Preparation of EEPP 49
7.2.2.3. Preparation of feruloyl esterase
hydrolysate from pollen’s exine
49
7.2.3. Application of SPE for pre-concentration
of phenolic compounds for further LC
and LC/MS determination
50
7.2.3.1. Application of SPE for
pre-concentration of phenolic
compounds from pollen
52
7.3. LC and LC/MS techniques 52
7.3.1. LC techniques 52
7.3.1.1. LC-DAD technique 52
7.3.1.2. LC-MS, MS/MS techniques 52
7.3.2. UHPLC-HESI-MS/MS technique for
determination of phenolic compounds
53
8. Standard method for
identification of bee pollen
mixtures through meta-barcoding
54
8.1. Meta-barcoding protocol 54
8.1.1. Required materials 54
8.1.1.1. Reagents 54
8.1.1.2. Laboratory equipment 54
8.1.2. Pollen acquisition 54
8.1.3. Laboratory workflow 54
8.1.3.1. DNA extraction 54
8.1.3.2. Amplification 55
8.1.3.3. Normalisation 58
8.1.3.4. Quality control and quantification 58
8.1.3.5. Quantification 59
8.1.4. Sequencing 59
8.1.4.1. Sample library 59
8.1.4.2. PhiX control 59
Table of contents
Page
No.
8.1.4.3. Preparing reagent cassette and loading
the sample
59
8.2. Bioinformatics 60
8.2.1. Required software 60
8.2.2. Classification 60
8.2.2.1. Reference databases 60
8.2.2.2. Preparation and classification of
sequencing data
61
8.3. Data analysis 63
8.3.1. Required software 63
8.3.2. Prepare sample meta-data 63
8.3.3. Importing data 63
8.3.4. Recommended packages for further
analysis
65
9. Standardmethod for the
determination of protein content of
pollen using the Kjeldahl method
65
9.1. Introduction 65
9.2. Determination of protein content
by Kjeldahl method
65
9.3. Development of the method 65
9.3.1. Volume of NaOH solution 65
9.3.2. Volume of H3BO3 solution 66
9.3.3. Distillation time 66
9.4. Sample preparation 66
9.5. Validation of the method 66
9.6. Application of the method 66
10. Standard method for
determination of lipid content of
pollen using Soxhlet extraction
66
10.1. Introduction 66
10.2. Methodology for total lipid content
determination
66
10.3. Development of the method 67
4 M. G. Campos et al.
Table of contents
Page
No.
10.4. Validation of the method 67
10.5. Application of the method 67
11. Determination of sugar content
in pollen using HPLC-RID
67
11.1. Introduction 67
11.2. Method for identification and
quantification of sugars by HPLC-RID
70
11.2.1. Principle 70
11.2.2. Reagents 70
11.2.3. Equipment 70
11.2.4. Preparation of stock standard solution
and working standard solutions of sugars
70
11.2.5. Sample preparation 70
11.2.6. HPLC-RID analysis 70
11.3. Validation of the method 71
11.4. Application of the method 71
12. Standard method for vitamin
analysis in bee pollen
71
12.1. Introduction 71
12.2. B complex vitamins 71
12.2.1. Methods for B complex vitamins 71
12.2.1.1. Simultaneous extraction of vitamins
B1, B2, vitamers from vitamin B6
and niacin
72
12.2.1.2. Chromatographic conditions for the
analysis of the B complex vitamins in
dried bee pollen
72
12.2.1.3. Determination of thiamin (vitamin B1),
with pre-column reaction
72
12.2.1.4. Determination of riboflavin
(vitamin B2)
72
12.2.1.5. Determination of vitamin PP through
niacin and niacinamide vitamers
72
Table of contents
Page
No.
12.2.1.6. Determination of vitamin B6 through
the vitamers pyridoxol, pyridoxal and
pyridoxamine
72
12.3. Methods recommended to analyze
antioxidant vitamins of bee pollen
72
12.3.1. Determination of vitamin C 72
12.3.2. Determination of vitamin E 72
12.3.3. Determination of b-carotene
(provitamin A)
72
12.4. Concluding remarks 73
13. Standard methods for mineral
and trace elements
73
13.1. Introduction 73
13.2. Material and methods 73
13.2.1. Sample preparation 73
13.2.2. Methods of analysis 74
13.2.2.1. Flame and electrothermal atomic
absorption spectrometry
74
13.2.2.2. Inductively coupled plasma atomic
emission spectroscopy
(ICP-AES) methods
74
13.3. Concluding remarks 74
14. Standard methods for
pyrrolizidine alkaloid profiling
by UHPLC-HRMS
75
14.1. Introduction 75
14.2. Materials and methods 76
14.2.1. Equipment and chemicals 76
14.2.2. Collection of plant pollen from
PA-containing plants
76
14.2.2.1. Pollen collection from E. vulgare 76
14.2.2.1.1. Collection with forceps 76
14.2.2.1.2. Collection with cyclohexane 76
Standard methods for pollen research 5
Table of contents
Page
No.
14.2.3. Pollen collection from E. cannabinum 76
14.3. Extraction of PAs from plant pollen 77
14.4. Profiling of PAs in plant pollen with
UHPLC-HRMS
77
14.5. Data processing 78
14.6. Concluding remarks 79
15. Standard methods for total
phenolic compounds, total
flavonoids and antioxidative
capacity bee pollen evaluation by
FTIR-ATR spectroscopy
80
15.1. Introduction 80
15.2. Materials and methods 85
15.2.1. Sample characterization and preparation 85
15.2.2. Reference analytical methods 85
15.2.2.1. Total phenolic and flavonoids
content analysis
85
15.2.2.2. Antioxidant activity analysis 85
15.2.3. FTIR-ATR collection and data processing 85
15.3. Concluding remarks 85
16. Standardmethods for determination
of antioxidant bioactivity
86
16.1. Introduction 86
16.2. Methods 87
16.2.1. 2,2-diphenyl-1-picrylhydrazyl (- DPPH
)
free radical scavenging activity principle
87
16.2.1.1. Variations in the -DPPH
 assay for
bee pollen
88
16.2.1.2. Material and methods for
-DPPH
 assay
88
16.2.2. Antioxidant capacity using ORAC (oxygen
radical absorbance capacity)
assay principle
88
16.2.2.1. Material and methods for antioxidant
activity (ORAC) assay
89
16.3. Concluding remarks 89
Table of contents
Page
No.
17. Standardmethods for anti-
inflammatory activities in bee pollen
89
17.1. Introduction 89
17.2. Determination of standard methods 89
17.2.1. Anti-inflammatory activities of BP in vitro 89
17.2.1.1. BP extract preparation 89
17.2.1.1.1. Materials and reagents 89
17.2.1.1.2. Procedure 89
17.2.1.2. BP sample preparation 90
17.2.1.2.1. Materials and reagents 90
17.2.1.2.2. Procedure 90
17.2.2. Anti-inflammatory activities of BP in vivo 90
17.2.2.1. BP extract preparation 90
17.2.2.1.1. Materials and reagents 90
17.2.2.1.2. Procedure 90
17.2.3. Animals 90
17.3. Concluding remarks 90
18. Standardmethods for antimicrobial
activities in bee pollen
91
18.1. Introduction 91
18.2. Materials and methods 91
18.2.1. BP sample collection 91
18.2.2. Extract preparation of BP for antimicrobial
analysis
91
18.2.3. Antimicrobial analysis by agar
disk diffusion
92
18.2.3.1. Preparation of inoculum suspension 92
6 M. G. Campos et al.
Table of contents
Page
No.
18.2.4. Antimicrobial analysis by agar
well diffusion
92
18.2.4.1. Preparation of inoculum suspension 92
18.2.5. Antimicrobial analysis by micro-dilution in
microtiter plate (96 wells)
92
18.2.5.1. Preparation of inoculum suspension 92
18.2.5.2. Calculating the lethal concentrations 93
18.3. Concluding remarks 93
19. Determination of standard
methods for anti-mutagenic
activities in bee pollen
94
19.1. Introduction 94
19.2. Determination of standard methods
for anti-mutagenic activities in BP
94
19.2.1. Preparation of BP extracts 94
19.2.2. Anti-mutagenic activities in vitro 94
Table of contents
Page
No.
19.2.2.1. Materials and reagents 94
19.2.2.2. Procedure 94
19.2.3. Anti-mutagenic activities of BP in vivo 95
19.2.3.1. Materials and reagents 95
19.2.3.2. Experimental animals 95
19.2.3.3. Preparation of BP extracts 95
19.3. Concluding remarks on
anti-mutagenic effects of BP
95
Acknowledgements 95
Disclosure statement 96
References 96
Appendix: authors 108
Standard methods for pollen research 7
1. Introduction
Pollen production on an industrial scale can be consid-
ered as a recent development, so there is a need for an
international directive for quality control. In this chap-
ter, the authors propose optimized methodologies
based on the major skills developed in the field which
can be used in the near future to fill that gap. This sec-
tion aims to aid the understanding of actual production
and the development of sanitary standards in the light
of food security, standardized beekeeping management
actions, collection of the product in the bee hives,
transportation, sanitary treatment, processing and stor-
age by agribusiness, resulting in a product with flavor,
very delightful aroma and texture of excellent quality
for food or nutraceutical purposes. To achieve the
above characteristics, we also discuss corbicular pollen
itself in terms of weight, size, color, shape and surface
texture and, unified systems (descriptors) for the evalu-
ation of corbicula morphological parameters of mono-
floral pollen.
Pollen is considered by some researchers to be one
of the naturally available resources that may be
regarded as a perfect food, since it is rich in essential
amino acids, trace elements, enzymes, B-complex and
some C and E vitamins. The pollen grains present a
huge variation of morphological characteristics, estab-
lished by genetic heredity and do not vary depending on
environmental events or changes. In this context, the
pollinic analysis of bee pollen is the best method to
determine the plants used by honey bees, making it pos-
sible, among other inferences, to establish the plant taxa
in the regions used for the pollen forage. The goal is to
provide global information regarding the pollinic analysis
of this matrix and the analytical procedures that are
most frequently applied, from the more laborious and
time consuming to the quicker and more advanced.
Among the last, flavonoid / phenolic acid profiles using a
gradient HPLC/DAD method is described. Given the
specie-specific feature of those profiles obtained with dif-
ferent taxa collected by bees, they represent an import-
ant tool for quality control concerning the botanical
origin of both monofloral and complex mixtures of
bee pollen.
The first step in the determination of phenolic
compounds is their appropriate extraction. The com-
plexity of the structure of pollen grains requires mul-
tiple-stage extraction of polyphenols. Here there will
be explained different procedures for the preparation
of pollen extracts containing phenolic compounds and
proposed a general procedure. Also presented are
the applications of SPE for pre-concentration of the
obtained phenolic fractions which facilitate further
analysis. One of the sections describes LC and LC/MS
techniques used for the identification and quantifica-
tion of phenolic compounds with a detailed proced-
ure for their determination by the UHPLC- HESI-MS/
MS technique.
Another important question to be discussed is the
fact that traditional pollen analysis via light microscopy
has limitations in sample throughput as well as taxo-
nomic resolution. Recently, pollen meta-barcoding
methods have been developed as alternative
approaches, where plant species identification of pollen
grains works via DNA sequencing. However, these cur-
rently utilize different genetic markers and sequencing
platforms lessening study comparability. We here
describe a detailed protocol of the latest development
in this field as a standard method for pollen meta-bar-
coding. It is highly cost-efficient, requires no palyno-
logical knowledge, is performable in standard
laboratories and profits from a well-established refer-
ence database.
For the main macro and micronutrients standard
methods are proposed. Bee pollen has been consid-
ered by some researchers as Mother Natur
es perfect
food by containing nearly all nutrients required by
humans as proteins, sugars, amino acids, lipids, vita-
mins, minerals, nucleic acids, enzymes, phenolics and
many more compounds. Pollen can also be character-
ized as a supplementary food with varied enhancing
effects in human health due to its nutritional proper-
ties. Once pollen was exclusively a source of protein
for bees’ diet, but recently, it has gained further atten-
tion as a potential food source for human consump-
tion too. The Kjeldahl method is widely used for
protein determination. In this section, we examine the
methodology concerning the volumes of the reagents,
as well as the distillation time and the quantity of the
sample, in order to elucidate the most suitable pro-
cedure and a method been validated using bee pollen.
There is increasing knowledge of the nutritional value
of pollen as a supplementary food or medicine. One
species of pollen is different from another, and not
one has the same standard chemical composition, nor
contains all the characteristics attributed to "pollen" in
general. The lipid fraction is one of the main pollen
constituents, and its determination is highly important.
In this chapter, we describe a method for the deter-
mination of lipid content of pollen investigated and
optimized, by minimizing the reagent consumption and
the sample quantity.
Suggested methods to analyze micronutrients are
also presented. Pollen vitamins as well as the literature
found about them are included in the text. Methods for
B-complex (B1, B2, B6 and PP, with vitamers) and anti-
oxidant vitamins (vitamin E, C and carotenes as provita-
min A) are explained and involve HPLC, open column
and titrimetric determinations. The composition of min-
erals in pollen collected by honey bees is dependent on
the plant species visited. The variety of the samples col-
lected is closely associated with the geographical apiary
locality given by the floral sources (taxa) and the envir-
onmental conditions in the area such as soil compos-
ition, climatic conditions and agricultural crops grown
8 M. G. Campos et al.
close to the apiary. Because plants can cause bioremedi-
ation of some pollutants and are part of some other
anthropogenic processes, they should also be consid-
ered as an additional source of minerals in pollen (e.g.,
Cd, Cr, Cu, Fe, Ni, Pb and Zn). Monitoring these min-
erals in this matrix, as well in other food products is
very important as a quality parameter. Despite this, the
publications in the field are scarce. Here we propose,
from the research published on the mineral composition
on pollen, different methodologies that can be car-
ried out.
Regarding toxic compounds, research has been lim-
ited. However, as a starting point, attention has focused
on pyrrolizidinic alkaloids (PA). Many plant species pro-
duce pollen containing PAs that honey bees collect and
bring as a protein source into the bee hive. Methods
for identification of PA in pollen, together with the
extraction and analysis of PAs, are essential tools for
plant and food research. Mainly two approaches are
commonly used: In the first approach, the total PAs (N-
oxides and free-bases) are analyzed as the sum of the
“necine bases” (the basic structures of PAs) by gas
chromatography coupled with mass spectrometry (GC-
MS). The second approach is based on liquid chroma-
tography coupled with mass spectrometric detection
(LC-MS). This method can detect both the free alkaloids
and their N-oxides simultaneously, and has become the
method of choice in recent years. Due to the lack of
standards on the market and the multitude of PAs dis-
covered so far, the detection, identification and quantifi-
cation of PAs still represent a continuous challenge. In
the following chapter, methods for the collection,
extraction and profiling of PA plant pollen from Echium
vulgare L. and Eupatorium cannabinum L. are presented.
In the advanced methods that allow multiple analysis
with a small amount of sample, it will be shown the
potential of new integrated methods as the Fourier
Transform Infrared (FTIR) spectroscopic method with
Attenuated Total Reflectance (ATR) used, for example,
to predict the total content of phenolic and poly-
phenolic compounds. Determination of Free radical
scavenging (EC50) and antioxidative capacity (ABTS) of
bee pollen is also included. FTIR-ATR can be a useful
technique for the quantification of total phenolic com-
pounds, total flavonoids and antioxidant capacity of bee
pollen that could improve the speed of laboratory anal-
yses of this product.
Concerning the bioactivity of pollen, several studies
have been developed in order to elucidate the useful-
ness of pollen, and therapeutic effects have been
reported among others in the treatment of benign pros-
tatitis and for oral desensitization of children who have
an allergy, effects as anti-inflammatory, antimicrobial,
anti-fungicidal, anti-mutagenic and immunomodulatory.
Some studies reported that bee pollen accelerates
mitotic rate, promotes tissue repair, enhances greater
toxic elimination and reduces excessive cholesterol
levels. Pollen has also been used to alleviate or cure
conditions such as colds, flu, ulcers, premature aging,
anemia, colitis and today is increasingly used as a health
food supplement.
For these reasons, bee pollen has attracted espe-
cial attention by researchers throughout the world,
and many scientific publications have been produced.
In the last 17 years, at least 42 reports have been
published about the antioxidant properties from dif-
ferent taxa, evaluating these properties by various
methods. The available information corresponds to
bee pollen of 21 countries including Brazil, Mexico,
New Zealand and Portugal, the countries where most
attention has been dedicated to this topic. The
reports cited in the current work give information
about the antioxidant capabilities of bee pollen com-
ing from around 75 well identified plant species
belonging to approximately 52 families. The 2,2-
diphenyl-1-picrilhydrazyl (DPPH
) assay is the most
popular method to evaluate the free radical scaveng-
ing properties of bee pollen. It is based on the single
electron transfer mechanism, is rapid, easy, accurate,
reproducible, and presents good correlations with
other assays like TEAC and b-carotene bleaching. The
oxygen radical absorbance capacity (ORAC) assay is
based on the hydrogen atom transfer mechanism,
which reflects biological conditions, and has been pro-
posed for evaluating antioxidant properties in foods.
For all these reasons, DPPH
 and ORAC assays are
proposed as standard methods to assess the antioxi-
dant properties of bee pollen. The steps to carry out
both assays are described. Due to related properties
for human health, some information about standard
methods commonly used for the determination of
antimicrobial and antimutagenic activities in bee pollen
are provided.
In conclusion, a significant quantity of standard meth-
ods for pollen research is now validated, and will be help-
ful to be included in the future International Directive for
Quality Control of bee pollen in the near future.
2. Producing bee pollen from harvest to storage
To provide a good quality product it is necessary to
get results in the conception of a professional, tech-
nical and competitive apiculture, mainly dealing with
the opening trade between the nations, MERCOSUR,
European Union, and others (Barreto et al., 2006;
Panetta, 1998). The beekeeping segment in the indus-
try context, fits into the conception, “in search of
excellence and quality,” concerning the significant role
that the products represent to humanity (Bassi, 2000;
Germano et al., 2000).
Bee pollen harvesting requires special skills and
guidelines. In this work we will represent the best
methods for a better quality of product in line with the
good practices in the field.
Standard methods for pollen research 9
The main guidelines to obtain the best raw material
are
1. Production plan – Such planning should aim at the
highest yield in both quantity and quality. Inside an
organization, technical support is crucial, and can be
achieved by courses, training, internships and even
from specialized consultancy, according to the needs
of the production (Barreto et al., 2006);
2. Directing the raw material – It is of an unquestionable
importance to direct the raw material for processing
and acquisition in an ideal standard with the following
sequence: production, harvest, transport, processing
and storage.
 Production: Standardized handling results in the har-
vest schedule.
 Harvest: Harvest is the first stage from raw material
operations to processing; this process should undergo
rigorous contamination control.
 Transport: Incorrect transport can negatively affect the
quality of raw material. The following parameters
should be considered (Barreto et al., 2006):
i. Adaptability of the vehicle to the type of
raw material;
ii. Suitable packaging for the quality and type of the
collected material;
iii. Necessity of traveling long distances (weather con-
ditions, temperature, sun exposure, dust, rain-
fall, etc.);
iv. Internal environment of the transport (odor-free).
 Processing: In this phase, the facilities should be planned
according to the flowchart that is demonstrated, as
well as the necessary equipment to process the prod-
uct such as freezers, dehydrators, sieves, cold chambers
and others (Barreto et al., 2006).
 Storage: In this phase, consider the following
requirements:
a. Type of packing;
b. Product identification, label;
c. Temperature, humidity, luminosity;
d. Time of travel to the facilities should be the lowest
possible (Barreto et al., 2006).
 During the next steps, other situations can occur, for
example: Deterioration of raw material: It usually
occurs by the action of living organisms (insects, micro-
organisms, etc.), by physical processes (freezing, heat-
ing, pressurizing, etc.) and biological (fermentation,
rancidity, etc.). Follow all the previous recommenda-
tions to avoid this situation, otherwise do not use the
product (Franco, 2002; Silva, 2000).
 Contamination of raw material: Avoid the following
causes of contamination (Gilliam et al., 1989;
Silva, 1999):
a. Lack of cleanliness in rooms and personnel;
b. Exposure of raw material to fresh air;
c. Exposure of raw material to environment with
dirt, garbage, animals with injuries and infectious
illness and poor and badly kept premises.
2.1. Production and handling of bee pollen
before processing
Bee pollen is the result of specific conditions of the
bee colonies, normally with regular population density
of a bee hive in development, where the bees are
stimulated to intensify pollen collection. Since it is a
protein material production, this raw material will be
susceptible to a more rigorous standard in the
hygiene requirements, from the production up to its
processing and storage (Barreto et al., 2006).
2.1.1. Quality control in the apiary
The quality control of bee products begins in the api-
ary because these products do not get their quality
by handling, but by nature’s attribute. The regulation
of good practices in the elaboration and food produc-
tion is a broad subject that is applicable in most of
the production. It is known that the microflora in the
bee pollen can have two different origins: (1) by the
normal microflora, originated from pollen, being
yeast, mold and bacteria; (2) by outside microflora,
originated from practices of handling, elaboration and
product storage.
To achieve a better-quality product some steps
should be done follow this check list:
a. Systematized handling of bee colonies;
b. Feeding the bees with high quality and reliable prod-
ucts when forage is scarce;
c. Daily collection of bee pollen;
d. Disposal of bee pollen collected on rainy days;
e. Sanitation of equipment in contact with the pollen
at the harvest period (pollen press, hive tool
and brush);
f. Sanitized bee gloves and beekeeper’s clothing;
g. Packing of harvest and transport previously sanitized
and dried;
h. Enough harvest time and transport to keep the integ-
rity level of the product until the processing facility.
2.1.1.1. Smoke. Smoke is a safety procedure used
during the handling of bee colonies. However, the
fuel materials, sawdust, should be sieved to avoid
producing unwanted dust that can affect bee pollen
pellets. The smoke should have a pleasant odor, and
pine, cypress and eucalyptus leaves and lemongrass
can be used as additives to the mixture. Animal
manure must not be used. The smoke should not be
aimed at the combs with honey (avoiding that the
product acquires unpleasant flavor) or brood. The
10 M. G. Campos et al.
smoke should be applied horizontally in the bee hive
when there is no exposed frame. It should never be
aimed directly at the raw material in production
(Barreto et al., 2006).
2.1.1.2. bee hives. Nowadays bee hives are often
already impermeable, which avoids the previous practice
of painting then, with the risk of heavy metal contami-
nated paints (Barreto et al., 2006)
2.1.1.3. Clothing. The overalls or coats used in the api-
ary should be cleaned after each handling session to
avoid contamination in the bee hive by moth eggs and
microorganisms that may adhere to clothes (Barreto
et al., 2006).
2.1.1.4. Tools. Hive tool, clean rubber gloves. After
each handling, in case any bee hive is found conta-
minated, these utensils should be substituted so as
not to contaminate the next hive (Barreto
et al., 2006).
2.1.1.5. Inputs. The main input used in the bee hive is
the bees wax. The beekeeper should acquire this
material in stores or suitable processing centers.
Whenever buying bees wax, make sure it is packed
with PVC film, whether there is moth or not
(crumbled aspect in the plastic foil). Any wax that is
directly packed with newspaper should be avoided as
printing ink is a powerful source of heavy metals
(Barreto et al., 2006).
2.1.1.6. Handling. The choice of technology to be
used in production handling is crucial. The material
applied in the production has to be standardized, and
it needs to be easily handled to avoid higher bee
losses, as well as allow perfect sanitation of the prod-
uct. For example, some hive boxes made of wood
are now being replaced by nontoxic plastic boxes
that permit efficient sanitation. Another point is the
handling standard; it refers to the conduction of the
production in a way that the harvest can be realized
in all bee hives in a timely fashion (Petersen
et al., 2011).
2.1.2. Types of pollen traps
Among frontal, floor, intermediate or bottom pollen
traps (Figure 1), each one has its advantages and dis-
advantages. The ideal pollen trap is one that can
receive the pollen in a safe and hygienic manner.
The frontal pollen trap is practical and hygienic but
because it is attached to the main floor, it receives
a significant amount of dirt resulting from the
internal bee hive cleaning. The bees continuously
remove particles from inside to outside of the hive,
making the pollen trap dirty. In the intermediate
internal pollen trap, it is opened a secondary
drawer, the main one being closed with a ventilation
screen. It is noticed that the dirt level decreases and
even the pollen has a lower humidity level than the
previously mentioned. However, it can cause an
increase in the bees’ defensibility due to the
obstruction of the main entrance. It is recommended
to choose a pollen trap that can be easily placed
and removed. In addition, a pollen trap that makes
the handling easier, allowing better ventilation and
sanitation, allowing the flight of drones and pro-
tected against weather changes. In tropical countries
such as Brazil, beekeepers have been using the
“Tropical Africanized Type” (Figure 1) with good
results (Barreto, 2004).
The collection of pollen (Figure 2) should be done
daily, there are some devices that prescribe the weekly
collection of the product, but there is no consensus
among scientists about such procedure.
Figure 1. A tropical Africanized pollen trap (TTA).
Figure 2. Pollen collection by beekeepers.
Standard methods for pollen research 11
2.1.3. Chosen area of apiary production
Choice of apiary location can be problematic.
Today, the choice should be careful, taking into
account the use of agrochemicals, both in crops and
in animals. Pollution is not exclusive to urban areas,
and unfortunately some rural areas can show con-
tamination in air, water and or soil. Rubbish in the
apiary attracts undesirable insects, exposing the raw
material to the risk of unwanted contamination
(Figures 3 and 4).
2.1.4. Quality control in the collection
Damp compressed pollen should be collected separately
from the raw materials intended for processing, and
should only be used for bee feed (Figure 5). The
collecting box should not be exposed to contamination
from the ground.
2.1.5. Quality control in the transport of raw material
At harvest, the beekeeper should be aware of the type
of packaging used to transport the product from the
apiary to the processing sector, thus avoiding:
a. crushing of raw material;
b. contamination with dust;
c. transfer of odors and humidity as well as high temper-
atures, which may interfere with the quality of the
product to be processed.
2.2. Handling of bee pollen in the processing sector
This sector should be provided with a reception room
for production (raw material) and processing rooms,
according to the Technical Regulation on Hygiene and
Sanitary Conditions and Good Preparation Practices for
Developers of Establishments/Industrialized Food
(Example for Brazil – Ordinance 368 of 4 September
1997 (Brasil, 1997) obtained in www.agricultura.gov.br/
ada/dipoa/). Other countries have similar regulations.
In these areas, verify the rules:
a. machinery that will be in contact with the raw mater-
ial should be in stainless steel (AISI 304);
b. manipulation, the process of cleaning and sanitizing
should be adopted integrally;
c. use of heat, dehumidifiers, chemical agents, must have
precise control;
d. water used for cleaning the room and machinery
should be of known origin with recognized purity.
Figure 3. Conditions to be avoided in the apiary and surround-
ing area.
Figure 4. Bad conditions of the hives and packaging containers;
hives on the ground are subject to rot and termite attack.
Figure 5. Damp compressed pollen.
12 M. G. Campos et al.
2.2.1. Flowchart from production to processing of pollen
2.2.2. Processing bee pollen
In the processing area, the handler must necessarily fol-
low the sequence below (Barreto et al., 2006):
a. Reception of raw material ("In Natura" pollen);
b. Primary labeling, date and place of harvest;
c. Removal of coarse dirt, dead bees, chalk brood mummies;
d. Screening;
e. Freezing to 18 C for 48 hours minimum and max-
imum 15 days;
f. Dehydration step I at a maximum temperature of
42 C for 8 hours;
g. Cleaning by ventilation;
h. Manual cleaning;
i. Dehydration step II maximum temperature of 42 C
for 4 hours;
j. Checking moisture content of dehydrated bee pollen;
k. Final filling of the dehydrated product;
l. Final product labeling valid up to 12 months;
m. Storage;
n. Shipping.
2.3. Materials and equipment involved in
the process
The collection tray often used is a model made of non-
toxic, broad and shallow plastic, suitable for food. Fresh
pollen contains high humidity content (20 to 30%), conse-
quently, fragile to large overlap, becoming brittle and caus-
ing an increase of pollen powder, representing the most
significant loss at the time of processing. These trays must
be provided with a cover which will prevent dirt from
becoming incorporated into the product during transport
(Almeida-Muradian et al., 2012; Brasil, 1997, 2001).
In Figures 6 and 7 can be seen the steps involved
from the reception to the delivery of bee pollen. When
the product arrives at the facilities, it has to be:
a. Transferred to lots;
b. Pass through wet sieving (usually with maximum cap-
acity of 5 kg);
c. Freeze. Pollen should be in a sterile polyethylene con-
tainer and immediately placed in a freezer. Bee pollen
will remain there for a minimum of 48hours to destroy
Standard methods for pollen research 13
possible mites, eggs or larvae of the wax moth (Galleria
spp.), and other insects. The freezing will also have the
property of stabilizing and controlling the development
of microorganism related to normal microflora con-
tained in pollen (Ibrahim & Spivak, 2006). The collected
pollen can be stored in the freezer for a period of a
few weeks. After a month, it begins to show change in
taste, texture and color;
d. Dehydrated. For this step is necessary to have a green-
house with adjustable thermostat, keeping the tempera-
ture between 40 and 42 C. Internally, the equipment
consists of stainless steel with shelves that receive trays
with screens in stainless steel AISI 304 that will allow effi-
cient cleaning and sanitizing. The dehydrating aims at the
water removal until it reaches the maximum humidity of
4–8% (Campos et al., 2008), but for tropical countries 4%
is often recommended (Thakur & Nanda, 2020).
Frozen pollen is usually placed in a greenhouse, which
facilitates its uniform distribution in the screened trays,
but it should not be submitted to an immediate tempera-
ture of 42 C. It is essential to respect a thawing period of
the material to prevent damaging the product due to ther-
mal shocks. It is suggested that after removing the pollen
from the freezer, it should be transferred to a refrigerator
for approximately 2 hours for defrosting. The distribution
of pollen in the tray should be done in thin layers. Every
instrument in contact with the pollen must be stainless
steel to prevent contamination. The pollen must remain in
the greenhouse from 8 to 12 hours. After dehydration, the
packing of bee pollen will involve three steps:
a. Transferring dehydration, sealed loads;
b. Cleaning – First undergo the pollen in sieve particle
size to have the larger lumps undone and separate the
commercial pollen powder after applying dry air to
remove dirt such as bee fragments that are possibly
present. The final phase of cleaning is grooming mate-
rials not previously removed, for example, any prop-
olis that is still mixed with bee pollen;
c. Packing – Should be done after cleaning, reducing the rehy-
dration possibilities of the processed material. For this pro-
cedure, you must use a nontoxic plastic bag, suitable for
food packaging as primary, which remains protected by sec-
ondary packaging of nontoxic plastic barrels of first use or
food cardboard. Before closing it, air must be removed
from the bag with help of a vacuum pump. Storage should
be in dry conditions, with mild temperature and away from
light. For fractional packing of bee pollen, it is suggested
placing silica gel pads for foods – that will keep humidity
level of the product stable (Brasil, 1997).
3. Inexpensive pollen load authentication bymeans
of computer vision and classification algorithms
Bee-keepers, bee-keeping associations, and laboratories are
interested in detecting fraud in pollen, and require tools to
standardize and authenticate bee pollen origin to guarantee
their nutritive and health benefits. Microscopic analysis of
pollen grains, which form bee pollen loads, is a precise
method for identifying pollen origin. However, this process
requires the laboratory work of melissopalynology experts,
and is thus time consuming and costly. There have been
many attempts to automate pollen grain identification by
computer algorithms but there is no inexpensive, complete,
and automated process (Allen, 2006; Boucher et al., 2002;
Rodr
ıguez-Dami
an et al., 2006). Image processing techni-
ques and one-class classification algorithms can be used to
identify unknown pollen types. Concretely, the mean shift
algorithm (Comaniciu & Meer, 2002) filters the pollen
image and is used to homogenize pollen load color infor-
mation. Then, one-class classification algorithms (Chandola
et al., 2009; Moya et al., 1993) identify each local pollen
type. The method makes use of a multi-classifier algorithm,
designed to aggregate one-class classifier outputs, given a
unique response with a confidence measure. Finally, an
ambiguity discovery algorithm is also included to detect
identical pollen type and reduce misclassification.
The chapter provides general details of the algo-
rithms involved in the methodology, describes the
necessary hardware for the standard method, and
presents the results of applying the image processing
algorithms and one-class classifiers. In addition, the
chapter presents the overall process to create a soft-
ware dictionary with known local pollen types, how to
use the graphical user interface, and a guide to separate
and analyze the pollen loads of the images by type. The
validation of the proposed standard methodology is car-
ried out for the authentication of four pollen types,
Cistus ladanifer, Rubus, Echium, and Quercus ilex, which
serve as an example of the methodology (Otero &
Losada, 2002). Totally, a dataset of around 2000 instan-
ces has been used to validate the trained system.
Figure 6. Processing “fresh” bee pollen; first screening.
14 M. G. Campos et al.
3.1. General overview of the method
Figure 8 shows the outline of the method. In a nutshell, it
starts with preparation of a pollen load sample to acquire the
image by the computer vision system. Second, the pollen
loads of the image are segmented from background (see
Section 3.2.1) and filtered (see Section 3.2.2). Third, the color
instances of the processed pollen loads are used to train a
multi-classifier model based on one-class classifiers (one for
each local pollen type). Finally, the multi-classifier outputs the
authentication of each color instance, classifying them as a
known local or non-local (outlier) pollen type. Section 3.2.3
provides more details about the classification algorithms.
3.2. Description of the computational techniques
3.2.1. Segmentation algorithm
The Otsu segmentation algorithm (Otsu, 1979) is the one
applied to the gray-scale image to extract the pollen loads
from the background. Later, a morphological opening oper-
ation is applied to the thresholded binary image (Gonzalez
& Woods, 2008). The goal of this computer vision oper-
ation is to remove those small objects having less than 50
connected pixels in an 8-connected neighborhood.
Those pixels extracted in the latter phase are analyzed by
the remaining processing algorithms. Their color information
can be represented in several ways. The most common is the
RGB space where colors are represented by their red, green,
and blue components in an orthogonal Cartesian space.
However, the RGB space does not represent the higher levels
processes which allow human color perception. Color is bet-
ter represented in terms of hue, saturation, and intensity, as
HSI or HSV spaces do (Lucchese & Mitra, 2001). However,
the latter color spaces are not perceptually uniform. The CIE
and L
u
v
 and L
a
b
 are ideal for color recognition
because of the following three properties: a) separation of
achromatic information from chromatic information, b) uni-
form color space, and c) similarity to human visual perception.
In these color spaces, for instance, the Euclidean distance
between two color points can be easily calculated. This prop-
erty will ease the work of the classification algorithms.
3.2.2. Homogenizing the pollen loads of the image by mean
shift filtering
Each extracted pollen load from the image has many dif-
ferent L
u
v
 color values, possibly as many as pixels
contained in the load. This has a negative impact on pol-
len load color authentication since human experts iden-
tify each pollen load as a unique color. Therefore, we
Figure 7. Example of a warehouse for bee pollen production in Brazil. This is one of the largest in Latin America, located in the state
of Bahia, northeast Brazil. The main source of pollen is the coconut palm, achieving high standards of excellence and quality of the
product: (a) Dehydration room 40 C; (b) distribution trays; (c) aeration of the dehydrated product room.
Standard methods for pollen research 15
need a procedure of homogenizing the images of pollen
loads, which is divided into three different steps:
1. First, we apply an image processing algorithm before using
the classification methods. The goal is to homogenize and
smooth the large quantity of different color points of a pol-
len load in just a few unique and representative color
instances. One of the best methods for discontinuity-pre-
serving smoothening in image processing is the mean shift
algorithm, proposed in Comaniciu and Meer (2002), and in
line with the feature space analysis. The main strengths of
the mean shift algorithm are: (a) it is an application inde-
pendent tool, (b) it is suitable for real data analysis, (c) it
does not assume any prior shape on data clusters, (d) it can
handle arbitrary feature spaces and (e) it has only one par-
ameter, the bandwidth selection window size. The mean
shift procedure, originally presented by Fukunaga and
Hostetler (1975), is a procedure for locating the maxima of
a density function given discrete data sampled from that
function. It is also an iterative method, starting with an initial
estimate or points xi and a G kernel function. This kernel
function determines the weight of nearby points for the re-
estimation of the mean. Denote by yj with j ¼ 1, 2, ::: the
sequence of successive locations of the kernel G :
yjþ1 ¼
Pn
i¼1xiG k xxih k
 2
Pn
i¼1G k xxih k
 2 , (1)
is the weighted mean at yj computed with the kernel G
and y1 is the center of the initial position of the kernel.
Taking this theory as a base, Comaniciu and Meer
(2002) extended the mean shift procedure for filtering
and segmenting images. A color image is typically
represented as a two-dimensional lattice of 3-dimen-
sional vectors (pixels). The space of the lattice is
known as the spatial domain, while the color is the
range domain. For both, the mean shift algorithm will
work with the Euclidean metric defined in the previ-
ous section for the color space.
2. When both domains are concatenated, the dimensions
of the joint spatial range domain are compensated by
a proper normalization. Thus, the multivariate kernel
is defined as the product of two radially symmetric
kernels, and the Euclidean metric allows a single band-
width parameter for each domain:
Khs , hr xð Þ ¼
C
h2s h
p
r
k k x
s
hs
k2
 
k k x
r
hr
k2
 
, (2)
where xs is the spatial part, xr is the range part of a
feature vector, kðxÞ the common profile used in both
domains, hs and hr the kernel bandwidths, and C the
normalization constant. A kernel process usually pro-
vides satisfactory results and then the user just has to
provide one parameter h ¼ ðhs, hrÞ which controls
the size of the kernel, and thus the smoothening reso-
lution. Nevertheless, replacing the pixel in the center
of the window by the average of the pixels in the win-
dow blurs the image. Discontinuity-preserving
smoothening techniques, on the other hand, reduce
the amount of smoothening near abrupt changes. The
mean shift algorithm uses a bilateral filtering which
works in the joint spatial-range domain. The data are
independently weighted in the two domains and the
Figure 8. The proposed standard method for detecting unknown pollen loads. The processing chain finishes with an authentica-
tion output.
16 M. G. Campos et al.
centered pixel is computed as the weighted average of
the window. The kernel in the mean shift procedure
moves toward the maximum increase in joint density gra-
dient, while bilateral filtering uses a fixed static window.
3. The last step is the application of the mean shift filter-
ing algorithm. The algorithm works as follows. Let xi
and zi, with i ¼ 1, 2, :::, n be the (pþ 2)-dimensional
input and filtered image pixels in the joint spatial-
range domain. Being the superscripts s and r the spa-
tial and range components of a vector, respectively,
and c the point of convergence, for each pixel:
a. Initialize j ¼ 1 and yi, 1 ¼ xi;
b. Compute yi, jþ1 according to Equation (2.1) until
convergence, y¼ yi, c;
c. Assign zi ¼ xsi , yri, c
 
:
The spatial bandwidth has a distinct effect on the out-
put when compared to the range (color) bandwidth. Only
features with large spatial support are represented in the
filtered image when hs increases. On the other hand, only
features with high color contrast remain when hs is large.
3.2.3. One-class multi-classification algorithm based on distance
In Chica and Campoy (2012) authors compared four differ-
ent one-class classification approaches for the problem:
Gaussian, Parzen classifier, SVDD and kNN: The kNN algo-
rithm with k ¼ 1 showed the best performance. Then, this
is the proposed algorithm for the pollen recognition system.
kNN, originally provided by Dasarathy (1991), is a dis-
tance-based one-class classifier based on the assumption that
normal data instances occur in dense neighborhoods while
anomalies occur far from their closest neighbors. The basics
of the algorithm for one-class classification is that the anomaly
Figure 9. The computer vision system needs inexpensive hard-
ware such as a camera and a stand to control illumination.
Figure 10. Different images of pollen load samples which were taken with our vision system. Left and central images belong to known
pollen types (Rubus, Cistus ladanifer, Quercus ilex, and Echium, respectively). Right images are non-local samples and must be rejected by
the system.
Standard methods for pollen research 17
score of a data instance is defined as the distance with its kth
nearest neighbor in each dataset.
Nearest neighbor classifiers always require the defin-
ition of distance or similarity measures between two data
instances. For continuous features, the Euclidean distance
is the most popular choice. In the case of choosing k ¼ 1
as the parameter of the algorithm, each new instance z will
be considered as target or outlier depending on the classi-
fication of its closest neighbor in the training data.
In multi-class anomaly detection (our pollen authentication
problem) training data contains labeled instances belonging to
multiple normal classes but it does not contain anomalous
instances. A test instance is considered anomalous if it is not
classified as normal by any of the classifiers. To do this, a confi-
dence score obtained from the prediction output of the clas-
sifier is normally provided. If none of the kNN classifiers are
confident in classifying the test instance, the instance is then
labeled as anomalous (Chandola et al., 2009).
We have followed the latter approach, modeled as
follows: jCj being a set of known local bee pollen load
types, the training data will contain instances belonging
to jCj classes. In order to use one-class kNN classifiers
and be able to reject unknown pollen load types, we
have to decompose the classification system in jCj binary
sub-problems. Thus, we will train jCj different kNN clas-
sifiers: f1, f2, :::fjCj: An ensemble scheme is used to fuze
all of them in a multi-class authentication output.
Therefore, for each pollen color instance x we first
map each one-class classifier output fiðxÞ to a posterior
probability Pðy ¼ cjxÞ: These probabilities are also nor-
malized in the range ½0, 1: The posterior probability of
each classifier’s target can be considered as the confi-
dence CFocðy ¼ cjxÞ for one instance x to belong to class
c: In order to classify an incoming pollen load sample as
one of the jCj possible pollen types we build a multi-clas-
sifier. The multi-classifier will compare the confidence
CFocðyjxÞ of all the one-class classifiers and will provide a
global prediction from the most reliable one-class classi-
fier. The multi-classifier prediction x is given by:
x ¼ max
1c Cj j
CFoc cjxð Þ: (3)
3.2.4. Multi-classification confidence and ambiguity discov-
ery process
It is also necessary to estimate the confidence of the
multi-classifier prediction. To do this we first introduce
two parameters, Toc and Tm, as done in Goh et al.
(2005) in Equations (4) and (5):
Toc ¼ CFoc xxð Þ; (4)
Tm ¼ Toc  max
1cjCj, c 6¼x
CFocðcjxÞ: (5)
Toc is the highest confidence factor from the jCj binary
one-class classifiers and determines the multi-classifier
prediction class x: However, Toc might not be enough to
estimate the global confidence of the multi-classifier
prediction. This is the reason for introducing the second
parameter, the multi-class margin Tm: Wrong predictions
could have high Toc but small Tm: Then, correct predic-
tions must have high multi-class margin values Tm:
Goh et al. (2001) showed that there is a better separ-
ation of correct from erroneous predictions if the multi-
class margin variable is used. After a preliminary experimen-
tation, we set parameters Toc and Tm to 0.5 and 0.01,
respectively, to be used in the final decision stage of the
multi-classifier, as shown in the rule of Equation (6).
x is accepted, if Toc  0:5 and Tm  0:001,
outlier, otherwise:

(6)
Sometimes, one or more bee pollen types could have
exactly the same color description as another. In that
case, the kNN multi-classification system must be able to
detect, during the training phase, that one incoming local
pollen type is identical to one already existing (i.e., it is
already included in the dictionary). The goal for the soft-
ware prototype is to also warm the user not to intro-
duce duplicated entries in the dictionary. This mechanism
is called ambiguity discovery. Being em1 a sensitivity-speci-
ficity error of a multi-classifier before the inclusion of
the new pollen type, and em2 the error of a multi-classi-
fier after the inclusion of the classifier of the new pollen
type, we define De ¼ ðem2  em1Þ as the difference
between them. In our case we have used the F-measure
(van Rijsbergen, 1979) as the em1 error measure.
The ambiguity discovery process is launched every
time the De parameter is higher than a fixed value.
Then, if De exceeds a threshold value, ambiguity discov-
ery is triggered, and the process works as follows:
1. The confusion matrix of the new multi-classifier for
the testing data is computed;
2. The maximum value of ðKi, LjÞ is calculated
with i 6¼ j, K being the vector of real classes and L
the vector of predicted classes;
3. The user is asked about merging conflicting classes ci
and cj into a unique class ci;
4. The multi-classifier is trained according to the
response of the user in the third step.
3.3. Experimentation
Section 3.3.1 explains the data used for the experimen-
tation. Section 3.3.2 explains the image processing and
filtering results. Section 3.3.3 presents the results to
analyze the behavior of the one-class classifiers and
multi-classification systems. Finally, Section 3.3.4 shows
the validation of both the software prototype and pro-
posed standardization process.
3.3.1. Experimental setup: equipment and data
The main equipment used in the experiments is com-
posed of:
18 M. G. Campos et al.
 Camera and optics device to obtain images of the pol-
len load samples. The resolution does not need to be
extremely high since only color and edge information
will be processed. (Example: color camera uEye UI-
1485LE-C (www.ids-imaging.com) with an Aptina
CMOS sensor in 5 M Pixels resolution (2560	 1920
pixels). The light-weight housing of the UI-1485LE fea-
tures a C/CS lens mount with adjustable flange back
distance. A focal lens obtained from Goyo Optical
(www.goyooptical.com) is also used.
 The lighting conditions are decisive and must be con-
trolled. An external lighting generator, which points out
the pollen load sample is employed to ensure that the
color of the pollen loads do not change for the different
runs. The external lighting consists of a 50 LED-based
ring light from CCS Direct Lighting (www.ccsgrp.com).
This lighting device can be mounted on a CS mount.
 A support device is needed to allow a fixed position for the
camera, lens, and lighting. The sample of pollen loads is
placed on the base of this supporting device in order to get a
stable and low-cost system to acquire images of the sample
for a posterior processing and analysis (see Figure 9).
 Finally, a computer must be connected to the camera
via USB to install the software. Proprietary software
can be used to take snapshots of the pollen load sam-
ples. Then, the proposed computer vision and classifica-
tion system will be controlled by the software
prototype to ease the labor of the expert.
Different samples from Spanish bee pollen loads were
obtained from beekeepers to build the authentication models
and validate them against nonlocal samples. Samples belong-
ing to four Spanish local pollen types (Rubus, Echium, Cistus
ladanifer and Quercus ilex) and non-local samples were identi-
fied, labeled, and grouped by experts. Figure 10 shows these
pollen loads samples. In this figure we can see how, even for
experts, the separation and classification of the pollen loads
by color is difficult and subjective. One can also see in the fig-
ure how non-local samples can be misleading (bottom right
loads of Figure 10).
3.3.2. Image processing results
The first step of the method is processing the images of
the pollen samples (taken in TIFF format and resolution
of 1024	 768 pixels). The software will apply the filter-
ing algorithm after segmenting pollen loads from back-
ground. As explained in Section 3.2.2, the selection of
an appropriate bandwidth parameter for the mean shift
algorithm is not trivial. This selection depends on the
type of images. The implemented mean shift algorithm
receives the spatial bandwidth hs, the range bandwidth
hs, and the minimum segment area in number of pixels.
The last parameter is fixed to a high value (20 pixels)
not to segment a pollen load into different parts.
The selection of the other two parameters is more
difficult. High spatial bandwidth values merge different
bee pollen loads because they are normally close to
each other. Low range bandwidths do not effectively
aggregate the entire color information of the pollen
loads. We conducted a preliminary experiment with dif-
ferent values. In order to illustrate the importance of
this parameter selection, we show in Figure 11 an ori-
ginal Rubus data sample together with three output
images after applying the filtering algorithm with differ-
ent values of hs : 7, 15, and 30. Additionally, it is
important to remark that higher bandwidth values mean
higher computational time. Thus, we set the bandwidth
values to h ¼ hs, hrð Þ ¼ ð15, 20Þ for the experimenta-
tion, which seems one of the best parameter combin-
ation for this pollen load authentication problem.
We generate, from the processed and filtered
images, a set of 3146 color instances. Each color
instance has three input features, corresponding to the
L
u
v
 color space values, and its class (one of the four
pollen types or an outlier class). From this set of color
instances, 400 of them are used for training the four
one-class classifiers (100 for each pollen type), 800 to
test the one-class classifiers (400 belonging to the four
know pollen types and 400 were outliers). The rest of
the 1946 instances are used to validate the multi-classifi-
cation system (Section 3.3.4). Table 1 summarizes these
groups of experimental data.
Table 3. Evaluation measures of the final multi-classifier.
kNN multi-classifier model Accuracy (%) FN rate FP rate
0% rejected during training 92.5488 0.0108 0.1161
10% rejected during training 94.6043 0.0935 0.0167
Two cases are listed; considering 0%, and considering 10% of data as
outliers when training the one-class classifiers. FN and FP rates are
calculated taking all the known pollen types as the positive class and
the outliers as the negative.
Table 2. Evaluation measures obtained by the one-class kNN
classifiers for each of the four pollen types.
Rubus Echium
FN rate FP rate F-measure FN rate FP rate F-measure
0.05 0 0.9744 0.09 0 0.9529
Cistus ladanifer Quercus Ilex
FN rate FP rate F-measure FN rate FP-rate F-measure
0.06 0 0.9691 0.01 0.0028 0.9851
Lower values mean better performance.
Table 1. Description of the three independent datasets used
to train, test the classifiers, and validate the one-class
multi-classifier.
Dataset
name
Total
size
Pollen
types instances
Outlier
instances
Training 400 400 0
Test 800 400 400
Validation 1946 1016 930
Standard methods for pollen research 19
3.3.3. Numerical performance of the multi-classification
algorithms
In this section we employ the classification accuracy, false
negative and positive rates, F-measure, and confusion
matrix to numerically validate the multi-classifier per-
formance. A false negative (FN) occurs when the out-
come of the classifier is incorrectly predicting as outlier
when it is actually a target. A false positive (FP), on the
other hand, occurs when the outcome is incorrectly pre-
dicting as target when it is actually an outlier. The FN
rate measures the number of FNs out of the total num-
ber of negatives or outliers, and the FP rate calculates
the fraction of FPs divide by the total number of posi-
tives or target instances (Witten & Frank, 2005). The F-
measure provides a relation between the precision and
recall of the classification results (van Rijsbergen, 1979).
First, we analyzed the performance of the kNN classifi-
cation method for the authentication of each of the
selected local bee pollen types in isolation. These results
are obtained by classifying the test dataset with the train
one-class classifiers. The classifiers are trained without
rejecting any training instance as outlier (rejection thresh-
old equal to 0%). Table 2 shows the evaluation measures
of the classifiers for each pollen type. By observing figures
of Table 2 we can conclude that the FP rate is almost 0.
This means that the classifiers are able to correctly iden-
tify all the outliers (non-local pollen types) without mis-
classifying them as local pollen types. The FN rate is
higher than the FP rate although its value is low, 9% in the
worst case (Echium pollen type).
Table 3 shows the performance measures of the final
multi-classifier. In the first block of figures we show the
accuracy and FP-FN rates when rejection threshold equals
0%. In the second block, a rejection threshold of 10% is
used for obtaining the results. The kNN algorithm, with a
rejection of 10%, is the model having the best results as it
has low FP and FN rates, and has the highest accuracy.
Finally, Table 4 shows the confusion matrix of the
best multi-classifier configuration with a kNN algorithm
with a rejection of 10%. The reader can see that, in
general, there is no miss-classification between the
known pollen load types. There is just one miss-classi-
fied instance by the kNN multi-classifier. The highest
error is in the right column. These are FNs because
real pollen load type instances are classified as outliers
(non-local pollen types). Nevertheless, as it can be seen
in Table 2, the FN rate is low.
From the results, we can also see that the tradeoff
between FP and FN rates is better than in the isolated
one-class classifiers. This fact comes from the confidence
mechanism of the multi-classifier, able to discard instan-
ces classified by one of the one-class classifiers as local
pollen types when they are not clearly confident about
their decision. Although accuracy is obtained when classi-
fying all the classes (four known pollen types and out-
liers), the FP and FN rates are calculated between the
classification of the instance as known pollen type (one of
the four known types) or as outlier. There is almost no
error when classifying among local pollen types.
Finally and by comparing the validation measures of
the multi-classifiers and the one-class classifiers, it can
be concluded that the overall results of the multi-classi-
fier are better than the independent one-class classifiers.
This fact indicates a good behavior of the whole fusion
scheme for detecting the known pollen type classifiers.
3.3.4. Validation of the complete prototype
The proposed software prototype includes the software
methods to manage and create local pollen types, train
the models, and validate the whole process. The soft-
ware prototype was programmed in MATLAB using
some functions of the DD tools library (Tax, 2011).
Initially, the user needs to train the system with the
known local samples. The user must manually separate
Table 4. Confusion matrix of the multi-classification system formed by the four one-class kNN classifiers (one for
each pollen type).
Predicted pollen type
Real pollen type
Rubus Echium Cistus Quercus Outlier Total
Rubus 248 0 0 0 33 281
Echium 0 357 0 0 43 400
Cistus 0 0 275 1 10 286
Quercus 0 0 0 48 1 49
Outlier 4 0 5 8 913 930
Total 252 357 280 57 1000 1946
10% of the training data were rejected as outliers during the training phase of the one-class classifier.
Figure 11. Original image and resulting images after applying the
mean shift algorithm (varying spatial bandwidth hs to 7, 15, and
30, respectively).
20 M. G. Campos et al.
pollen types as shown in Figure 12. By using the diction-
ary tool of the software prototype, the user can also
save her/his local (known) types. See for instance Figure
13 where an entry called Rubus is introduced into the
system. After performing the training phase of the sys-
tem, a new incoming pollen samples can be processed
to start with the automatic pollen loads separation by
color. Some screenshots of the software application are
provided in Figure 14. In these images, one can see the
graphical user interface with the list of separated pollen
loads by type, the entries of the dictionary, and final
output of the multi-classification algorithm.
3.4. Concluding remarks
The propose standard methodology used a chain of
methods based on computer vision and classification
techniques for authentication the origin of bee pollen
loads. This process can be used to check and detect
fraudulent samples. A one-class multi-classification
scheme based on kNN and its merging scheme improve
the accuracy of isolated classification methods and is
able to provide a final global output. The best multi-
classifier configuration, formed by kNN one-class classi-
fiers, is able to achieve a 94% accuracy when detecting
local pollen types. The model has been validated in
1946 colors instances for the authentication of four
Spanish pollen types against different outlier samples.
The use of the presented standard methodology drastic-
ally reduces the time and effort spent by experts to sev-
eral seconds and can be used as a standard method for
macroscopically rejecting unknown pollen loads. Future
work can be devoted to apply a more interpretable
multi-classification system such as linguistic rule-based
classifiers. In that way, the users of the standard
method could understand the reason why a sample is
rejected as known local pollen type.
4. Evaluation and classification of corbicular
pollen morphology
The evaluation of quality of the corbicular pollen is one
of the areas which lacks unified and standardized evalu-
ation systems. However, descriptors used in the evalu-
ation of genetic resources of cultivated species of plants
could be a suitable system for the evaluation and char-
acterization of various types of corbicular pollen.
Among other analyzes, the technical parameters of
the corbiculae, may be an important marker for assessing
the geographical and botanical origin and subsequently
the nutritional quality of corbicular pollen. The industrial
production, however, requires methods for the identifi-
cation of corbicular pollen that can be systematically
included into the production chain. What the plant spe-
cies a corbicula is composed of, and what the plant com-
position and their proportionality is in the overall pollen
lay, are the indicators of the geographic and botanical
origin of the corbicular pollen (Carri
on et al., 2003).
Briefly, the pollen color can distinguish the types of cor-
bicular pollen and it mainly depends on factors such as pol-
len maturity, moisture, drying, pigments and the actual plant
source (No^zkov
a, Brindza et al., 2010). The authors suggest
a variety of colors ranging from white, yellow, orange, red,
with a greenish tint, to black and its shades. The works
also present a common fact: the color of the flower pollen
varies with the color of the corbicular pollen of the same
plant species (Chica & Campoy, 2012; Reiter, 1947).
The information regarding the size of corbiculae is
very scarce in the literature. The authors mention the
size only indirectly – either using the parameter of
weight, or the number of pollen grains in one corbicula
(Rodr
ıguez-Dami
an et al., 2006). The size, shape and
surface of the pollen grains, the corresponding botanical
origin, size of the openings in the pollen trap, beekeep-
er's handling of the corbiculae after harvesting, etc., are
all factors that have a decisive effect on the shape and
Figure 12. Pollen load samples separated by an expert to train
the system.
Figure 13. Screenshot of the prototype pollen dictionary. The
ambiguity discovery algorithm checks if there is an existing pol-
len type having identical features compared to the new sample.
Standard methods for pollen research 21
size of the corbiculae, and so also on its the quality.
The ornamentation of the exina of the pollen grains can
cause differences in the structure of the cor-
bicula surface.
4.1. Characteristics of corbicular pollen samples
The selected samples of corbicular pollen, frozen and
store at the temperature of 11 C to 18 C, immedi-
ately after collection.
4.2. Methodology for assessing the morphometric
characters of corbicular pollen
Quantitative and qualitative characteristics will be under
evaluation of morphological parameters of corbicular
pollen collected by the honey bees. For each character-
istic is designed and defined the methodology of evalu-
ation. Then a database of detailed image recordings is
created. The corbiculae in each sample is placed on a
glass slide using a fully automatic magnifying glass, for
instance, Zeiss Discovery with a digital camera
Figure 14. (a and b). Two screenshots of the prototype software detecting a local pollen type in the pollen load images.
22 M. G. Campos et al.
AxioCam, or similar. The corbiculae are placed on the
glass slide in the same direction. The upper part of the
corbiculae in the image is always the depression created
by placing the pollen around the bee leg inside the pol-
len basket.
4.2.1. Methodological procedures for evaluating the quantita-
tive characteristics
In the selected quantitative traits evaluate the corbiculae
by using the image analysis software Zeiss Axiovision or
open source software like ImageJ with the correspond-
ing modules (plugins) for automatic measurement
(Figure 15). In the case of corbicular weight, conduct
the measurement on precision analytical scales. The fol-
lowing section presents the assessment methodology of
quantitative characteristics, which were published in
No^zkov
a, Ostrovsk
y et al. (2010).
1. Corbicula size. The size of corbiculae determines eight
quantitative traits – bound height (mm), bound width
(mm), Feret minimum (mm), Feret maximum (mm),
shape index, symmetry, diameter (mm) and area
(mm2). The methodological procedure for the assess-
ment of these characteristics is based on the manual
for the Axiovision software.
a. Corbicula bound height. Indicates the height (in
the y axis) of a bound square for the selected
area. The square is drawn parallel to the x and y
axis (Figure 16). Unit of measurement: mm.
b. Corbicula bound width. Indicates the width (in
the x axis) of a bound square for the selected
area. The square is drawn parallel to the x and y
axis (Figure 16). Unit of measurement: mm.
c. Feret maximum and Feret minimum. This trait is
based on the measurement of a minimum and a
maximum distance. Two parallel lines are place
on the opposite sides of the object in 32 posi-
tions and evenly rotated by the respective angle
(Figure 16). The corresponding distance is meas-
ured in each position. The maximum value is
Feret maximum and the minimum value is Feret
minimum. Unit of measurement: mm.
d. Shape index (bound width/bound height). This par-
ameter is determined by calculating the two values.
It is not expressed by any unit. It can be used only
if the width and height of the corbicula has been
correctly specified. The shape index helps us
express the shape of the object. If the value is > 1,
the object is flatter. If the value is < 1, the object is
taller. If the value approaches 1, the object has a
round shape – it is equally as wide as it is tall.
e. Symmetry or Feret index (Feret min/Feret max). This is
a parameter determined by calculating the two meas-
ured values. It is not expressed by any unit. With this
parameter, it is possible to express the shape of the
object. Possible values range from 0 to 1. The values
approaching 1 indicate the presence of compressed or
round objects. Feret min and Feret max have approxi-
mately the same value. If the parameter has a low value,
the objects are long, elongate or oblong.
f. Diameter. It is assumed that the measure area is
a circle. This surface is then use as a basis to cal-
culate the corresponding diameter. Unit of meas-
urement: mm.
Figure 15. Flowchart of image analysis.
Figure 16. Evaluation of corbicula size characteristics.
Standard methods for pollen research 23
g. Area. This is determined by the boundaries of
the object by distinguishing the limit color of
the object from the background (Figure 16).
Unit of measurement: mm2.
2. Corbicula weight (mg). Each corbicula is weighed on
accurate laboratory scales.
From each sample select healthy and typical corbicu-
lae. The minimum number is 5 and the maximum is 100
corbiculae per sample.
4.2.2. Methodological procedures for evaluating the qualita-
tive characteristics
For the evaluation of qualitative traits, the set of corbi-
culae should be as big as possible, that it will cover the
widest variability of the select characteristics. The quali-
tative characteristics of the corbiculae can be evaluated
visually in digitalized images. From each sample select
healthy and typical corbiculae. A database of detailed
recordings, should be done with a fully automatic macro
magnifying glass, for instance, Zeiss Discovery with a
digital camera AxioCam (or similar). The following sec-
tion presents the assessment methodology of the quali-
tative characteristics (No^zkov
a, Brindza et al., 2010).
1. Corbicula shape. For the evaluation of the corbicula
shape, from three characteristics can be chosen:
shape, heart shape and cutout depth. The variability in
these characteristics is shown in Figures 17 and 18.
2. Corbicula surface. The variability in this characteristic
is shown in the Figures 18 and 20.
3. Corbicula color. The color of the corbicula can only
be assessed on samples of homogeneous color, i.e.,
samples of monofloral pollen. Color evaluate by using
a visual color grading scale by Royal Horticultural
Society (RHS) (2004) (or similar), which represents a
universal color system with the respective
color codes.
4.2.3. List of descriptors and classification
4.2.3.1. General structure of descriptors list. The list of
descriptors designed for the evaluation of genetic
resources of cultivated plant species have always been
specifically attuned to a particular plant species
or genus.
Each list of descriptors consists of three main parts
– the manual, passport descriptors and descriptors for
evaluation and characterization. The manual contains
the information on the structure and use of the descrip-
tors, as well as the general principles for assessing the
samples. The passport descriptors are designed accord-
ing to the nature of the evaluated object and provide
identification and passport data on the assessed sample.
Figure 17. Evaluation of corbicula shape.
Figure 18. Evaluation of corbicula heart shape.
Figure 19. Evaluation of corbicula cutout depth.
24 M. G. Campos et al.
The descriptors for evaluation and characterization are
designed based on the study of variation in quantitative
and qualitative characteristics. Using these descriptors,
the morphological, biological and economic characteris-
tics are evaluated. When creating a list of descriptors
for the corbicular pollen, it will always be necessary to
specify the type of insect involved in the creation of the
corbiculae and the descriptors for characterization and
evaluation will be designed specifically for the different
types of monofloral pollen, the predominant plant spe-
cies and especially in the case of multifloral corbicular
pollen, according to the geographical origin (Figure 21).
4.2.3.2. Passports descriptors. The passport descriptors
provide general information about the accession
included in the collection and the management of
growth (characterize by the collection method, storage
and handling of the accession). The passport descriptors
should define the mandatory information to identify the
originality of the accession (Haussmann & Parzies,
2009). In the case of genetic plant resources, the pass-
port data are exchange, and for the individual
institutions to be able to communicate with each other,
an international standardize list of Multi-crop passport
descriptors (FAO/IPGRI, 2001) was created.
To evaluate the quality of corbicular pollen, it is also
necessary to design the passport descriptors, which will
guarantee the authenticity and identify the specific prop-
erties of the evaluated sample. What we consider to be
the most important step is that the passport descriptors
also determine the botanical and geographical origin.
This means, it should be clear whether the pollen in
question is monofloral or multifloral. In the case of mul-
tifloral pollen, the geographic origin should be identified,
which ensures the representation of specific
plant species.
Similar to plant species, even the corbicular pollen
should have passport descriptors universal for all types
of pollen. Despite the very high heterogeneity between
the types of corbicular pollen, they should be designed
in a way that they can be used to collect the necessary
information be it in monofloral or multifloral pollen.
The following section contains a draft structure of
passport descriptors (Figure 22). The information on
Figure 20. Evaluation of corbicula surface.
Figure 21. Flow chart of the descriptors list structure.
Standard methods for pollen research 25
the insect species or subspecies creating the corbiculae
should come first. It should be followed by information
on the collection by the beekeepers, i.e., the collection
frequency, pollen trap characteristics, growing season
and habitat characteristics. What is also important is
the information on storage (method, length, tempera-
ture and humidity) and pollen processing (moisture, dry-
ing method). They should also include information on
purity, i.e., presence of visible impurities, mold or
whether the pollen comes from genetically modified
plants (GMPs). The sensory characteristics such as fla-
vor, color and taste, are also an important identifier of
the specific characteristics and authenticity of pollen.
4.2.3.3. General structure of characterization and
evaluation descriptors. A descriptor is an independent
rating system for one specific characteristic. We use
them to describe quantitative or qualitative
characteristics.
The structure of each descriptor is the same (Figure
24), which is a big advantage ensuring their universal use.
Descriptors consist of a maximum of 9 classification
grades (Figure 23). The minimum number of classification
grades is two. Their number is proposed according to
the observed variability in the characteristic.
For each classification grade, the respective verbal
description is proposed. In the case of quantitative
characteristics, the individual classification grades should
come with intervals (ranges). An important part of the
descriptors is the assessment methodology with concise
diagrams and pictures that describe the evaluation
in detail.
4.2.3.4. Methods and techniques for creating and
modifying the descriptors. A descriptor, which was
included in the list of descriptors, covers the known
variability of morphological, biological or economic
characteristics and properties. If a new characteristic or
trait is discovered in the future, which better character-
izes and identifies the evaluated objects, the methods
and techniques for creating new or modifying the exist-
ing descriptors should be observed. In the next section
we present the procedures and rules for the creation
and design of various parts of the descriptor for the
purpose of characterization and evaluation.
a. Completeness. The descriptor grades should be
designed to allow for any measured value or trait in
the characterized sample to be included into the
respective grade. This rule applies to both quantitative
and qualitative characteristics (Table 5). In the case of
quantitative characteristics, the intervals must be
clearly defined. Also, a specification whether the range
is open or closed should be included (Table 6).
Figure 22. Descriptor structure.
Figure 23. Visual representation of the line for the limit points xmin ¼ 1.81; xmax ¼ 4.26 and the coordinate “x4” to the point “y4.”
26 M. G. Campos et al.
b. Unambiguity and uniformity. A descriptor is a ranking
system with a clear identification of what, how, when,
and to what extent it evaluates. An adequately
designed methodology, pictures, diagrams and classifi-
cation levels helps maintain the overall unambiguity.
The descriptor should be used uniformly to evaluate
the given characteristic. This means that all evaluators
should use the same procedure for characterization
and classification.
c. Measurement ranges. In order to ensure the unambi-
guity in the evaluation of quantitative characteristics
(in the case of a continuous scale), the classification
grades are recommended to be supplemented by
ranges. By specifying the respective ranges, misinter-
pretations in the characterization of samples are
avoided. One of the ways to create the evaluation
scales is using the principle of membership function in
the fuzzy set (No^zkov
a et al., 2011).
a. The first important step is to determine the limit val-
ues in the given characteristic. This means that it is
necessary to determine the minimum (xmin) and max-
imum limit (xmax), which can be achieved in the eval-
uated characteristic independent of the
environmental conditions.
b. Based on the defined limits, the equation is expressed
as follows:
The straight line is uniquely determined by two points
(xmin, ymin); (xmax, ymax).
[1]
y ¼ ymin þ ymaxyminxmax  xmin x  xminð Þ
In our case, xmin ¼ minimum limit and xmax ¼
maximum limit, ymin ¼ 0 and ymax ¼ 1 (Figure
23). Thus, the equation for the line is
[2]
y ¼ 1
xmax  xmin x  xminð Þ
The above equation is used to express the vari-
able "x" necessary to calculate the limit points of
the individual ranges
[3]
x ¼ xmin þ xmax  xminð Þy
c. The values of “y1” to “y19” (0.05, 0.1, 0.15 to 0.95)
are then sequentially assigned to variable "y." The
variable "x" represents the coordinate we are trying
to find to the respective point "y" (Figure 23). The
value of the variable "x" calculated according to
Equation (2.3) is the limit point for the respective
descriptor range (Figure 24).
An illustrative example of 5 levels in a descriptor for
the height characteristic where xmin ¼ 1.81; x max ¼
4.26 and the value of “y1” to “y19” (0.05, 0.1, 0.15 to
0.95) were attributed to variable “y” (Tables 7–9).
Figure 24. Flowchart of the passport data structure.
Table 5. Descriptor for quantitative character – principle of
completeness.
Corbicula – height (mm)
1 Very small < 2.30
3 Small 2.30 2.79
5 Medium 2.79 3.40
7 High 3.40 3.89
9 Very high 3.89 and more
Table 6. Structure of open and closed intervals in descrip-
tor grades.
Corbicula – height (mm)
1 Very small < 2.30 (0;2.30)
3 Small 2.30–2.79 (2.30;2.79)
5 Medium 2.79–3.40 (2.79;3.40)
7 High 3.40–3.89 (3.40;3.89)
9 Very high 3.89 and more (3.89;1)
Table 7. Calculated value of variable “x.”
x1 x2 x3 x4 x5 x6 x7 x8 x9 x10 x11 x12 x13 x14 x15 x16 x17 x18 x19
1.93 2.06 2.18 2.30 2.42 2.55 2.67 2.79 2.91 3.04 3.16 3.28 3.40 3.53 3.65 3.77 3.89 4.02 4.14
Table 8. Determination of limit points in ranges for five descriptor levels.
x1 x2 x3 x4 x5 x6 x7 x8 x9 x10 x11 x12 x13 x14 x15 x16 x17 x18 x19
1.93 2.06 2.18 2.30 2.42 2.55 2.67 2.79 2.91 3.04 3.16 3.28 3.40 3.53 3.65 3.77 3.89 4.02 4.14
Standard methods for pollen research 27
4. Grading scales for qualitative characteristics. In the
qualitative characteristics with a nominal range (color,
shape, etc.), the classification grades are consecutive.
Their number depends on the observed variability in the
characteristic. The classification grade "9 – Other" is open
to the newly detected variability in the characteristic
(Table 10). In the case of ordinal scales (color intensity,
etc.), the classification grades are not consecutive but fol-
low the polarity of verbal characteristics (Table 11). In
alternative characteristics, grade "1" refers to the absence
of the given feature in the characteristic, and grade "9"
refers to the full presence of the feature (Table 12).
5. Polarity of verbal description. In the descriptors for
quantitative characteristics, it is necessary to follow
verbal polarization (Figure 25). This means that the
classification grade 1 represents something extremely
small and grade 9 something extremely large.
6. Descriptor name. The name of the descriptor should
be concise, it should accurately determine the evalu-
ation area and it should be integral. The name should
employ correct terminology and its structure should
be uniform throughout the list of descriptors
(Biodiversity International, 2007).
7. Evaluation methodology. The methodology should contain
the evaluation details, i.e., how much material from the
sample is to be assessed in order to get a representative
result, what the assessment period is, as well as other
information specific to the given characteristic (Table 13).
8. Drawings and pictures. It is appropriate that the meth-
odology be supplemented with concise diagrams,
drawings and pictures. These help the evaluator to
clearly and consistently determine the correct classifi-
cation grade in the descriptor and to avoid confusion
caused by undesirable side effects.
5. Identification of the floral origin
by microscopy
Pollinic analysis by microscopy is one of the first steps
and one of the very important because once the pollen
morphology is identified, it is possible to infer the char-
acteristics of the geographical regions where the BP
was obtained, as well as the habitat of forest, fields and
crops (Barth et al., 2010; Corbi et al., 1985; Dias et al.,
2012; Rodr
ıguez-Dami
an et al., 2006). This method is
based on two stages: the first consists in the identifica-
tion of the pollen grains that are under observation and
the second one aims to count them (http://www.apiser-
vices.com/abeille-de-france/articles/pollinic_analysis.htm).
Pollinic analysis allows scientists to infer the vegetation
present in the area and to date and ascertain any bio-
diversity changes, as for example, the presence and dis-
tribution of invasive or exotic plants. Also, allows the
scientists to distinguish if the pollen is monofloral, biflo-
ral or multifloral, or if the pollen grains present differ-
ent color means that belongs to different botanical
species (Nogueira et al., 2012). The application of
Pollinic analysis in paleoclimatic reconstruction has been
reported (Li & Flenley, 1999; Rodr
ıguez-Dami
an et al.,
2006). Some other studies revealed that BP analysis
provides important information about the plant prefer-
ences of honeys bees (Chaturvedi, 1973). Therefore, BP
Table 9. Ranges of 5 descriptor levels for height.
Grade Verbal characteristics Intervals
1 Very small < 2.30
3 Small 2.30–2.79
5 Medium 2.79–3.40
7 High 3.40–3.89
9 Very high 3.89 and more
Table 10. Descriptor for qualitative character
– structure of grades.
Corbicula – heart shape
1 Elliptical
2 Spherical
3 Heart-shaped
9 Other
Figure 25. Polarity in verbal description.
Table 11. Descriptor for qualitative character –
structure of grades in the case of ordinal scale.
Corbicula – cutout depth
1 Very deep
3 Deep
7 Shallow
9 Very shallow
Table 12. Structure of alternative character.
Corbicula – presence of visible impurities
1 Absent
9 Present
28 M. G. Campos et al.
Table 13. Complete descriptor with methodology and pictures.
Grade Verbal characteristics Figures
3 Smooth
3 5 7
5 Moderately rough
7 Rough
Methodology: Evaluate not less than 50 undamaged pollen corbiculas.
Table 14. Descriptor for characteristic – corbicula – height (mm).
Grade Verbal characteristics Range Figure
1 Very low
3 Low
5 Moderately high
7 High
9 Very high
Methodology: Evaluate at least 50 undamaged corbiculae, to two decimal places.
The Figure shows the height (in the direction of axis y) of a bounded object.
Table 15. Descriptor for characteristic – corbicula – width (mm).
Grade Verbal characteristics Range Figure
1 Very narrow
3 Narrow
5 Moderately wide
7 Wide
9 Very wide
Methodology: Evaluate at least 50 undamaged corbiculae, to two decimal places.
The Figure shows the width (in the direction of axis x) of a bounded object.
Table 16. Descriptor for characteristic – corbicula – Feret maximum (mm).
Grade Verbal characteristics Range Figure
1 Very small
3 Small
5 Medium
7 Large
9 Very large
Methodology: Evaluate at least 50 undamaged corbiculae.
Standard methods for pollen research 29
load analysis indicates the behavioral pattern and sense
of selection of plants for food (Sharma, 1970). Moreti
et al. (2002) referred that allows the identification of
the major pollen sources used by the bees, as well as
the periods of pollen production in the field and pos-
sible times of shortage. By another side, Arroyo et al.
(1986) referred that the determination of the flora
where the bee products are originating is certainly of
great interest to plan rationally the exploiting of natural
resources. This knowledge allows us to know the geo-
graphic origin of these products, so avoiding counter-
feits of origin (Ricciardelli, 1982). While these analyzes
(Melissopalynology) are important to the aspects men-
tioned above, it’s important to highlight that the process
of counting and classification of pollen grains are much
laborious and time consuming due to the fact that they
are done manually by highly skilled experts. Therefore,
a high level of training is required for obtain reliable
identification results (Rodr
ıguez-Dami
an et al., 2004,
2006). The classification of the different aspects of the
structure used to the analysis of pollen by microscopy
will be referred in Section 5.3 and complement the full
characteristics. Accordingly to the Dictionary.com (http://
dictionary.reference.com/browse/pollinic), pollen is
defined as “male grain gametophyte generation of seed-
bearing plants. In gymnosperms, each pollen grain also
contains two sterile cells (called prothallial cells), thought
to be remnants of the vegetative tissue of the male
gametophyte.”
5.1. Structure of the pollen grain
Pollen is a granular mass of male reproductive cells pro-
duced in the anthers of a flower. A pollen grain is a liv-
ing cell surrounded by two protective coats, the intine
and exine. The cell contains cytoplasm and nucleus. On
the surface are apertures or germinal pores or furrows.
The detail structure of a pollen grain and pollen grain
wall is shown in Figures 26 and 27. The inner layer or
intine is thin, delicate and very elastic. It is semi-perme-
able membrane and does not stain. When the grain is
seen in section, the intine can be recognized as a thin,
clear line surrounding the cell contents. The outer layer
exine is thicker, more brittle, and often variously sculp-
tured or provided with various modifications such as
spines, outgrowths or reticulations. The exine is made
of an extremely durable material called sporopollenin
and is composed by four layers, as listed below:
1. Base: This is a clear, uniform layer, the outer part of
which can be stained to reveal a dark line in the optical
section of the grain;
2. Rods or columns: These are arranged radially from
the base;
3. Tectum: This layer forms a roof over the rods. It may
be an incomplete layer, leaving some of the rods
free-standing;
4. Ornamentation. This is provided by a layer of spines,
outgrowths, reticulation, and other processes on
the tectum.
Usually, all the layers are not present. Those which
are present, show many modifications that are very use-
ful in identifying a particular pollen grain.
Apertures are present on the surface of pollen grain.
These are formed by the thinning or the absence of
same of the layers of the exine. They are called furrows
Table 17. Descriptor for characteristic – corbicula – Feret minimum (mm).
Grade Verbal characteristics Range Figure
1 Very small
3 Small
5 Medium
7 Large
9 Very large
Methodology: Evaluate at least 50 undamaged corbiculae.
Table 18. Descriptor for characteristic – corbicula – shape
index (width/height).
Grade Verbal characteristics Range
3 ellipsoidal <0.9
5 spherical 0.9–1.2
7 heart shaped 1.2 and more
Methodology: Evaluate at least 50 undamaged corbiculae.
Table 19. Descriptor for characteristic – corbicula – sym-
metry (Feret min/Feret max).
Grade Verbal characteristics Range
1 Unsymmetrical 0 0.81
9 Symmetrical 0.81 1
Methodology: Evaluate at least 50 undamaged corbiculae.
30 M. G. Campos et al.
Table 20. Descriptor for characteristic – corbicula – area (mm2).
Grade Verbal characteristics Range Figure
1 Very small
3 Small
5 Medium
7 Large
9 Very large
Methodology: Evaluate at least 50
undamaged corbiculae, to two
decimal places. The area is
determined by the boundaries
of the object by distinguishing the
limit color of the object from
the background.
Table 21. Descriptor for characteristic – corbicula – weight (mg).
Grade Verbal characteristics Range
1 Very small
3 Small
5 Medium
7 Large
9 Very large
Methodology: Evaluate at least 50 undamaged corbiculae.
Table 22. Descriptor for characteristic – corbicula – texture.
Grade Verbal characteristics Figures
3 Smooth
3 5 7
5 Moderately rough
7 Rough
Methodology: Evaluate not less than 50 undamaged pollen corbiculas.
Table 23. Descriptor for characteristic – corbicula – shape.
Grade Verbal characteristics Figures
1 Irregular
2 Spherical
3 Ellipsoidal
4 Heart shaped
9 Other 1 2 3 4
Methodology: Evaluate not less than 50 undamaged pollen corbiculas
Table 24. Descriptor for characteristic – corbicula – heart shape.
Grade Verbal characteristics Figures
1 Ellipsoidal
1 2 3
2 Spherical
3 Heart shaped
4 Other
Methodology: Evaluate not less than 50 undamaged pollen corbiculas.
Standard methods for pollen research 31
or colpi, if they are elongated and tapering toward the
end or pores when they are round or oval (Figure 28).
The furrows run longitudinally with their ends toward
the polar areas. The apertures are generally situated on
the equator but pores are often found scattered over
the whole grain surface. Pores and furrows often occur
together, but at different levels of the exine. Apertures
allow the grain to dry or to absorb water and thus to
change into an expanded state. Also provide an easy
outlet for the pollen tube when the grain germinates.
Therefore, they are also called germ pores or germi-
nal furrows.
5.2. Floral origin
Considering the bee pollen loads, pollen may be classi-
fied as:
1. Unifloral, also named monofloral, when this product is
derived from a single botanical source (Campos
et al., 2008);
2. Bifloral when obtained from two species of bee plants;
3. Multifloral, also named heterofloral or plurifloral,
when the pollen originates from several botanical spe-
cies (Barth, 2004; Campos et al., 2008; Sharma, 1970;
Stanley & Linskens, 1974).
5.3. Features of pollen grains
Each pollen grain has its own peculiar structural fea-
tures (size, shape, apertures type, exine sculpture (exine
ornamentation) and surface and other features), which
helps us recognize its origin. Scientists-palynologists
Sawyer (1981) and Vorwohl (1990) have different opin-
ions of these features. Some of them will be show along
the text.
5.3.1. Size
The size (diameter) of the pollen grain is expressed in
microns or micrometers (mm), where 1 mm ¼ 1 / I
000mm. The maximum diameter of a pollen grain
includes the spines or other ornamentation. Sawyer
(1981) used five size classes:
1. Very small (<20 mm);
2. Small (20–30 (mm);
3. Medium (30–50 mm);
4. Large (50–100 (mm);
5. Very Large (>100 mm).
However, not all pollen grains are round. Many are
either elliptical or triangular, even elongated. Vorwohl
(1990) suggested that instead of defining size by the
diameter, it should be defined by the length and breadth
and defined nine classes of length and width (Table 28).
5.3.2. Shape
Shape is the outline; the optical section seen under the
microscope and has been considered as an important
feature in pollen identification (Sawyer, 1981). A pollen
grain can be round, oval, elongated, triangular, semi-cir-
cular or boat-shaped, and irregular or multi-sided. The
shape plays a very minor role in the identification of
pollen grains, because it varies depending upon the
Table 25. Descriptor for characteristic – corbicula – cutout depth.
Grade Verbal characteristics Figures
1 Very shallow
3 Shallow
7 Deep
9 Very deep
1 3 7 9
Methodology: Evaluate not less than 50 undamaged pollen corbiculas.
Table 26. Evaluation of corbicular color (No^zkov
a, Brindza et al., 2010).
Sample no. Color (Visual assessment) Color group name (RHS) Color (RHS)
2 brownish yellow gray-yellow 162 A
5 grayish green gray-green 194 A
66 brown brick gray-orange 165 B
80 lemon yellow yellow 8 A
81 brownish yellow gray-yellow 162 A
90 brownish yellow gray-yellow 162 A
98 brownish yellow gray-yellow 162 A
99 brownish yellow gray-yellow 162 A
100 carrot orange orange 24 A
113 brownish yellow gray-orange 164 B
32 M. G. Campos et al.
Table 27. Descriptor for characteristic – corbicula -color.
Grade Verbal characteristics
1 Black
2 Purple
3 Brown
4 Green
5 Orange
6 Yellow
7
8
9 Other
Figures
2 purple hue
3 brown hue
4 green hue
5 orange hue
(Continued)
Standard methods for pollen research 33
position in which the grain lies and upon the viewing
aspect (polar view or equator view). Many grains appear
round in one particular aspect. If they also show special
characteristic shapes, these have been given preference
in the key. These alternative shapes should be used for
identification. In addition to the definitely triangular or
tri-lobed appearance of many grains, other 3-furrowed
grains may show a shape intermediate between rounds
and triangular. Both these classifications should then
be considered.
5.3.3. Apertures
Pollen grains have apertures on the surfaces. The shape
(type) and the numbers of apertures are important
characteristics of a pollen grain and can be used in iden-
tification. Furrows containing pores are counted as one
aperture. In compound pollen grains, the numbers refer
to each single grain sub-unit. Nine classes of aperture
number (as given in Table 28) can be used in pollen
identification. Regarding aperture type, Vorwohl (1990)
used the following five classes in pollen identification
(Table 28). Whereas Sawyer (1981) recognized only
four classes of aperture types to be important in pollen
identification as:
1. Pores only;
2. Furrows only;
3. United pores and furrows;
4. Irregularly occurring furrows.
5.3.4. Exine sculpture (exine ornamentation) and surface
The exine, the outermost coat of the pollen grain, is
either smooth or various ornamented with different
outgrowths or projections, such as granules, various
kinds of spines, and reticulum or window-like struc-
tures. The ornamentation of the exine has a major
Table 27. (Continued).
Grade Verbal characteristics
6 yellow hue
Methodology:
1. Visual assessment: the most frequent corbicula color is taken into account and the corbicula is assigned its respective
classification grade.
2. Evaluation using RHS: the methodological rules set out in this classification system are observed. After determining the color
of the sample, the respective code (e.g., 164B) is recorded.
Figure 26. Diagrams showing the structure of a pollen grain in
optical section (Sawyer, 1981).
Figure 27. Diagrams showing details of pollen grain walls
(Sawyer, 1981).
Figure 28. Diagrams illustrating the morphological regions of a
pollen grain and pollen grain apertures: (a) equatorial view; (b)
polar view (Sawyer, 1981).
34 M. G. Campos et al.
effect on the surface of pollen grain as viewed under
the microscope. The surface of pollen grains, as seen
under the microscope, can sometimes change with a
change in focus. For example, the pollen grain of Luffa
cylindrica may appear granular under one focus and
reticulate under another.
1. Nine classes of exine ornamentation can be identify:
thin; medium, no rods; medium with spaced rods or
beaded; medium or thick with coarse external rods;
layer of closed thin rods; long thin spines; large broad-
based spines; small or very small spines; or warts and
other projections.
2. Five classes of surface as: smooth or indefinite; granu-
lar or warts; striated; net or pitted and Isolated dots
due to spines or other projections. (exine ornamenta-
tion can also be divided into seven classes of exine
sculpture – Table 28):
a. Psilate, Faveolate and Fossulat: When exines are
smooth, they are known as psilate, e.g., Betula
spp and Pyrus spp. Exine sculpture with little pits
is reported as faveolate and is known as fossulate
with relatively larger pits. In viewing the surface,
the pollen grains with psilate, faveolate, and fossu-
late exine sculpture appear smooth or some-
times indefinite;
b. Scabrate, Verrucate and Gemuate: Exines with little
warts and diameters of not more than one pm
are scabrate, e.g., some species of Quercus. In
verrucates, the diameter of the warts is more
than one pm, e.g., members of the Ranunculaceae
family, some Asteraceae, and Nigella spp. Gennuate
is an exine sculpture with round warts which are
narrower at the base, for example, Ilex spp. In
optical sections, this kind of exine appears to
have small projections. Pollen grains with this
kind of exine sculpture appear granular in viewing
the surface;
c. Echinate: Exines provided with different kinds of
spines, e.g., Malvaceae, Cucurbitaceae,
Campanulaceae, and many Asteraceae are echinate.
Pollen grains with echinate exine sculpture appear
to have dots on the surface;
d. Clavate and Bacculate: Exines with rods which
have thicker ends are clavate and those with
stick-like rods are bacculate. Pollen grains with
clavate or bacculate exine sculptures are very
rare. On examining the surface, the pollen grains
with this kind of exine sculpture appear granular
(they have larger grains than scabrates, verru-
cates, and gennuates);
e. Rugulate and Striate: Exine sculptures with stria-
tions, e.g., species of Acer, Prunus, Datura, and
Fragaria, belong to these categories. In optical
section, the exines appear to have close, thin
rods. On examining the surface, the pollen grains
also appear striated;
f. Reticulate: This term is used when the exine
sculpture is reticulate or net-like. This category is
divided into fine reticulate, or beaded, or coarse
reticulate. Examples are Lilium spp, Hedera helix,
and Ligustrum spp. In optical sections, reticulate
exines appear to have either thin or coarse rods,
or they appear to be beaded. Pollen grains with
reticulate exine sculpture appear reticulate on
the surface also;
g. Fenestrate: Exines in which there are window-like
holes between the ribs of the exine belong to
this category. Important examples include
Taraxacum officinale (Ap.: and members of
Amaranthaceae). On the surface, the pollen grains
appear to have window-like holes.
5.3.5. Other features
According to Vorwohl (1990), aggregation of pollen grains
is one of the important features in identifying pollen
grains. Pollen grains of some plant species have other
diagnostic features. For example, the pollen grains of
many plant species in the family’s Ericaceae and
Table 28. Diagnostic pollen grain features used for identifica-
tion (Vorwohl, 1990).
1. Length and breadth 1 <10 mm
2 10–15mm
3 15–20mm
4 20–25mm
5 25–30mm
6 30–35mm
7 35–40mm
8 40–50mm
9 >50 mm
2. Aperture number 1
2
3
4
5
6
7
8 >7
9 Absent or not clear
3. Aperture type 1 Pore
2 Colpa
3 Heterocolpa
4 Syncolpate
5 Heterocolporate
4. Exine Sculpture 1 Psilate, Faveolate, Fossulate
2 Scabrate, Verrucate, Gennuate
3 Echinate
4 Clavate, Bacculate
5 Rugulate, Striate
6 Reticulate
7 Fenestrate
5 Aggregation 1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 >7
Standard methods for pollen research 35
Leguminosae (sub-family, Mimosoidae) are polyads, i.e.,
compound. In Ericaceae, four grains are aggregated to
form a tetrad, e.g., Rhododendron spp., whereas in the
Leguminosae, eight, twelve, or sixteen pollen grains are
aggregated to form a polyad, e.g., Acacia spp, Albizia spp,
Calliandra spp, and Mimosa spp. Also another important
feature for pollen determining as: Air sacs; Thickened or
projecting edges to apertures; A cap or streak on the
apertures; Granules or projections scattered on the
apertures; Intine swollen below the apertures; Intines
thick or very thick and cell content’s granular has been
reported by Sawyer (1981).
5.3.6. Irregular grains
In the study of pollen found abnormal, not typical,
grains. Sawyer (1981) allocates such grains in the follow-
ing groups:
1. Shriveled or abortive grains. Particularly in Rubus
(blackberry), Trifolium repens (white clover) and many
garden cultivars (e.g., Saxifraga and Oxalis) we find
shriveled grains which do not expand when mounted.
Such grains may even form the majority in a sample;
2. Giant and dwarf grains. Many pollen samples show the
occasional giant grain; much larger than usual and
often irregular. Others, for example Plantago (plantain)
and Fagopyrum (buckwheat), produce a number of
unusually small perfect grains among their pollen;
3. Aperture numbers. Variants with 4 furrows are occa-
sionally found among 3-furrowed grains. This is par-
ticularly noticeable in some of the Rosaceae (e.g.,
blackberry and fruit blossom), Rumex (dock) and
Hippophae (sea buckthorn). Regular variation of the
number of pores is found in Carpinus (hornbeam) (3, 4
or 5); Alniis (alder) (4, 5 or 6) and Epilobium angustifo-
lium (rosebay willowherb) (3 or 4);
4. Abnormal furrows. An interesting variation occurs in
Veronica (speedwell) and occasionally in other
Scrophulariaceae. This results in a square grain (PI.
188). Irregular and united furrows are common in
Hypericum while the 'furrows' of Berberis (barberry),
Crocus and Populus (poplar) appear as irregular breaks
in the exine rather than true furrows (Crocus). The
study of atypical pollen grains used for palinoindication
of environment. (See also the paragraph regarding to
Palinoindication of environment).
5.3.7. Pollen “types”
Although the pollen from each species of plant differs
from that of all others, the differences are often slight
and with the multitude of plant species and hybrids in
the world, it will be appreciated that we cannot expect
to identify each one by its pollen alone. Here they are
group according to their “type.”
The pollen of some species is unique in its charac-
ters. Thus, Trifolium repens (white clover) and Trifolium
pratense (red clover) may be distinguished from one
another and from all others. But more often we find a
“type” of pollen common to a genus or other group of
plants. The pollen of all Tilia (lime) is unmistakable and
although careful study may indicate the particular
Figure 29. Photos to illustrate the variety of colors of pollen loads of different species of plants: (a) gray – Vicia faba; (b) lemon yellow
– Brassica napus; (c) dark gray – Papaver rhoeas; (d) blue – Phacelia tanacetifolia; (e) mixed yellow pollen – mixed pollen from fruit trees
(Prunus, Pyrus, Malus) and bright orange – Taraxacum officinale.
36 M. G. Campos et al.
species, the classification “Tilia type” is usually sufficient
in limiting.
5.4. Colors of pollen
The colors of bulk pollen, dry from the anthers or the
pollen loads of bees, have been arranged in spectral
order. They show considerable variation but may at
times provide a valuable clue to identity (Hodges, 1952;
Kirk, 1994; Partap, 1997; Sawyer, 1981). Hodges (1952)
and Kirk (1994) consider that the first steps for identifi-
cation of pollen loads are
1. sorting by color;
2. selecting one or two pollen loads from each
color group;
3. pollen loads are put on white paper and lightly patted
to make a uniform flat surface. This is necessary in
order to find an exact match to a known color chart.
The nice color charts in Hodges (1952) and Kirk
(1994) illustrate the different color pollen loads of dif-
ferent plants. The Figure 29 presents an example of the
different colors of pollen loads. Important is the lighting,
in which the identification is carried out. Pollen loads
that are to be identified must therefore be matched to
the color chart under these same standard conditions.
If the chart is used in direct sunlight or under electric
light, the result may well be different. According to Kirk
(1994), many plant species have pollen loads with very
similar colors, so the particular species from which a
pollen load has been collected can rarely be identified
from just the pollen color. However, a combination of
the color, flowering time and knowledge of local plants
can usually narrow down the possibilities to one or a
few species. However, this identification cannot be
accurate. Therefore, for more exact identification
microscopic examination pollen grains is necessary.
For instance, it has been proved that the color of
pollen influences the color of honey; pollen and pollen
loads from Taraxacum and Helianthus annus are bright
orange – the color of pure honey with this plants – also
is bright orange; pollen from Phacelia – blue, pure honey
from Phacellia as a pale-blue color, pollen loads from
Aesculus hippocastanum give a burgundy, honey with red-
dish color, etc. Most of the loads collected by honey
bees consist only of pollen, but bees also collect prop-
olis and occasionally fungal spores and various dusts
(Hodges, 1952).
5.5. Methods of pollinic pollen analysis
5.5.1. Types of microscopy examination
The exact definition of the botanical origin of bee prod-
ucts, especially honey and bee pollen, requires the use
of microscopic examination. The literature describes
three methods of microscopic procedures.
5.5.1.1. Light microscope.
 Use a light microscope at a magnification of which is
100, 500, 1000 makes it possible to determine the
external characteristics of the pollen grain: the shape,
size, surface sculpture, and the presence of pores, fis-
sures, etc.). And identify it up to the family, genus and
species (Tripathi et al., 2003). But this method has its
drawbacks: the internal structure of the grain is
unattainable because of the opacity of the outer shell
(exine) and pollen grains of some plants are too large
and therefore it is impossible to get a clear image.
 Preparation of pollen analysis for microscopic prepara-
tions. There are several techniques, which differ mainly
in the medium fixing pollen grains turpentine, alcohol,
honey solution or acetolized method (Dustman, 1992).
Since the methods of preparation of pollen in micros-
copy greatly affect the appearance of pollen grains and
noted the difference in the results of analysis, it is
advisable to reference preparations of pollen grains and
pollen description made in a relatively simi-
lar procedure.
Louveaux et al. (1978) proposed an affordable and sim-
ple method for the preparation of dust melisopalinology
research involves the use of glycerol-gelatin. This
method is used in many countries.
5.5.1.2. Transmission electron microscope (TEM).
 Microscopic examination of pollen using a transmission
electron microscope, which allows the study the ultra-
thin sections of pollen grains;
 The image has pollen with unlimited depth transparency
(X-ray allows us to study the internal structure of an
object is significantly different than the other methods)
(Kedves et al., 2002, 2003; Laere et al., 1969).
5.5.1.3. Scanning electron microscope (SEM) (three-
dimensional image). Accuracy to distinguish varieties
even within the same species of plants. Especially for
minor differences in the size and structure of the sur-
face of pollen grains of different varieties of fruit crops
(Brindza & Brovarskyi, 2014; Fogle, 1977; Klungness &
Peng, 1983; Lindtner, 2014).
At the present stage of research melisopalinologiy
widely practiced a combination of all methods of
microscopy (Tripathi et al., 2003). For carrying pollen
analysis must have reference preparations of pollen
grains of the plants that could potentially occur in the
samples of pollen, keys and atlases pollen. Help in defin-
ing various botanical origin keys pollen grains photo-
graphs, drawings. Modern pollen atlas published in 2004
by German melisopalinology. It contains pictures and
descriptions of pollen about 170 species of plants and
honey flora characterizes his country (Bucher et al.,
2004). Several studies focused on computerized texture
analysis of Scanning Electron Microscopy (SEM) images
have been published (Langford et al., 1990; Li &
Flenley, 1999).
Standard methods for pollen research 37
5.6. Microscopic techniques
Identification of pollen is carried out in 3 stages.
5.6.1. Preparation of pollen grain slides from pollen loads for
microscopic research
Method of preparation of pollen grain slides based on
the method of mellisopalinology Louveaux et al. (1978).
1. Take 1mg of bee pollen loads with tweezers and
transfer to a glass slide, add a drop of distilled water,
which cement pollen grains will be separate each
other. The amount of water is such that pollen grains
freely melt away in a single layer on a slide.
2. Evaporate the water.
3. Apply one drop of warm (þ40 C), heat by thermal
table glycerin-gelatin and cover with cover slip.
4. Preparation of glycerin jelly. Weight 10 g of gelatin and
place them place for two hours in distilled water (gel-
atin absorbs 34.2ml of water) and add 66.3 g of gly-
cerin. This ratio of ingredients allows to save pollen
grain shape unchanged. To prevent microbial growth
in glycerol-gelatin add 0.5 g of carbonic acid (phenol)
and to get a clear picture of the structures of pollen
grains – 0.1 g Basic Fuchsine. 12 hours later for the
prevention of further drying up of the slides aids its
conservation – edge of the coverslip treat with dec-
orative lacquer (nail polish).
5.6.2. Identification of pollen under microscope
Determination of botanical species of pollen can be
done using the computer program identification of pol-
len LUCIA (Rodlauer & H€uttinger, 1992) (or similar).
As an example, Digital camera Sony with light micro-
scope Nicon transfer image of pollen on the computer
screen. The operator determines the size, shape, num-
ber of pores, fissures and other parameters of the pol-
len grain – creates a digital key for determine (details in
Introduction). Using this key, another computer pro-
gram (electronic database of pollen) "Pollen Data Bank,"
provides a search and similar pollen grains select and
begins clarification of affiliation pollen to a particular
plant species. For comparison also use own standard
slides and/or slides from other collections as, for
example, the collection of the Austrian Institute
of Beekeeping.
Determination of the botanical origin of pollen will
be performed by comparing investigational pollen grains
with descriptions and photographs of existing atlases
and reference books, as for instance: Ambruster and
Oenike (1929), Burmistrov and Nikitina (1990), Faegri
et al. (1993), Fossel and Pechhacker (2001), Hodges
(1952), Maurizio and Louveaux (1965), Sawyer (1981)
and Zander (1935).
Microscopic magnification of X 450 (Zea mays) and
	1000. For the magnification of 	1000 used oil
immersion (cedar oil). A calibration line representing
10mm should be add to each photograph.
5.6.3. Description of pollen grains
Description submitted by the following scheme:
1. Latin name of the plant;
2. The family of plant;
3. Plant life form (tree, shrub, herb; an annual or peren-
nial plant);
4. Morphological characteristics of pollen grains:
a. size of the diameter and the equator;
b. number of furrows, pores;
c. surface and exine sculpture;
5. Color of bee pollen loads.
5.7. Palinoindication of environmental pollution
The knowledge of the morphology of pollen grains from
different species (plants) allows simultaneous pollen ana-
lysis of bee products, also to carry out environmental
monitoring, determine ecologically disadvantaged
regions. It has been proved that the formation and
development of pollen is very sensitive to external fac-
tors (radiation, volcanic activity, solar activity, heavy
metal contamination, etc.) (Bogdanov et al., 2005).
The result is a significant amount of pollen grains with
modified morphological structures. Research shows that
in harsh conditions, plants may produce large amounts
of atypical imperfect, irregular and abnormal pollen
grains - the worse the state of the environment, the
greater the percentage of abnormal pollen and vice
versa. (Bessonov, 1992; Dzyuba, 2006, 2007; Dzyuba &
Tarasovich, 2001; Gilber & Lisk, 1978; Glazunov, 2001;
Muszynska & Warakomska, 1999).
5.8. Concluding remarks
Research using light and scanning electron microscopy
show the changes in size and shape of the pollen grains,
the number, the shape and type of apertures, their size
and location relative to each other. Main thing changes
the most stable structure of pollen grains – sporoderm
surface structure (shell), the number and thickness of
the layers. Russian scientists have developed a method
for determining the quality of the environment by the
number of abnormal, atypical pollen grains (Dzyuba,
2006). In selected samples investigated at least 200 pol-
len grains, each of which is examined by the follow-
ing parameters:
1. The form of the pollen grain;
2. The size of the pollen grain;
3. The number of apertures;
4. The nature of the location and type of apertures;
5. The thickness and number of layers sporoderm;
6. The nature and type of sculpture sporoderm;
38 M. G. Campos et al.
7. Symmetry breaking pollen grain.
Thus, each sample was fixed in the presence/absence
teratomorfic (pathologically developed) pollen grains,
their quantity and analyzed morphological features. The
Figure 30 illustrates the typical, normal pollen grains.
Pollen grains have no deviations in the morphological
parameters (size and shape of grains, number, shape
and type of apertures, their size and location relative to
each other, the sculpture's surface the number and
thickness of layers). In the Figure 30 presents the pho-
tographs of pollen grains morphological features which
correspond to norm. Photos made with the help a scan-
ning electron microscope (Figure 30) and the Figure 31
shows some pollen grains (scanning electron micro-
scope) with changed morphological structures.
Particularly large number of abnormal pollen grains in
Phacelia tanacetifolia (atypical shape and size) (Figure
31a). Pollen grains of Taraxacum officinale observed sym-
metry breaking (Figure 31b) (Lokutova, 2014).
Therefore, numerous experiments have shown that
during pollen analysis of bee products to determine
their botanical and geographical origin there is a possi-
bility at the same time estimate the state of the
environment. According the quantitative content abnor-
mal pollen grains can draw conclusions about environ-
mental situation in the region, where these bee
products were received.
6. Phenolic and polyphenolic profiles as
fingerprints of floral origin of bee pollen
Rapid and easier method developed by Campos et al.
(1997) use High Pressure Liquid Chromatography/
Diode Array Detector (HPLC/DAD) profiles, of phen-
olic and polyphenolic compounds, made with hydroalco-
holic extracts of pollen and bee pollen loads. It is easy
to do, reliable, highly effective, and requires minimum
sample preparation, as indicated by the several reports,
from worldwide, in which the method has been used to
analyze the phenolic composition of pollen. This method
allows the identification of the genus and the species
(sp.) and provides valuable information related to chem-
ical structure of the constituents under analysis. The
phenolic acids and polyphenolics in pollen are expressed
as specie-specific, which represents an important tool
for taxonomy studies. It should be do a screening of dif-
ferent color pellets from a mixture of bee pollen and
the identification of the potential floral source using a
Figure 30. Typical (normal) pollen grains (scanning electron microscopy): (a) Aesculus hippocastanum; and (b) Vicia faba.
Figure 31. Pollen grains with changed morphological structures (scanning electron microscopy): (a) Phacelia tanacetifolia; (b)
Taraxacum officinale; and (c) Scilla bifolia.
Standard methods for pollen research 39
data base build with floral hand collect pollen or herbar-
ium specimens. The first approach to the flavonoid
structures can be based on Campos and Markham the-
ory (2007).
6.1. Materials and methods
6.1.1. Sample management
Sample management is an important issue for the ana-
lysis. It is essential to specify details about where and
when specific pollen samples are collected, sample stor-
age, handling, and preparation, as details of any of these
aspects might affect the phenolic composition of pollen.
More details found in Section 2 of this chapter.
6.1.2. Sample collection
6.1.2.1. Pollen hand collected – reference
samples. Reference samples should be collected from the
specimens at the Herbariums if they show a very good
preservation or they can be hand collected from the sta-
mens of flowers at the field and bring to the laboratory.
The correct taxonomic identification should be assured.
The HPLC/DAD fingerprints of the phenolic and poly-
phenolic compounds of these well-known samples will be
used as a data base for match of the unknown samples
of bee pollen. For that, it is important to:
1. record location of harvesting;
2. record date of sampling;
3. collect the respective voucher specimen, which must
to be deposited at herbarium to be identified or
authenticated.
6.1.2.2. Bee pollen pellets/loads. Samples collected dir-
ectly from bee pollen traps or purchased from beekeep-
ers during the collection period. Information about the
apiary locations and harvesting data is always welcome,
as well the processing of drying, if it was applied.
6.1.3. Methods
6.1.3.1. HPLC/DAD analyses.
6.1.3.1.1. Sample preparation (hand collected pollen and/or
pollen from herbarium).
1. Weight ten milligrams of dried pollen of each sample.
If the sample is from herbarium lower amounts can be
used but the proportion 10mg to 1000 ml should be
maintained.
Note: Using always the same concentration among
samples makes easy to compare the approximate
amount of phenolic and polyphenolic compounds in
the fingerprints, without the need of quantification. If
need, a rigorous determination of the total amounts
of each compound can be done, using a calibration
with standard solutions.
2. Add an ethanol (EtOH)-water solution (1ml, 50%
V/V).
3. Sonicate for 60min.
4. Centrifuge the resultant extract at
5000 rpm for10min.
5. Take out the supernatants and do a microfiltration
with a 0.45 mm membrane for HPLC/DAD analysis as
previously described (Campos et al., 1996).
6.1.3.1.2. Sample preparation (bee pollen). Usually bee
pollen samples are made up of different color loads/pel-
lets and set up a “multifloral bee pollen.” However,
“monofloral” samples are also found in the market and/
or produced with this purpose according to Campos
et al. (2008).
1. Separate pollen loads by colors (Figure 32).
Note: commonly each color corresponds to a botan-
ical origin and, for each one, it will be carried out a
HPLC/DAD profile (fingerprint of the taxon). This fin-
gerprint will be compared to others in the data base to
obtain the match with a similar one.
2. Weight one pellet, keeping the same ratio for extrac-
tion (10mg/ml EtOH 50%), for instance, if the weight
is 6.78mg the solvent will be 0.678ml of EtOH 50%
(V/V).
3. Proceed the same way for mixtures of bee pollen. The
extraction should keep the same ratio. If necessary, du
an aliquot that represents the entire weight of the
sample under analysis, for example, 10 g/l EtOH 50%
(V/V). The portion of the mixture to be weighed is
removed after sample homogenization.
4. Proceed as explained in points 3–5 in sSection 6.1.3.1.1.
Note: In addiction a standard of quercetin-3,5,7,30,40-
O-pentamethyleted can be add to samples for HPLC/
DAD injection, as an internal standard (Campos &
Markham, 2007).
6.1.3.1.3. HPLC/DAD analysis. For the HPLC/DAD pollen
analysis several steps must be performed rigorously to
have a correct identification of the compounds in the
different samples with different origins:
1. HPLC/DAD system with a RP18 column;
2. Eluent formed by acidified water with ortho-phos-
phoric acid to pH 2.4 (solvent A) and acetonitrile
HPLC grade (solvent B), combined in a linear gradient
according to the following: starting with 100% A,
decreasing to 91% over the next 12min, to 87% over
the next 8min, to 67% over the next 12min, to 57%
over the next 10min, and holding that last proportion
until the end of analysis (a total of 60min).
Note: This gradient was optimized to avoid sample
purification, which decrease time of preparation, sample
manipulation and possible loose or degradation of com-
pounds to be used on the fingerprints for each taxon.
40 M. G. Campos et al.
Among phenolic and polyphenolic compounds many
other polar constituents will be extracted with the
hydroalcoholic solvent used; however, the contamin-
ation does not interfere with the HPLC analyze because
at the beginning a water gradient is used and the com-
pounds that will be separated stay in the top (head) of
the reverse phase column choose, allowing the more
polar, i.e., amino acids, proteins, sugars, hydrosoluble
vitamins and minerals to elute in the first minutes with-
out interfering in the fingerprint. Phenolic and poly-
phenolic compounds only start to be eluted 15min
later with the increase of acetonitrile in the gradient;
3. Flow rate 0.8ml/min;
4. Temperature stated at 24 C;
5. Injection of the samples extracts (20 to 100 ml,
depending on phenolic concentration);
6. Standard chromatograms plotted at wavelengths 260
and 340 nm (the choice of these two wavelengths con-
sist in the knowledge that flavonols and flavones glyco-
sides, the main compounds in pollens, have two
characteristic bands of ultraviolet – absorption around
these values; phenolic acids and its derivatives can also
be identified using the same range);
7. Spectral data for all peaks is accumulated in the range
220–400 nm (the phenolic/flavonoid spectra are all in
this display) using DAD;
8. Apparatus: HPLC system, Diode Array Detector with
respective software;
9. The phenolic/polyphenolic profile of each sample is
made up with all compounds resolved in its respective
chromatogram;
10. The structural information for each compound is made
by direct comparison of retention time and ultraviolet
absorption spectra with standards, and according to
the theoretical rules developed and compiled by
Campos and Markham (2007).
Note: the compounds can also be isolate and the
structures determined by Nuclear Magnetic Resonance
(NMR) and/or LC-MSMS (Campos et al., 1996, 1997).
Briefly: The extracts are applying in sheets Whatman
3MM chromatography paper for 1D separation. Each
compound is prepared for further studies by chroma-
tography on a C-18 reversed phase column using water:
acetonitrile gradient and checked for purity by HPLC/
DAD, according to Markham and Campos (1996). For
the full structure elucidation by NMR spectra can be
carry out in Bruker AC apparatus at 300MHz (or simi-
lar) with the isolated compounds, solved in DMSO-d6,
with TMS as intern standard at 30 C (Markham &
Geiger, 1993).
6.2. Theoretical rules for structural determination
of the main phenolic acids and flavonoids
The identification of the compounds is carried out by
direct comparison of retention time and ultraviolet
absorption spectra (Campos & Markham, 2007). In
Figure 33 flavonoid structures are shown to a better
understanding of the explanation of the UV spectra
related to the position of the substitution in the mol-
ecule. The majority of the flavonoids in pollen are deriv-
atives of flavones and/flavonols. The majority of the
flavones found in pollens are apigenin or luteolin deriva-
tives, and the compounds with a flavonol nucleus, are
mainly kaempferol, quercetin or isorhamnetin. In the
most samples these compounds appear as glycoside
derivatives. Examples of different flavonol and flavone
structures and its respective UV-spectra are show in
Figure 34(a–g). Isoflavones, as daidzein, glycitein and
genistein derivatives can also be found in certain pollens
but are not so common.
Examples, of rules developed in Campos and
Markham (2007) for identification of details in the
UV-spectra:
1. In all of them can be observed a clear existence of two
intense bands of absorption which indicate the presence
of the rings A, B and C from flavonoids. In Figure 34a
are pointed the three important absorption points:
a. Band I, usually with absorption around
340–390 nm. Flavones 340–350 nm; Flavonols-3-O-
derivatized (Table 29) (as glycosides, for instance)
Figure 32. Separation of bee pollen by color pellets from a
sample that include different floral origin pollens.
Figure 33. Basic flavonoid structure in order to show a better
understanding of the explanation of the UV spectra related to
the position of the substitution in the molecule (From Campos
& Markham, 2007. Reproduced with permission).
Standard methods for pollen research 41
350–360 nm; Flavonols with the C-3-O-H
free 370–390 nm;
b. Band II, usually with absorption
around 240–280 nm;
c. shoulder between 280 and 310 nm – substitution
in C7;
The absence of the shoulder between 280 and
310 nm is indicative of a substitution in C7-OH.
Example, flavones as Apigenin (Figure 34a) and apigenin-
7-O-derivative (Figure 34b). This –OH is free in
Apigenin itself with no substitution, and the shoulder
remain in the UV-spectra.
Figure 34. Examples of the more common UV- spectra found in bee pollens: (a) apigenin; (b) apigenin-7-O-derivative; (c) luteolin; (d)
kaempferol-3-O-derivative; (e) quercitin-3-O-derivative; (f) isorhamnetin-3-O-derivative; (g) chalcone.
42 M. G. Campos et al.
Example, flavonols as Kaempferol-3-O-derivatives
(Figure 34d) and Kaempferol-3,7-O-di-derivatives (sub-
stitution in C7-OH) (Figure 34e). The same absence of
the shoulder at 280–310 nm can also be seen for
Quercetin-3,7-O-di-derivative (Figure 34g). When the
C7-OH is free the shoulder is present at 280–310 nm
(a flat region in this part of the spectrum), examples of
luteolin, Kaempferol-3-O-derivative and Quercetin-3-O-
derivative (Figures 34c, d and f, respectively – all have
the C7-OH free);
2. Differences between isorhamnetin and quercetin:
Comparing isorhamnetin-3-O-and quercetin-3-O-deriva-
tive, the difference in the structure is in B-ring with a
methyl group in C3’ hydroxyl in the first. This implies
an intensification of Band IIb and IIa (wavelengths of
UV absorption for both around 255 and 265shoulder
nm). Quercetin show a smoother absorption among
these two bands;
3. To distinguish both structures, flavones or flavonol 3-
O-substituted the relative intensity of absorption
among Band I and Band II should be analyzed. If Band I
is higher than Band II, usually are flavones but in the
inverse case, Band I lower then Band II is a high prob-
ability to be a flavonol-3-O-substituted (glycosylated in
the majority of the times) (examples in Figure 34);
Note: More details for UV interpretation of the
spectra in Campos and Markham (2007).
4. In Tables 30 and 31, the wavelengths related to the
oxygenation pattern, for the more common flavones
and flavonols can be consulted and there, a prevision
of the bathochromic effect (relating to a shift to a lon-
ger wavelength in the absorption spectrum of a com-
pound) in band I with the increase of the oxygenation
pattern on B-ring in these structures. However, when
these changes occur, usually the UV-absorption of
Band II remain the same (Table 31).
6.3. Quantification of the major floral specie (floral
source) in a mixture
With this method it is also possible the identification of
the major taxon source in mixtures of various floral spe-
cies (different color loads/pellets in general correspond-
ing to various floral sources), if one of them is in a
high percentage.
1. Analyze the phenolic HPLC/DAD profiles (finger-
prints) for the different taxon found in the mixture.
2. In each taxon select the compound that can be used
as a marker for this specie. The compound used as
marker should not be found in the other floral species
in the full mixture. Note: This approach is supported
by microscopic analysis. The flavonoid/phenolic HPLC/
DAD fingerprint of the full mixture should be the
same as the profile corresponding to the major floral
specie, i.e., if the major floral source in the bee pollen
sample (batch) is Salix atrocinerea, the fingerprint
match with the one obtained with same floral source
in the data base made up with the hand/herbarium
pollen sample.
3. To achieve the percentage of this major monofloral
identification in a bee pollen mixture:
a. Run HPL/DAD profile of the mixture and com-
pare to the data base to verify if it is the same as
one of them;
b. Once the floral specie is identified choose one of
the major compounds as a marker and its quanti-
fication will correspond to 100%. Note: analyze
the fingerprints obtained with the various pellets
(other floral origins) in the mixture and choose a
compound that does not belong to them;
c. Quantify the same compound in the mix. The
amount determined will be a % of the marker
and corresponds to the percentage of the major
specie in the entire batch. In Figure 35 an
example involving a mixture of bee pollen with
about 60% of Eucalyptus globulus is shown. In this
case, quercetin-3-O-sophoroside (di-glycoside)
was chosen for quantification and the calibration
curve was made with rutin, which is a similar di-
glycoside of quercetin as the one used as marker
for this taxon.
Note: The majority of the flavonoids in pollen are
kaempferol, quercetin, isorhamnetin, apigenin or luteolin
Table 29. Amplitude of the Wavelengths in Band I of flavones
and flavonols UV spectra.
Flavonoid type k (nm)
Flavones 304–350
Flavonols (OH-3 derivative) 328–357
Flavonol (OH-3 free) 352–385
Table 31. Wavelengths of Band II in UV spectra of flavones
with oxygenation only in A-ring.
Flavone Oxygenation pattern of A-ring kmax. (nm)
Flavone 250
5-hydroxiflavone 5 268
7-hydroxiflavone 7 252
5,7-di-hydroxiflavone 5.7 268
Baicalein 5,6,7 274
Norwogonin 5,7,8 281
Table 30. Wavelengths of Band I in UV spectra of flavonols
with a different pattern of oxygenation in B-ring.
Oxigenation pattern
Flavonols Ring-A and C Ring-B kmax. (nm)
Galangin 3,5,7 – 359
kampferol 3,5,7 40 367
Quercetin 3,5,7 30, 40 370
Myricetin 3,5,7 30, 40, 50 374
Standard methods for pollen research 43
derivatives, mainly glycosides. However, many different
structures can be found and is not difficult found a
good marker for each taxon. A combination of two or
more markers can also be used if necessary.
7. Determination of phenolic profile of bee
pollen extracts by LC and LC/MS
Polyphenols are known to interact with compounds
present in foods, like carbohydrates, proteins or lipids
(Jakobek, 2015; Ozdal et al. 2013). Comprehensive
analytical methods such as LC and LC/MS can be used
to obtain more detailed information on phenolic
profile of bee pollen as well as on interactions in food
systems. Standard procedures for the extraction of
phenolic compounds from bee pollen, pre-concentra-
tion of phenolic extracts prior to analysis, LC
and LC/MS techniques for their determination will
be described.
7.1. Polyphenolics: importance, structure and
classification
The phenolic compounds in pollen grains are important
as a part of reproductive process in plants. They pro-
tect pollen load from destructive exterior factors such
as UV-radiation (LeBlanc et al., 2009). During the ger-
mination of pollen load, the formation of two protective
membranes – exine and intine is occurred. Both mem-
branes are rich in content of phenolics. Campos et al.
(1997) have shown that it is possible to establish the
correlation between the polyphenols and palynological
origin of bee-collected pollen.
There are thousands phenolic structures of substances
present in plants (Boudet, 2007). It can be the simple
molecular structure with one substituted benzenoic ring or
complex polyphenolic structure with two or more inter-
connected aromatic rings. Natural phenolic compounds
can be classified according to their structure in two
major classes:
1. non-flavonoids polyphenols (Lignans and lignins, phen-
olic acids, coumarins, stilbene oligomers, gallotannins,
ellagitannins and procyanidins (Soto-Vaca et al., 2012))
(Table 32);
2. flavonoids (large group of molecules with complex
structure derived from p-coummaric acid. Basically, it
consists of two aromatic rings with different alkyl and
(or) O-substituents and with multiple chiral centers
(Y
a~nez et al., 2013). Benzene rings A and B are con-
nected with C3 chain which are often closed and form
heterocyclic ring C (Figure 36).
There are a significant number of subclasses of fla-
vonoids based on different reactions that occur on the
rings over the lateral groups – hydroxylation, prenyla-
tion, alkalinisation and glycosylation (Stalikas, 2007):
chalcones (without formed C ring), aurones, flavanols
(includes catechins, epicatechins and procyanidins), fla-
vonols, flavones flavanones, isoflavones and antocya-
nins. The general structures of these subclasses of
flavonoids are presented in Table 33. Presence of sugar
unit (in form of glycoside) and OH-groups increase
water solubility of these molecules, meanwhile the
presence of alkyl groups make them lipophilic. The
most abundant form of flavonoids is the glycosidic
form with different sugars linked (glucose, ramnose,
etc.). If there is no bonded sugar unit, flavonoids are in
aglycone form.
7.2. Preparation of pollen extracts for determination of
phenolic compounds
Bee pollen can be considered as a complex mixture of
different compounds which differ, both, in chemical
composition as well as in their solubility. The complex-
ity of the structure of the pollen grains are also
reflected in the fact that it is protected by two mem-
branes with the main outer, exine, that consists of poly-
mer sporopollenin. For this reason, isolation and
extraction of material from pollen requires prior soften-
ing or maceration of pollen grain so that membranes
Figure 35. Fingerprint of Eucalyptus globulus obtained with a
pollen pellet (100% pure) and a mixture of bee pollen with
about 60% of Eucalyptus globulus pollen pellets.
Figure 36. General formula of flavonoids.
44 M. G. Campos et al.
Table 32. Model of chemical structures of subclasses or representative molecule of subclasses of non-flavonoids phen-
olic compounds.
Subclass Model of chemical structure or model of representative molecule Name
Lignans virolin
Lignins pinoresinol
Phenolic acids hydroxycinnamic acids
hydroxybenzoic acids
Coumarins coumarin
Stilbenes resveratrol
Tannins gallotannin
Standard methods for pollen research 45
slightly relaxed or their physical disruption (by using
ultrasound or sonification) in order to increase the
extraction efficiency. Application of ultrasound (which
increases penetration of solvent’s molecules into the
pollen matrix thereby enhance contact of solvent and
the respective soluble components therein) is especially
recommended due to the presence of thermolabile
components in plant materials (Wang & Weller, 2006),
including pollen. Generally, during the entire extraction
process application of higher temperature should be
avoided in order to preserve the original compos-
ition of the compounds present in the pollen
Table 33. Model of chemical structures of flavonoids.
Subclass Model of chemical structure
Chalcones
Aurones
Flavanols
Flavonols
Flavones
Flavanones
(Continued)
46 M. G. Campos et al.
because some of them are volatile at elevated tem-
peratures (lipids, phenolic compounds, etc.) or they
can lose their native structure (free amino acids,
proteins, polysaccharides). All phenolic compounds
present in plants can be regarded as more or less
water soluble (flavonoids, most of the phenolic acids,
gallotannins and ellagitannins) and water insoluble
(coumarine, stilbenes, proantocyanidins, some phen-
olic acids bonding in the cell wall, lignins). Solubility/
Insolubility in water and similar solvents depends pri-
mary on the presence of side groups attached to
the aromatic rings.
7.2.1. Extraction of phenolic compounds from pollen – gen-
eral procedure
Based on the aforementioned the following procedure
for the extraction of phenolic compounds from pollen
can be proposed:
1. Measure the appropriate mass of pollen sample
(e.g., 0.5 g);
2. Add 20 cm3 (calculated for 0.5 g of pollen) of heptane
in order to remove lipid components;
3. Shake thoroughly sample on Vortex and place it in an
ultrasound bath for 15min at 30 C;
4. Centrifuge sample (5000 rpm) to separate non-polar,
lipid fraction;
5. Add 20 cm3 (calculated for 0.5 g of pollen) of 50%
ethanol on the residual undissolved portion of pollen
and place it in an ultrasound bath for 15min at 30 C;
6. Centrifuge sample (5000 rpm) in order to separate
liquid and solid phases;
7. Evaporate ethanol from supernatant by vacuum-rotary
evaporator at temperature of 35 C;
8. Add ethyl acetate to the remaining aqueous solu-
tion (volumes of solution and extraction solvent
should be equal) in order to perform the extrac-
tion of the phenolic compounds and their separ-
ation from the remaining organic substances
(proteins and sugars);
A phenolic fraction, obtained in this way, is prepared
for further concentration using the solid-phase extrac-
tion (SPE) method.
7.2.2. Preparation of enzymatic hydrolysates from pollen
Besides the common extracts for determination of
phenolic profile of pollen, some authors (Rzepecka-
Stojko, Pilawa et al., 2012; Rzepecka-Stojko, Stec et al.,
2012) have prepared two types of enzymatic hydroly-
sates from pollen: pepsin hydrolysates from pollen (PEP)
and ethanol extracts of pepsin-digested bee pollen
(EEPP). On the other hand, Smirnova et al. (2012) pre-
pared enzymatic hydrolysate of pollen’s exine by using
feruloyl esterase.
7.2.2.1. Preparation of PEP.
1. Measure 20 g of pollen samples.
2. Add distilled water (volume of distilled water should
be five times the volume of sample) and acidify with
conc. HCl until reach pH 2.
3. Add pepsin to obtain its concentration of 1%.
4. Incubate sample at 37 C for 48 h to per-
form hydrolysis.
5. Boil sample for 10min to stop enzymatic activity.
6. Filter the obtained enzymatic hydrolysate under
reduced pressure.
7. Centrifuge the filtrate for 10min at 10,000 rpm.
8. Evaporate the supernatant at rotary vacuum-evapor-
ator to yield dry extract.
Table 33. (Continued).
Subclass Model of chemical structure
Isoflavones
Antocyanins
Standard methods for pollen research 47
T
ab
le
34
.
So
m
e
pu
bl
ic
at
io
ns
on
LC
an
d
LC
/M
S
an
al
ys
is
of
po
ly
ph
en
ol
ic
co
m
po
un
ds
fr
om
be
e
po
lle
n
sa
m
pl
es
.
C
om
po
un
ds
M
et
ho
d
St
at
io
na
ry
ph
as
e
M
ob
ile
ph
as
e
t
(
C
)
Fl
ow
ra
te
(m
L/
m
in
)
k
(n
m
)
R
ef
er
en
ce
s
A
B
F
la
vo
no
id
s
H
PL
C
-P
A
D
C
18 1
1.
9
cm
	
0.
4
cm
,
5
mm
O
rt
ho
ph
os
ph
or
ic
ac
id
pH
2.
5
A
ce
to
ni
tr
ile
24
0.
8
22
0-
40
0
C
am
po
s
et
al
.(
19
97
)
F
la
vo
no
id
s
(a
gl
yc
o
ns
)
H
PL
C
-D
A
D
C
18 4
.6
m
m
	
25
0
m
m
,
10
mm
O
rt
ho
ph
os
ph
or
ic
ac
id
pH
2.
6
M
et
ha
no
l
2
27
8-
28
2
27
8-
35
0
Se
rr
a-
Bo
nv
eh

 ı
et
al
.(
20
01
)
P
he
no
lic
ac
id
s,
fla
vo
no
id
s
N
an
o-
LC
-U
V
C
18 1
00
mm
I.D
.	
10
cm
0.
5%
fo
rm
ic
ac
id
M
et
ha
no
l
50
0
nl
/m
in
20
0,
28
0
Fa
na
li
et
al
.(
20
13
)
A
nt
ho
cy
an
in
s
H
PL
C
-D
A
D
-E
SI
-M
S/
M
S
C
18 1
50
m
m
	
4.
6
m
m
,
5
mm
0.
1%
tr
ifl
uo
ra
ce
tic
ac
id
A
ce
to
ni
tr
ile
35
0.
5
52
0
D
iP
ao
la
-N
ar
an
jo
et
al
.(
20
04
)
F
la
vo
no
id
s
H
PL
C
-P
A
D
-E
SI
-M
S/
M
S
C
18 2
50
m
m
	
4
m
m
,
5
mm
1%
fo
rm
ic
ac
id
M
et
ha
no
l
1
35
0,
52
0
Fe
rr
er
es
et
al
.(
20
10
)
F
la
vo
no
id
s,
hy
dr
o
xy
ci
nn
am
ic
ac
id
am
id
e
de
ri
va
ti
ve
s
H
PL
C
-P
A
D
-E
SI
-M
S/
M
S
C
18 2
50
m
m
	
4.
6
m
m
,
5
mm
0.
1%
ac
et
ic
ac
id
M
et
ha
no
l
28
0.
5
27
0,
35
0
N
eg
ri
et
al
.(
20
11
)
F
la
vo
no
id
ag
ly
co
ne
s
H
PL
C
-D
A
D
-A
PC
I/M
S
C
8 1
50
m
m
	
4.
6
m
m
,
5
mm
0.
1%
fo
rm
ic
ac
id
in
5%
ac
et
on
itr
ile
A
ce
to
ni
tr
ile
30
1
25
8
Lv
et
al
.(
20
15
)
48 M. G. Campos et al.
9. Complete drying in a laboratory incubator (desicca-
tor) at 38 C.
10. Measure the dry extract and dissolve it in distilled
water to prepare PEP-solution with concentration of
2mg/cm3.
7.2.2.2. Preparation of EEPP.
1. Extract the supernatant obtained after pepsin-treat-
ment of pollen with 200 cm3 of 50% ethanol for
60min at room temperature with intensive shaking.
2. Filter the extract under reduced pressure.
3. Centrifuge the filtrate for 10min at 10,000 rpm.
4. Evaporate the supernatant at rotary vacuum-evapor-
ator to yield dry extract.
5. Complete drying in a laboratory incubator (desiccator)
at 38 C.
6. Measure the dry extract and dissolve in 50% ethanol
solution to prepare EEPP-solution with concentration
of 2mg/cm3.
7.2.2.3. Preparation of feruloyl esterase hydrolysate
from pollen’s exine. For the preparation of this type of
hydrolysate, Smirnova et al. (2012) used enzyme feruloyl
esterase derived from the mold of genus Aspergillus sp.
according to the following procedure:
1. Incubate the exine of pollen in acetate buffer solution
pH 4.5 containing enzyme feruloyl esterase in quantity
of 5 g/dm3 for 24 h at 50 C;
2. Thoroughly wash the digested material with dis-
tilled water;
3. Incubate in 1% sodium-dodecylsulfate (SDS) solution;
Table 35. Mass spectral data of compounds detected in extracts of bee pollen.
No
[MþH]þ,
m/z
[M-H]-,
m/z
MS/MS positive
ion mode
MS/MS
negative ion
mode Proposed identification References
1 685.0 Rosmarinic acid dihexoside derivative Negri et al. (2011)
2 627.2 481.0, 319.0 Myricetin-3-O-rhamnosyl-glucoside Negri et al. (2011)
3 435.0 303.0 Quercetin-3-O-arabinoside Negri et al. (2011)
4 641.2 479.0, 317.0 Isorhamnetin-3-O-diglucoside Negri et al. (2011)
5 595.0 449.0, 287.0 Kaempferol-3-O- rhamnosyl-glucoside Negri et al. (2011)
6 771.2 625.1, 479.1, 317.1 Isorhamnetin-3-O-
(20 ',30’-dirhamnosyl)glucoside
Negri et al. (2011)
7 611.1 303.1 Rutin Negri et al. (2011)
8 625.2 479.0, 317.0 Isorhamnetin-3-O- rhamnosyl-glucoside Negri et al. (2011)
9 449.0 287.0 Kaempferol-3-O-glucoside Negri et al. (2011)
10 641.2 495.0, 333.0 Patuletin-3-O- rhamnosylglucoside Negri et al. (2011)
11 449.2 303.1 (100%) Quercitrin Negri et al. (2011)
12 381.4 261.0, 235.2, 147.2 Chalcone Negri et al. (2011)
13 761.3 615.3, 381.1, 287.0 Kaempferol-7-O-rhamnosyl-3-O-galloyl
glucuronide
Negri et al. (2011)
14 584.3 438.2, 420.2 N’ ,N’’ ,N’’’-tris-p-coumaroylspermidine Negri et al. (2011)
15 674.3 498.2 , 480.3 N’ ,N’’ ,N’’’-tris-p-feruloylspermidine Negri et al. (2011)
16 465 303 Delphinidin-3-O-glucoside Di Paola-Naranjo et al. (2004)
17 611 303, 465 Delphinidin-3-O-rutinoside Di Paola-Naranjo et al. (2004)
18 449 287 Cyanidin-3-O-glucoside Di Paola-Naranjo et al. (2004)
19 595 287, 449 Cyanidin-3-O-rutinoside Di Paola-Naranjo et al. (2004)
20 479 317 Petunidin-3-O-glucoside Di Paola-Naranjo et al. (2004)
21 625 317, 479 Petunidin-3-O-rutinoside Di Paola-Naranjo et al. (2004)
22 609 301, 463 Peonidin-3-O-rutinoside Di Paola-Naranjo et al. (2004)
23 639 331, 493 Malvidin-3-O-rutinoside Di Paola-Naranjo et al. (2004)
24 535 287, 449 Cyanidin-3-(60’-malonylglucoside) Di Paola-Naranjo et al. (2004)
25 625 463, 445, 300 Quercetin-3- sophoroside Ferreres et al. (2010)
26 623 315 Isorhamnerin-3- rutinoside Ferreres et al. (2010)
27 609 463, 445, 300 Quercetin-3- neohesperidoside Ferreres et al. (2010)
28 609 447, 429, 285 Kaempferol-3- sophoroside Ferreres et al. (2010)
29 739 593, 575, 284 Kaempferol-3-(4- rhamnosyl)-
neohesperidoside
Ferreres et al. (2010)
30 593 447, 429, 284 Kaempferol- 3- neohesperidoside Ferreres et al. (2010)
31 635 593, 489,
471, 284
Kaempferol-3-(acetyl)-
neohesperidoside
Ferreres et al. (2010)
32 739 593, 285 Kaempferol-3- neohesperidosyl-
7- rhamnoside
Ferreres et al. (2010)
33 447 285 Kaempferol-3- glucoside Ferreres et al. (2010)
34 593 285 Kaempferol-3- rutinoside Ferreres et al. (2010)
35 635 593, 489,
471, 284
Kaempferol- 3- (acetyl)-
neohesperidoside isomer
Ferreres et al. (2010)
Standard methods for pollen research 49
4. Wash again with distilled water;
5. Dry at 55 C until constant mass has been reached.
7.2.3. Application of SPE for pre-concentration of phenolic
compounds for further LC and LC/MS determination
Solid-phase extraction (Figure 37) is one of the methods
of sample preparation for the further analysis. It was
developed during the 1980s. Currently, SPE is applied
for pre-concentration of samples or for their isolation
and purification (Bertoncelj et al., 2011; Gasi
c et al.,
2014a; 2014b; Keckes et al., 2013; Santana et al., 2009).
It is based on interaction of some components from
liquid phase with solid (sorbent) phase. The desired
compound retains on column, meanwhile, all the
unwanted components have been washed and removed
with rinsing of column in convenient medium. The ana-
lyte is than recovered with small quantity of appropriate
solvent. In this way, apart from the isolation of the sub-
stance, its pre-concentration is achieved which facilitate
Table 36. Primer Sequences with indexes SA501 – SB712 (adapted from Kozich et al., 2013).
Forward
Name Sequence
SA501 AATGATACGGCGACCACCGAGATCTACAC ATCGTACG CCTGGTGCTG GT ATGCGATACTTGGTGTGAAT
SA502 AATGATACGGCGACCACCGAGATCTACAC ACTATCTG CCTGGTGCTG GT ATGCGATACTTGGTGTGAAT
SA503 AATGATACGGCGACCACCGAGATCTACAC TAGCGAGT CCTGGTGCTG GT ATGCGATACTTGGTGTGAAT
SA504 AATGATACGGCGACCACCGAGATCTACAC CTGCGTGT CCTGGTGCTG GT ATGCGATACTTGGTGTGAAT
SA505 AATGATACGGCGACCACCGAGATCTACAC TCATCGAG CCTGGTGCTG GT ATGCGATACTTGGTGTGAAT
SA506 AATGATACGGCGACCACCGAGATCTACAC CGTGAGTG CCTGGTGCTG GT ATGCGATACTTGGTGTGAAT
SA507 AATGATACGGCGACCACCGAGATCTACAC GGATATCT CCTGGTGCTG GT ATGCGATACTTGGTGTGAAT
SA508 AATGATACGGCGACCACCGAGATCTACAC GACACCGT CCTGGTGCTG GT ATGCGATACTTGGTGTGAAT
SB501 AATGATACGGCGACCACCGAGATCTACAC CTACTATA CCTGGTGCTG GT ATGCGATACTTGGTGTGAAT
SB502 AATGATACGGCGACCACCGAGATCTACAC CGTTACTA CCTGGTGCTG GT ATGCGATACTTGGTGTGAAT
SB503 AATGATACGGCGACCACCGAGATCTACAC AGAGTCAC CCTGGTGCTG GT ATGCGATACTTGGTGTGAAT
SB504 AATGATACGGCGACCACCGAGATCTACAC TACGAGAC CCTGGTGCTG GT ATGCGATACTTGGTGTGAAT
SB505 AATGATACGGCGACCACCGAGATCTACAC ACGTCTCG CCTGGTGCTG GT ATGCGATACTTGGTGTGAAT
SB506 AATGATACGGCGACCACCGAGATCTACAC TCGACGAG CCTGGTGCTG GT ATGCGATACTTGGTGTGAAT
SB507 AATGATACGGCGACCACCGAGATCTACAC GATCGTGT CCTGGTGCTG GT ATGCGATACTTGGTGTGAAT
SB508 AATGATACGGCGACCACCGAGATCTACAC GTCAGATA CCTGGTGCTG GT ATGCGATACTTGGTGTGAAT
Reverse
Name Sequence
SA701 CAAGCAGAAGACGGCATACGAGAT AACTCTCG AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SA702 CAAGCAGAAGACGGCATACGAGAT ACTATGTC AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SA703 CAAGCAGAAGACGGCATACGAGAT AGTAGCGT AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SA704 CAAGCAGAAGACGGCATACGAGAT CAGTGAGT AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SA705 CAAGCAGAAGACGGCATACGAGAT CGTACTCA AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SA706 CAAGCAGAAGACGGCATACGAGAT CTACGCAG AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SA707 CAAGCAGAAGACGGCATACGAGAT GGAGACTA AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SA708 CAAGCAGAAGACGGCATACGAGAT GTCGCTCG AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SA709 CAAGCAGAAGACGGCATACGAGAT GTCGTAGT AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SA710 CAAGCAGAAGACGGCATACGAGAT TAGCAGAC AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SA711 CAAGCAGAAGACGGCATACGAGAT TCATAGAC AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SA712 CAAGCAGAAGACGGCATACGAGAT TCGCTATA AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SB701 CAAGCAGAAGACGGCATACGAGAT AAGTCGAG AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SB702 CAAGCAGAAGACGGCATACGAGAT ATACTTCG AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SB703 CAAGCAGAAGACGGCATACGAGAT AGCTGCTA AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SB704 CAAGCAGAAGACGGCATACGAGAT CATAGAGA AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SB705 CAAGCAGAAGACGGCATACGAGAT CGTAGATC AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SB706 CAAGCAGAAGACGGCATACGAGAT CTCGTTAC AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SB707 CAAGCAGAAGACGGCATACGAGAT GCGCACGT AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SB708 CAAGCAGAAGACGGCATACGAGAT GGTACTAT AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SB709 CAAGCAGAAGACGGCATACGAGAT GTATACGC AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SB710 CAAGCAGAAGACGGCATACGAGAT TACGAGCA AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SB711 CAAGCAGAAGACGGCATACGAGAT TCAGCGTT AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
SB712 CAAGCAGAAGACGGCATACGAGAT TCGCTACG AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
Index and Read
Name Sequence
Read1 CCTGGTGCTG GT ATGCGATACTTGGTGTGAAT
Read2 AGTCAGTCAG CC TCCTCCGCTTATTGATATGC
Index GCATATCAATAAGCGGAGGA GG CTGACTGACT
Index sequences indicated in bold.
50 M. G. Campos et al.
further analysis. Also, the use of large amounts of
organic solvents for extraction can be avoided
(Rouessac & Rouessac, 2007). This is important due to
three reasons – medical, environmental and economic.
Typical solid phase extraction involves several steps
(Rouessac & Rouessac, 2007) (Figure 37):
1. activation and rinsing of the sorbent before using it –
performed by passing a certain volume of solvent simi-
lar polarity as well as samples that will be loaded to
the cartridge;
2. adsorption of the analyte by loading of known volume
of sample;
3. elimination of all interfering compounds by rinsing
cartridge with adequate solvent;
4. desorption of an analyte by percolation of cartridge.
The solid phase in cartridge is made of similar mate-
rials like solid phase in LC-columns depending of type
of analyte. A number of sorbents may be used such as
silica gel, cellulose, aluminum oxide, polyamide and
bonded silica gels (Milojkovi
c-Opsenica et al., 2015).
Polar compounds are isolated from cartridge with non-
polar matrix. If the analyzed sample contains non-polar
components, they can be isolated from cartridge with
polar matrix made of C-18 derivatized silica gel. This
gel has attached a hydrocarbonic C18-chain which
increases its hydrophobicity. They were the first type of
sorbents applied for SPE pre-concentration of phenolic
compounds from water samples (Mußmann et al., 1994;
Rodr
ıguez et al., 2000; Tian et al., 2005). Interactions
between phenolic compounds and this type of sorbent,
based on Van der Waal’s forces, can be weak if phenols
are in ionized states, so it is necessary to adjust the pH
at the proper value. Also, since silica materials are sen-
sitive on pH value, it is possible to substitute this type
of sorbent with some other, more polar matrices such
as modified polystyrene-divinylbenzene polymeric
(XAD) resins (Fontanals et al., 2007; Li et al., 2002;
Wissiak & Rosenberg, 2002). The procedure is similar
to those applied for C-18 derivatized silica gel but
adsorption area is larger and presence of some aro-
matic parts in polymeric structure made it more suit-
able for phenols extraction (Santana et al., 2009).
A typical cartridge SPE manifold can accommodate
up to 24 cartridges. Most SPE manifolds are equipped
with exit for connection to vacuum or water pump.
Application of reduced pressure speeds up the extrac-
tion process by pulling the liquid sample through the
stationary phase. The analytes are collected in sample
tubes inside or below the manifold after they pass
through the stationary phase. Before the loading on the
cartridge samples are usually passed through a disk for
extraction 0.5mm thickness and 25 to 90mm in diam-
eter (Rouessac & Rouessac, 2007). It serves to remove
the low quantity of organic substances which are pos-
sibly present in the tested aqueous solution.
Table 37. International data on the protein content (% D.M) of bee pollen.
Crude protein content (% D.M)
Type of samples Number of samples Values References
Multifloral 21 21.0–29.3 Herbert & Shimanuki, 1978
Multifloral 20 12.6–18.2 Serra-Bonveh
ı & Escol
a Jord
a, 1997
Monofloral 194 9.2–37.4 Somerville, 2001
Multifloral 15 9.8–16.5 Villanueva et al., 2002
Monofloral 29 13.6–31.9 Andrada & Telleria, 2005
Multifloral 10 21 ± 4
 Almeida-Muradian et al., 2005
Multifloral 6 batches 23.59

 Melo et al., 2005
Monofloral 123 13.0–24.5 Szcze^sna, 2006a
Multifloral (Poland) 13 15.8–24.1 Szcze^sna, 2006b
Multifloral (Korea) 9 17.6–24.5 Szcze^sna, 2006b
Multifloral (China) 5 17.8–26.1 Szcze^sna, 2006b
Monofloral 42 6.3–26.3 Forcone et al., 2011
Multifloral 154 12.2–27.0 Martins et al., 2011
Multifloral 14 18.4–22.4 Balkanska & Ignatova, 2012
Multifloral 7 15.2–28.5 Stanciu et al., 2012
Monofloral 12 14.2–28.9 Yang et al., 2014
Multifloral 7 batches 23.38 ± 1.24
 (Arruda et al., 2013)
Multifloral 106 13.65–26.5 (Liolios et al., 2014)
Monofloral 58 9.7–30.11 (Liolios et al., 2014)

Average and standard deviation.

Average.
Figure 37. Solid phase extraction.
Standard methods for pollen research 51
7.2.3.1. Application of SPE for pre-concentration of
phenolic compounds from pollen. The literature for
pre-concentration of phenolic compounds from pollen
by SPE is very limited (Di Paola-Naranjo et al., 2004).
According to our laboratory practice and experience
the following procedure for pre-concentration of phen-
olic compounds from pollen phenolic fraction prepared
by general procedure can be proposed:
1. Condition and activate the SPE cartridge by passing 10
cm3 of methanol and 10 cm3 of 0.5% hydrochloric
acid, successively. Do not allow air to enter the cart-
ridge packing;
2. Load 2 cm3 of pollen phenolic fraction by passing
slowly (no faster than one drop per second) into the
cartridge. Do not allow air to enter in to the resin
bed. Collect all cartridge effluent in clean tubes;
3. Wash the cartridge by passing 10 cm3 of 0.5% HCl
(for purpose of elimination of potentially present sug-
ars or some other polar substances). Collect all cart-
ridge effluent in clean tubes;
4. Eluate the phenolic compounds by slowly passing (less
than one drop per sec) of 5 cm3 of acidic solution of
methanol (add 0.5% HCl). Do not allow air to enter
into the resin bed. Collect phenolic solution in
clean tubes;
5. Tap tubes and pack it into the refrigerator or freezer
until further analysis.
7.3. LC and LC/MS techniques
7.3.1. LC techniques
High-performance liquid chromatography or high-pres-
sure liquid chromatography (HPLC) is highly improved
form of column chromatography. It is based on the
principle of forcing solvent through column, packed
with stationary phase consisting of small particles, under
high pressures (up to 400 atmospheres). Nano-liquid
chromatography method with UV-Vis detection is also
developed for analysis of phenolic compounds in bee
pollen (Fanali et al., 2013). The main advantages of this
method are short analysis time and reduced costs due
to the small dimensions of the column.
7.3.1.1. LC-DAD technique. The photodiode array
detector has been extensively used for the detection of
polyphenolics due to their absorption in the UV region
(Robards & Antolovich, 1997). Detection is based on
the absorptive measurement at characteristic wave-
lengths. Phenolic acids are usually detected at wave-
lengths between 210 and 320 nm, anthocyanidins and
anthocyanins (glycosides or acylglycosides of anthocyani-
dins) between 500 and 530 nm, flavanols between 210
or 280 nm, flavanones at 280 nm, and possibly at 252,
285, 290 and 365 nm, flavones at 360 or 370 nm, flavo-
nols and their glycosides at 270, 365, and 370 nm and
isoflavones at 260 nm, and possibly between 230 and
280 nm. UV spectra of flavonoids, for example, have
two absorption bands. Band I, with maximum absorp-
tion in the range of 300 to 550 nm, arises from the A
aromatic ring and band II, with maximum absorption in
the range of 240 to 285 nm, comes from the B ring. In
LC-DAD system, polyphenolic standards are used to
generate characteristic UV-Vis spectra and retention
times. Identification of individual polyphenolic com-
pounds is achieved by comparison with retention times
and UV-Vis spectra of standards. Quantification of these
compounds is performed from the peak areas by com-
parison to calibration curves obtained with standard sol-
utions. As general procedure for determination of
polyphenolic compounds from bee pollen extracts can
be HPLC-DAD method developed by Campos
et al. (1997).
7.3.1.2. LC-MS, MS/MS techniques. Although HPLC-
DAD method is the reference method for the analysis
of phenolic compounds, certain problems, like co-elu-
tion and similar UV-Vis spectra of many compounds,
can occur. As well, most glycosides and acyl residues
are poor chromophores, so no further distinguishing
can be achieved by means of DAD UV detection. In
that case mass spectrometry (MS), more precisely, tan-
dem mass spectrometry (MS/MS) detection is suggested
for further characterization and identification of
eluted compounds.
The MS principle consists in ionizing chemical com-
pounds to generate charged molecules or molecule
fragments and measuring their mass-to-charge ratios
(Sparkman, 2000). LC-MS provides structural informa-
tion about eluted compounds but is rarely used for full
structure characterization. Using LC-MS/MS more frag-
mentation of the precursor and product ions is formed,
providing additional structural information for the iden-
tification of phenolic compounds like details about the
aglycone moiety, the types of carbohydrates (mono-, di-
, tri- or tetrasaccharides and hexoses, deoxyhexoses or
pentoses) or other substituents present, interglycosidic
linkages, attachment points of the substituents to the
aglycone, etc. Aforementioned useful structural informa-
tion could be obtained directly and easily applying elec-
trospray ionization (ESI) and atmospheric pressure
chemical ionization (APCI) techniques, and so, many
purification procedures can be circumvented. Both are
the soft ionization techniques and the most widely used
in analysis of polyphenolic compounds in bee pollen and
other food. ESI is gentler method and is generally used
for polar compounds that can be ionized in solution,
while APCI is used for less polar molecules that can
undergo acid-based reactions in the gas phase (Biesaga
& Pyrzynska, 2009). Both positive and negative ioniza-
tion are applied. Negative ion mode provides the high-
est sensitivity in flavonoid analysis and fragmentation is
limited (Cuyckens & Claeys, 2002; de Rijke et al., 2003;
Fabre et al., 2001; Wang et al., 2007), while mass
52 M. G. Campos et al.
spectrum in the positive ion mode shows different frag-
mentation pattern. Analyte responses can vary consider-
ably so combined use of both EI and PI modes can
prove complementary information’s which are helpful in
the structure determination, especially for the minor
compounds (Cuyckens & Claeys, 2004; de Rijke et al.,
2006). The eluent system also has a distinct influence
on the ionization efficiency and therefore on structural
information. The most common additives are formic
acid, trifluoroacetic acid, ammonium acetate and ammo-
nium formate. They are volatile and thus compatible
with LC-MS. Combination of methanol-ammonium for-
mate at pH 4 gives the highest response in negative
mode APCI (de Rijke et al., 2003), while the best sensi-
tivity in ESI is achieved with acidic ammonium acetate
buffer and organic component consisted of methanol
and acetonitrile as mobile phase (Rauha et al., 2001).
Formic acid has slight negative effects on the ionization
efficiency but improves chromatographic separation and
peak shapes (Prokudina et al., 2012). For trifluoroacetic
acid is reported that suppresses the ionization due to
ion-pairing and surface-tension effects (Costa
et al., 2000).
Chromatographic retention times provide useful
structural information. In reversed-phase HPLC separ-
ation of polyphenolics on the basis of their polarity, the
elution order of polyphenolics may be predicted
Glycosylation increases polarity of flavonoids and there-
fore their mobility in the reversed-phase system.
Acylation and methylation have the opposite effect but
their position have significant influence on the retention
time (Greenham et al., 1995; Harborne & Boardley,
1984). In case of flavonoid glycosides rutinosides pre-
cede neohesperidosides, galactosides precede glucosides
(Robards & Antolovich, 1997), glucosides precede arabi-
nosides and arabinosides precede rhamnosides
(Harborne & Boardley, 1984; Schieber et al., 2002).
Linkage position can have an influence on the retention,
too. The elution order for different classes of flavonoids
is flavanones, flavonols and flavones, for compounds
with equivalent substitution pattern. The elution order
for hydroxycinnamic acids is caffeic acid, p-coumaric
acid, sinapic acid, ferulic acid and cinnamic acid while
for benzoic acids order is gallic acid, protocatechuic
acid, vanillic acid and syringic acid (Rodr
ıguez-Delgado
et al., 2001).
Quality and reliability of analysis depend on the com-
bination of the mass analyzer with the detector.
Analyzers can belong to magnetic sector, time-of-flight
(TOF) or Fourier transform (FT) generic type, depend-
ing on the physics of mass analysis. There are analyzers
that can perform MS and MS/MS (MS2) analysis, some-
times to a high MSn stage or different analyzers could
be combined together into one hybrid instrument. Mass
analyzers used for analyzing polyphenolic compounds
are usually quadrupole and ion-trap instruments. Ion-
trap instruments can perform MSn experiments and that
is its main advantage (de Rijke et al., 2003). Using ion-
trap mass spectrometry Ferreres et al. (2010) reported
the presence of non-colored flavonoids in bee pollen
from Echium plantagineum and Lv et al. (2015) identified
and quantified flavonoid aglycones in rape bee pollen
collected from Qinghai-Tibetan Plateau. In order to
optimize signals and to obtain maximal structural infor-
mation from ion of interest, the different collision
energy values should be applied. Methods based on
low-energy collision induced dissociation permit the
characterization of most structural features of flavonoid
glycosides. High-energy CID methods provide additional
structural information through more fragmentation
processes. Collision energy directly affects abundance
ratio of the various fragment ions. Optimum CE varies
for different compounds. Gu et al. (2012) showed that
a CE of 70 eV is sufficient to cause abundant product
ions for structural elucidation of flavonoids and 30 eV
for generation characteristic fragment ions for chloro-
genic acid (Gu et al., 2012). Recent publications on LC
and LC/MS analysis of polyphenolic compounds from
bee pollen samples are listed in Table 34.
In the past, fragmentation behavior of phenolic com-
pounds, especially flavonoids, have been extensively
investigated. A great number of rules for structural
characterization of unknown compounds, even without
the reference standards, were summarized (de Rijke
et al., 2006). As experience shows, fragmentation path-
ways are largely independent of the ionization mode
(ESI or APCI) and the type of instrument (triple quadru-
pole or ion trap) used (de Rijke et al., 2003; Rauha
et al., 2001) so that rules can be applied for resolving
spectral data obtained with modern instruments.
Characteristic m/z values of fragment ions from mass
spectral data used for identification of certain poly-
phenolic compounds from bee pollen are listed in
Table 35.
7.3.2. UHPLC-HESI-MS/MS technique for determination of
phenolic compounds
The LTQ-Orbitrap spectrometer supports a wide range
of applications. Its MS/MS capabilities make this mass
spectrometer extremely powerful, among other things,
for unambiguous detection of polyphenolic compounds
(Keckes et al., 2013; Nati
c et al., 2015; Panteli
c et al.,
2014; Ristivojevi
c et al., 2014). Some chromatograms
and mass spectra from our previous research are
shown on Figures 38–40. Instrumentation and
Table 38. Determination of the protein content (%) using dif-
ferent volumes of the H3BO3 solution.
Volume of H3BO3 (ml)
Protein content (%)
n¼ 5
30 20.49 (± 0.07)a
40 20.44 (± 0.06)a
Analysis performed under the same conditions except the volume of
the H3BO3 solution used.
Standard methods for pollen research 53
conditions for determination of phenolic compounds
using this method are as follows:
1. Pollen extracts (Section 7.2);
2. Thermo Scientific liquid chromatography system
equipped with:
a. Quaternary Accela 600 pump;
b. Accela Autosampler;
c. Linear ion trap-Orbitrap hybrid mass spectrom-
eter (LTQ OrbiTrap XL, Thermo Fisher
Scientific, Bremen, Germany) with heated electro-
spray ionization (HESI);
d. Hypersil gold C18 column (50	 2.1mm, 1.9 mm)
from Fisher Scientific.
3. Mobile phases:
a. water þ 0.1% formic acid;
b. acetonitrile þ 0.1% formic acid.
4. Linear gradient: 0–5min from 5% to 95% (B), 5–6min
95% (B) then 5% (B) for 3min;
a. HESI-source parameters:
b. Source voltage 4 kV;
c. Capillary voltage 47 V;
d. Tube lens voltage 159.11 V;
e. Capillary temperature 275 C;
f. Sheath and auxiliary gas flow (N2) 25 and 8 (arbi-
trary units);
g. 100–1500 m/z;
h. Collision energy 35 eV.
5. Injection volume 5 ml;
6. Flow rate of 0.400ml/min.
8. Standard method for identification of bee
pollen mixtures through meta-barcoding
Pollen analysis is a central part of bee ecology research
(Carvell et al., 2006; Danner et al., 2014; K€oppler et al.,
2007). Identification of plant species origin of bee col-
lected pollen traditionally relies on light microscopy and
discrimination based on morphological differences of
pollen grains (Mullins & Emberlin, 1997). This is labor-
and time-intensive (Galimberti et al., 2014), requires
expert knowledge (Keller et al., 2015) and lacks dis-
criminative power at lower taxonomic levels
(Galimberti et al., 2014; Williams & Kremen, 2007),
which means that pollen from closely related plant spe-
cies often has to be analyzed at the family or genus
level. Recently, meta-barcoding has emerged as a suit-
able alternative for pollen analysis (Keller et al., 2015;
Kraaijeveld et al., 2015; Richardson et al., 2015;
Valentini et al., 2010). However, due to a missing con-
sensus on the best marker for plant species identifica-
tion and the variety of DNA sequencing platforms
available, different methods and protocols exist (e.g.,
Bruni et al., 2015; Galimberti et al., 2014; Keller et al.,
2015; Kraaijeveld et al., 2015; Richardson et al., 2015),
which makes it difficult to compare independent studies.
We here present a detailed protocol of the method
described recently (Sickel et al., 2015) as a research
standard that is highly cost-efficient and overcomes
those limitations. It is based on ITS2-meta-barcoding,
which has been validated for plant barcoding (Chen
et al., 2010) and for which a comprehensive database
has been established (Ankenbrand, 2015). A variant of
this method has been recently developed also for the
rbcl gene, and analyzing both markers in parallel with
this workflow is recommended for most applications
(Bell et al., 2019). Beside the laboratory process, we
also provide information on data processing
and analysis.
8.1. Meta-barcoding protocol
8.1.1. Required materials
8.1.1.1. Reagents. DNA isolation kit suitable for pollen
grains (e.g., Macherey-Nagel NucleoSpin Food, D€uren,
Germany); PCR grade water; Ethanol (96–100%);
Primers as given in Table 1; Polymerase with proof-
reading ability including dNTPs, GC buffer and co-fac-
tors (e.g., 2 	 Phusion Master Mix); Agarose, suitable
buffer (e.g., TAE), intercalating dye (e.g., Midori Green
Advance, Biozym Scientific GmbH, Hessisch Oldendorg,
Germany), 6 	 loading dye, DNA ladder (e.g., FastRuler
Low Range DNA Ladder, Life Technologies, Carlsbad,
CA, USA); SequalPrepTM Normalisation Kit 96 wells
(Invitrogen, Carlsbad, CA, USA); Bioanalyzer High
Sensitivity DNA Chip (Agilent Technologies, Santa
Clara, CA, USA); dsDNA High Sensitivity Assay (Life
Technologies, Carlsbad, CA, USA); MiSeq Reagent Kit
v2 2	 250 bp (Illumina Inc., San Diego, CA, USA); 1N
NaOH (stock solution); PhiX Sequencing Control v3
(Illumina Inc., San Diego, CA, USA)
8.1.1.2. Laboratory equipment. Microliter pipettes and
tips; Microcentrifuge tubes; Electronic pestle; Bead mill;
Incubator; Vortexer; Table centrifuge; 96 well PCR
plates and PCR foils; 96 well plate cooling block; 96
well plate centrifuge; Thermal cycler; Agarose gel for-
mer; microwave, gel electrophoresis chamber; UV illu-
minator; Bioanalyzer, chip vortexer; Qubit Fluorometer;
Access to an Illumina MiSeq desktop sequencer with
MiSeq Control Software version 2.2 or later.
8.1.2. Pollen acquisition
Pollen sampling should be performed as described in
the pollination chapter of the COLOSS BEEBOOK Vol.
1, chapters 3.1. and 4.1.1 (Delaplane et al., 2013). For
long-term storage, we recommend lyophilization before
freezing at 80 C.
8.1.3. Laboratory workflow
8.1.3.1. DNA extraction. For the DNA extraction step,
we recommend using the Macherey-Nagel (D€uren,
Germany) NucleoSpin Food Kit and following the sup-
plementary guidelines for pollen samples, but equivalent
54 M. G. Campos et al.
extraction procedures may also be comparable. The
DNA extraction steps are as follows:
1. Take 2 g of pollen and add 4ml bidest H2O;
2. Homogenize the sample with an electronic pestle;
3. Take 200 ml (
50mg pollen) of the emulsion and
grind it in a bead mill;
4. Add 400 ml Buffer CF (preheated to 65 C) and 10 ml
Proteinase K and mix carefully;
5. Incubate at 65 C for 30min;
6. Centrifuge the mixture for 10min (>10,000 	 g);
7. Transfer the supernatant into a new microcentrifuge
tube and add 1 vol Buffer C4 and 1 vol ethanol;
8. Vortex for 30 s;
9. Pipette 700 ml mixture onto a NucleoSpin Food
Column placed in a Collection Tube;
10. Centrifuge for 1min at 11,000 	 g;
11. Discard the flow-through;
12. Repeat steps 9-1;
13. Add 400 ml Buffer CQW onto the spin column;
14. Centrifuge for 1min at 11,000 	 g;
15. Discard the flow-through;
16. Add 700 ml Buffer C5 onto the spin column;
17. Centrifuge for 1min at 11,000 	 g.
18. Discard the flow-through;
19. Add 200 ml Buffer C5 onto the spin column;
20. Centrifuge for 2min at 11,000 	 g;
21. Place the spin column into a new 1.5ml microcentri-
fuge tube;
22. Add 100 ml Elution Buffer CE (pre-heated to 70 C)
onto the membrane;
23. Incubate for 5min at room temperature (18–25 C);
24. Centrifuge for 1min a 11,000 	 g;
25. Proceed with amplification or keep frozen until fur-
ther processing.
8.1.3.2. Amplification. This protocol utilizes a dual-
indexing strategy (Kozich et al., 2013) amplifying the
ITS2 region, using the primers ITS-S2F (Chen et al.,
2010) and ITS4R (White et al., 1990), but can be
adapted to amplify other markers (Bell et al., 2019).
The primer sequences are as follows: forward: 50-
AATGATACGGCGACCACCGAGATCTACAC
XXXXXXXX CCTGGTGCTG GT
ATGCGATACTTGGTGTGAAT-30; reverse: 50-
CAAGCAGAAGACGGCATACGAGAT XXXXXXXX
AGTCAGTCAG CC TCCTCCGCTTATTGATATGC-
30, where XXXXXX indicates the variable index
sequences (Table 36). The detailed protocol is
described below:
1. Sample index combinations should be planned before-
hand according to the scheme in Figure 41;
Table 39. Protein content (%) using different distillation times.
Distillation time
(min)
Protein content (%)
n¼ 5
4 20.48 (± 0.07)a
3 20.29 (± 0.25)a
2 20.41 (± 0.04)a
Figure 38. Base peak chromatograms of (a) cherry wine and (b) Cabernet Sauvignon-Central Serbia in positive ion mode. (Panteli
c
et al., 2014).
Standard methods for pollen research 55
2. Prepare 3	 10 ml reaction mixes for each sample con-
taining (also see PCR sample design below for details):
a. 5 ml 2 	 Phusion Master Mix (New England
Biolabs, Ipswich, MA, USA) or equivalent;
b. 0.33 mM each of the forward and reverse primers
(sample-specific combinations of forward and
reverse index sequences);
c. 3.34 ml PCR grade water;
d. 1 ml DNA template.
3. Carry out the PCR with a program of:
a. 95 C for 4min, then;
b. 37 cycles of 95 C for 40 sec;
c. 49 C for 40 sec;
d. 72 C for 40 sec. and
e. a final extension at 72 C for 5min.
4. Combine the triplicate PCR reactions of each sample
and mix well.
For quality control purposes, successful amplification
can be checked on a 1% agarose gel using 5 ml of the
combined PCR product.
The design used in 96-well PCR sample is:
Design 1: Well-equipped laboratories with pipet-
ting robots or 96-channel pipettes can directly fill
each well with a different sample and generate three
replicates of these. This will result in 4	 3 replicateFigure 39. Extracted ion chromatograms and MS/MS spectra ofcyanidin-3-rutinoside (Panteli
c et al., 2014).
Figure 40. (a) Extracted ion chromatograms of phenolic glyceride derivatives with their retention times and accurate masses; (b and
c) mass spectra of two phenolic glyceride derivatives with the same accurate masses, m/z 413 (Ristivojevi
c et al. 2014).
56 M. G. Campos et al.
Figure 41. Planning scheme for samples and the corresponding index-combinations. Roman numbers indicate PCR plate numbers,
bold Arabian numbers on 96 well plates indicate normalization plate number.
Figure 42. Detailed workflow (schematic), suitable for laboratories with limited access to equipment for automated pipetting. Bold
numbers indicated step number of Design 2 in Section 8.1.3.2.
Standard methods for pollen research 57
96-well plates according to Figure 41 used for amplifi-
cation. After amplification one can proceed with
normalization.
Design 2: For laboratories with little equipment for
automated pipetting, the workflow described above is
impractical, since manual pipetting in that format is
time-intensive and pipetting errors can be easily intro-
duced. To facilitate the process, we recommend to
work with all triplicates but only 24 samples on one 96
well plate (Figure 42). This way, 16 PCR plates will be
produced, but pipetting effort is minimized. PCR plate
labeling is therefore of utter importance, for example
with roman numbers, I – XVI to be able to map the
samples back to the scheme in Figure 41. The complete
workflow is shown schematically in Figure 42 and
described in the following:
1. Prepare two PCR master mixes, each containing one
forward primer, corresponding to the samples you
want to amplify; each master mix contains:
a. 200 ml 2 	 Phusion Master Mix (New England
Biolabs, Ipswich, MA, USA) or equivalent.
b. 13.2 ml forward primer (10 mM).
c. 133.6 ml PCR grade water.
2. Place a new PCR plate into a cooling block;
3. Distribute 26 ml of the master mixes into row A
(Master Mix 1) and F (Master Mix 2);
4. Add 1 ml of the correct reverse primer;
5. Add 3 ml of the correct DNA template;
6. Using a pipette set to 10 ml, pipette up and down to
mix and distribute 10 mL each into the two rows
below: from row A into rows BþC; from row F
into rows GþH;
7. Seal with a foil, spin down briefly;
8. Perform PCR;
9. Prepare a 1% agarose gel;
10. After PCR, briefly spin down again;
11. Lift the foil carefully and combine the triplicate reac-
tions, pipette up and down to mix;
12. For gel electrophoresis, add 1 ml of 6	 loading buffer
into the so far unused rows Dþ E;
13. Add 5 ml PCR product to the loading buffer;
14. Briefly spin down;
15. Load the gel, add a DNA ladder;
16. Run the gel (e.g., 25min, 120 V);
17. Check under UV illuminator for successful PCR
amplification;
18. Freeze PCR product until further processing;
19. Continue with normalization.
8.1.3.3. Normalisation. To ensure more equalized
library sizes, DNA amounts in each PCR product are
normalized using the SequalPrepTM Normalisation Kit
(Invitrogen, Carlsbad, CA, USA). For 384 samples, four
normalization plates are needed. After normalization,
samples from each plate will be combined in “plate
pools” for the following quality control.
Design 1: Pool the samples of all three repli-
cates together by keeping the sample scheme.
Transfer 25 ml of PCR products onto the normal-
ization plates. Proceed with the normalization as
described below.
Design 2: For normalization, PCR plates I – IV; V –
VIII; IX – XII and XIII – XVI will be combined to nor-
malization plates 1, 2, 3 and 4. The pipetting scheme is
as follows:
1. Thaw the PCR plates.
2. Briefly spin down.
3. Use four normalization plates and add 25 mL of PCR
product into the wells following this scheme.
a. Normalization Plate 1: PCR plates I –IV
b. PCR plate I: row A !row A; row F ! row B
c. PCR plate II: row A !row C; row F ! row D
d. PCR plate III: row A !row E; row F ! row F
e. PCR plate IV: row A !row G; row F !row H
f. Repeat analogous for the other three normaliza-
tion plates.
g. Proceed with the normalization as
described below.
Design 1 & 2: Continue for both designs with the
normalization:
1. Add 25 ml of Binding buffer;
2. Mix by pipetting up and down or seal the plate with
foil tape, vortex to mix and briefly centrifuge
the plate;
3. Incubate for 1 hour at room temperature; alterna-
tively leave to incubate overnight;
4. Aspirate liquid from wells, do not scrape the
well sides;
5. Add 50 ml Wash buffer, mix by pipetting up
and down;
6. Completely aspirate the buffer from wells, you may
need to invert and tap the plate on paper towels;
7. Add 20 ml of Elution buffer;
8. Mix by pipetting up and down or seal the plate with
foil tape, vortex and briefly spin down;
9. Incubate for 5min at room temperature;
10. Combine 5 ml of each sample (plate-wise) in a new
microcentrifuge tube, mix well;
11. Prepare 1:10 dilutions of each plate pool;
8.1.3.4. Quality control and quantification. Quality
control is performed on a Bioanalyzer High Sensitivity
DNA Chip (Agilent Technologies, Santa Clara, CA,
USA) to ensure that the correct fragment size (peak at
approximately 450 bp; target plus adapters) has been
amplified. Additionally, libraries are quantified using the
dsDNA High Sensitivity Assay on the Qubit fluorometer
(both Life Technologies GmbH, Darmstadt, Germany)
in order to combine the four plate pools equimolarly to
the final sequencing library. We recommend preparing
58 M. G. Campos et al.
three independent concentration measurements per
plate pool.
Bioanalyzer:
1. Prepare a Bioanalyzer Chip according to
the protocol;
2. Allow all reagents to equilibrate to room
temperature;
3. If not ready, prepare a gel-dye mix;
4. Add 15 ml of the dye concentrate (blue lid) to a gel
matrix vial (red lid);
5. Vortex well and spin down, transfer to spin filter;
6. Centrifuge at 2240 	 g for 10min;
7. Protect solution from light, store at 4 C, use within
6 weeks;
8. Put a new chip on the chip priming station;
9. Pipette 9 ml gel-dye mix into the well marked with a
white “G”;
10. Close the chip priming station, with the plunger at
position 1ml;
11. Press plunger until held by the clip;
12. Wait for 60 s then release clip;
13. After 5 s slowly pull back the plunger to the
1ml position;
14. Open the priming station, pipette 9ml gel-dye mix in
the wells marked with black ‘G’s;
15. Pipette 5 ml marker (green lid) into all sample wells
and the ladder wells;
16. Pipette 1 ml of ladder (yellow lid) in the well-marked
with a ladder symbol;
17. In each sample well, pipette 1 ml of sample (concen-
trated and diluted Plate pools) or 1 ml marker
(unused wells);
18. Put the chip horizontally in the adapter and vortex
for 1min at 2400 rpm;
19. Run the chip within 5min;
20. The samples are of sufficient quality, if the electrophero-
grams show a single peak at approximately 450bp; this
peak can be rather wide due to different lengths of the
ITS2 region, a minor peak shortly after the lower
marker is acceptable and corresponds to left-over pri-
mer dimers, which will not interfere with sequencing.
8.1.3.5. Quantification.
1. Measure concentrations of plate pools with the
dsDNA High Sensitivity Assay on the Qubit
Fluorometer.
2. Mix 1 	 n ml Qubit reagent with 199 	 n ml Qubit
buffer (working solution).
3. For each measurement, mix 180–199 ml working solu-
tion with 1–20 ml sample.
4. Vortex and incubate at room temperature for 2min.
5. Combine plate pools to final library equimolarly, start-
ing with the least concentrated library of which take
20 ml.
6. Quantify the final pool and dilute to 2 nM, if final pool
contains less than 2 nM proceed without dilution.
8.1.4. Sequencing
Before sequencing, a sample sheet has to be prepared.
This can be done at the MiSeq instrument using the
Illumina Experiment Manager (IEM). However, due to
the high number of samples, we recommend to pre-
pare the sample sheet in advance, which can be done
on a different computer with the IEM installed, and
load it on to the MiSeq instrument when starting the
sequencing procedure. Due to using custom index
sequences, a new assay has to be added to the IEM,
this is described in the Supplemental Material of
Kozich et al. (2013).
For library dilution, we follow the Illumina Sample
Preparation Guide for a 2 nM library, with some modifi-
cations. In order to increase read quality, 5% PhiX con-
trol is added to the sample library. Additionally, the
reagent cassette of the sequencing kit (e.g., Illumina
MiSeq Reagent Kit v2 2	 250 bp) is spiked with the cus-
tom Read1, Read2 and index primers (for primer
sequences, see Table 1).
8.1.4.1. Sample library.
1. Remove Buffer HT1 from freezer.
2. Prepare a fresh dilution of 0.15 N NaOH (less than a
week old).
3. Mix 5ml of the sample library with 5 ml of 0.15
N NaOH.
4. Vortex briefly and centrifuge at 280 	 g for 1min.
5. Incubate at room temperature for 5min.
6. Add 990 ml Buffer HT1 (10 pM library).
7. Mix 480 ml of 10 pM library and 120 ml Buffer HT1
(8 pM library).
8.1.4.2. PhiX control.
1. Thaw PhiX control at room temperature.
2. Mix 2 ml 10 nM PhiX control with 3 ml H2O
(4 nM PhiX).
3. Add 5 ml 0.15 N NaOH.
4. Vortex briefly and centrifuge at 280 	 g for 1min.
5. Incubate at room temperature for 5min.
6. Add 990 ml Buffer HT1 (20 pM PhiX).
7. Mix 375 ml of 20 pM PhiX and 225 ml Buffer HT1
(12.5 pM PhiX).
8. Mix 570 ml 8 pM library with 30 ml 12.5 pM PhiX.
8.1.4.3. Preparing reagent cassette and loading
the sample.
1. Remove the reagent cassette from the freezer.
2. Rename the sample sheet to match the barcode of
the reagent cassette.
3. Place in water bath, do not fill higher than maximum
water line.
4. Prepare 3 ml each of Read1, Read2 and index pri-
mers in new microcentrifuge tubes.
5. Remove cassette from water bath, dry with
paper towel.
6. Invert the cassette several times to mix.
Standard methods for pollen research 59
7. Inspect wells, make sure all reagents are thawed and
there are no precipitates.
8. Gently tap the cassette on the bench to remove
air bubbles.
9. With a 1000 ml pipette tip, break the foils over wells
12–14 and well 17.
10. With a 100 ml pipette set to 75 ml, transfer the read
and index primers to the following wells of the
reagent cartridge: Read1 ! Well 12; Index ! Well
13; Read2 ! Well14, mix well by pipetting up
and down.
11. Load 600 ml of the spiked library to well 17.
12. Load the cassette, PR2 bottle and flow cell as
prompted by the instrument.
13. Sequence.
8.2. Bioinformatics
8.2.1. Required software
a. up to date Linux or Unix-based OS.
b. usearch, version v11 (Edgar, 2010), if necessary add
location to your system PATH.
c. SeqFilter, https://github.com/BioInf-Wuerzburg/SeqFilter
d. usearch additional python scripts, https://drive5.
com/python
8.2.2. Classification
8.2.2.1. Reference databases.
a. Reference database for direct classification
We recommend to first use a reference database that is most
specific to your project design and limited in represented taxa.
This helps strongly with the quality of the assignments. Examples
for restrictions are species lists for the biogeographic region or
specific taxonomy groups of interest. Such databases can be
created with the BCdatabaser software (Keller et al., 2020). If
you have unpublished sequences of your laboratory that should
be considered as well, add them into this file.
a. Reference database for hierarchical classification
All remaining sequences not classified with the direct database
can be additionally classified to a hierarchical taxonomic level as
deep as possible. For this it is important to have a database that
covers all taxonomic groups. One such database could be for
ITS2 accessible under the DOI https://doi.org/10.5281/zenodo.
3339028 which was also created with the BCdatabaser.
Table 41. International data on the lipid content (% DM) of
bee pollen.
Crude lipid content (%D.M)
Type of samples Number of samples Values References
Multifloral 21 4.0–5.5 (Herbert & Shimanuki, 1978)
Multifloral 3 2.0–18.0 (Echigo et al., 1986)
Multifloral 20 4.8–7.1 (Serra-Bonveh
ı & Escol
a Jord
a, 1997)
Monofloral 172 0–11.2 (Somerville, 2001)
Multifloral 15 3.6–8.9 (Villanueva et al., 2002)
Multifloral 10 7 ± 2
 (Almeida-Muradian et al., 2005)
Multifloral 6 batches 4.97

 (Melo et al., 2005)
Multifloral (Korea) 9 3.3–9.8 (Szcze^sna, 2006a)
Multifloral (China) 5 3.5–8.8 (Szcze^sna, 2006b)
Multifloral (Poland) 13 6.74–10.99 (Szcze^sna, 2006c)
Multifloral 154 4.0–13.3 (Martins et al., 2011)
Multifloral 7 4.3–5.0 (Stanciu et al., 2012)
Multifloral 14 6.3–8.7 (Balkanska & Ignatova, 2012)
Monofloral 12 0.6–6.5 (Yang et al., 2014)
Multifloral 7 batches 5.39 ± 0.60
 (Arruda et al., 2013)
Multifloral 99 1.15–6.97 (Liolios et al., 2014)
Monofloral 43 0.4–13.6 (Liolios et al., 2014)

Average and standard deviation; 

 Average.
Table 40. Protein content (%) analyzing different quantities
of pollen.
Pollen quantity (g)
Protein content (%)
n¼ 5
0.5 19.12 (± 0.16)b
1 20.59 (± 0.03)a
2 20.64 (± 0.06)a
Different letter shows statistically significant changes (Duncan’s mul-
tiple range test, a¼ 0.005).
Table 42. Total lipid content (%) by using different amount of
pollen for the analysis.
Pollen amount (g)
Total lipid content %
(n¼ 5)
1 2.24 (±0.113)a
2 2.87 (±0.056)b
3 2.83 (±0.098)b
The elution time was equal (5 h) and the petroleum ether served as
solvent. Different Latin letter indicates statistically significant differ-
ences, according to Duncan’s multiple range test (a¼ 0.05).
Table 43. Total lipid content (%) in pollen samples, by using
different elution times (3 h, 5 h, 7 h).
Elution time (h)
Total lipid content %
(n¼ 5)
3 2.56 (±0.080)a
5 2.95(±0.040)b
7 2.96 (±0.025)b
The pollen amount was 2 g and the petroleum ether served as solv-
ent. Different Latin letter indicates statistically significant differences,
according to Duncan’s multiple range test (a¼ 0.05).
60 M. G. Campos et al.
8.2.2.2. Preparation and classification of sequencing data. The sequence reads created in step 2.3.5 have to be
joined, quality filtered and classified. For this purpose:
1. Copy all R1 and R2 fastq files and the reference database folders into a single folder.
2. Navigate on the command line shell to this folder.
3. Merge forward and reverse reads:
usearch -fastq_mergepairs 
R1
.fastq \
-relabel @ \
-fastq_minmergelen 150 -fastq_maxmergelen 350 \
-fastqout analyse_merged.fq -fastq_maxdiffs 15
4. Quality filtering and format parsing:
usearch -fastq_filter analyse_merged.fq
-fastq_maxee 1 -fastaout reads.fa
sed"s/
>\([a-zA-Z0-9-]
\)\.\([0-9]
\)/>1_\2;barcodelabel¼\1;/g"\
reads.fa>reads_bc.fa
5. Cluster amplicon sequence variants:
usearch -fastx_uniques reads_bc.fa \
-sizeout -minuniquesize 3
-fastaout reads_bc.derep.fa
usearch -sortbysize reads_bc.derep.fa \
-fastaout reads_bc.derep.sort.fa
usearch -unoise3 reads_bc.derep.sort.fa \
-zotus reads_bc.zotus.fa \
-tabbedout unoise3.txt
6. Classify with direct reference database:
usearch -usearch_global reads_bc.zotus.fa \
-db PATH-TO-DIRECT-DB/sequences.tax.fa \
-id 0.97 -uc reads_bc.zotus_direct.uc -strand both
7. Classify remaining unclassified sequences with hierarchical reference database:
grep "
N" reads_bc.zotus_manRefs.uc j cut -f 9 > nohits
SeqFilter reads_bc.zotus.fa --ids nohits \
-o reads_bc.zotus.nohits.fa
usearch -sintax reads_bc.zotus.nohits.fa \
-db PATH-TO-HIERARCHICAL-DB/sequences.tax.fa \
-tabbedout reads_bc.zotus.sintax -strand plus \
-sintax_cutoff 0.8
8. Combine classifications:
cut -f1,4 reads_bc.zotus.sintax> reads_bc.zotus.sintax.cut
cat reads_bc.zotus.sintax.cut j \
sed -E -e "s/\_[0-9]þ//g" -e "s/,s:.
$//" \
> reads_bc.zotus.sintax.cutx
cut -f 9,10 reads_bc.zotus_ direct.uc j \
grep -v "
" j sed "s/[A-Za-z0-9]
;tax¼//" \
> reads_bc.zotus.cutx
echo ",kingdom,phylum,order,family,genus,species" > header.cutx
cat header.cutx reads_bc.zotus.cutx \
reads_bc.zotus.sintax.cutx>taxonomy.data
9. Count read abundance in the sample data:
usearch -usearch_global reads_bc.fa -strand plus \
-db reads_bc.zotus.fa -id 0.97 -uc zotus.uc
python2.7 PATH-TO-PYTHON-SCRIPTS/uc2otutab.py sample_zotus.uc \
> combined.txt
sed -e "s/OTUId//" combined.txt>combined.out
This procedure will end with the following final files used for further analyses:
(a) combined.out (the abundance of each taxonomic unit in the samples)
(b) taxonomy.data (taxonomy assignments)
Standard methods for pollen research 61
Table 45. Literature data on the carbohydrate content of bee pollen calculated as: 100 – (g fatþ g proteinþ g ashþ g water) (g/
100 g DM).
Type of samples Number of samples Values References
Monofloral
(Aloe greatheadii var. davyana)
6 59.5 (±1.3)
 Human and Nicolson (2006)
Multiflora 22 67.7 (±2.6)

61.2–70.6


Feas et al. (2012)
Multifloral 8 69.68–84.25

 Nogueira et al. (2012)
Monoflora
(Brasica napus L., Citrullus lanatus L.,
Camellia japonica L., Dendranthema
indicum L., Fagopyrum esculentum L.,
Helianthus annus L., Nelumbo nucifera
Gaertn., Papaver rhoeas L., Rosa
rugosa, Schisandra chinensis, Vicia faba
L., Zea mays L.)
12 59.43–77.82

 Yang et al. (2014)
Multifloral – 13–55

 Bogdanov (2015)
Multifloral 26 64.42–81.84

 Kosti
c et al. (2015)

Average and standard deviation; 

min – max.
Table 44. Total lipid content (%) in pollen samples, by using different solvents (acetone, pet-
roleum ether, hexane).
Solvent
Total lipid content %
n¼ 5
Acetone 2.99 (±0.155)a
Petroleum ether 2.91 (±0.084)a
Hexane 2.9 (±0.268)a
The pollen amount was 2 g and the elution time was 5 h.
Table 46. Literature data on the content of sugars in bee pollen (g/100 g DM).
Type of samples/
Country of origin Number of samples Sugars Values
Method
used References
Multiflora (with a
predominance of Cistus
landaniferus)/ Spain
20 Fructose 15.20–23.90
 GC-FID
Serra-Bonveh
ı & Escol
a
Jord
a (1997)
Glucose 10.86–18.19

Saccharose 4.20–9.40

Maltose 0.79–3.20

F/G



 1.13–1.53

Multifloral/Poland 95 Fructose 13.38–21.80
 GC-FID
SzczeRsna et al. (2002)Glucose 5.99–18.17

Saccharose 3.74–7.59

Maltose 0.96–3.32

F/G



 1.03–2.51

Multifloral/ Poland,
South Korea, China
27 Fructose 9.74–22.06
 HPLC-RID
SzczeRsna (2007)Glucose 8.45–20.06

Saccharose 1.13–4.84

Maltose 1.07–3.51

F/G



 1.02–1.63

Multifloral/Israel, China,
Romania, Spain
5 Fructose 15.9–19.9
 HPLEC-PAD
Qian et al. (2008)Glucose 8.2–13.1

Saccharose 14.8–18.4

F/G



 1.33–1.66

Multifloral/Romania -


 Fructose 19.31

 HPLC-RID
Bobis et al. (2010)Glucose 17.86


Saccharose 0.85


Maltose 0.64


Multifloral/Brazil 154 Fructose 12.59–23.62
 HPLC-RID
Martins et al. (2011)Glucose 6.99–21.85

F/G



 1.01–2.24
Multifloral/Romania 16 Fructose 8.44–15.39
 HPLC-RID
Margaoan
et al. (2012)





Glucose 4.37–16.14


F/G



 0.78–1.44

Monofloral (Cucurbita
pepo Thunb, Phoenix
dactylifera L., Helianthus
annus L., Medicago sativa
L., Brasica napus L.)/
Saudi Arabia
-


 Fructose 17.13–21.30
 HPLC-RID
Taha (2015)Glucose 15.44–17.06

F/G



 1.07–1.25


min – max; 

average; 


no information; 



fructose to glucose ratio; 




results for fresh pollen.
62 M. G. Campos et al.
These files can be loaded directly into R with the
package phyloseq (McMurdie & Holmes, 2013). In add-
ition, also the results of the intermediate steps are
retained and can be used for troubleshooting, archiving
or further analyses.
8.3. Data analysis
8.3.1. Required software
a. Up to date R distribution (R Core Team, 2014).
b. R package: phyloseq (McMurdie & Holmes, 2013);
https://joey711.github.io/phyloseq
8.3.2. Prepare sample meta-data
Additionally, to the sequencing data, phyloseq is intended
to read in a meta-data file. This can be a simple
spreadsheet in csv format with any number of variables
to be investigated. For example, continuous vectors like
“altitude” or “temperature” or categorical factors as
“bee species” or “site” can be used. For this, open the file
with your preferred text-editor or spreadsheet applica-
tion and add columns according to the sampling design.
Save the file again in tab-separated format.
8.3.3. Importing data
The data generated above can be directly imported into
R as a phyloseq class object. This allows a variety of
analytical procedures and is recommended. However,
other software tools handling community datasets may
be equally well used for the task of analyses. The fol-
lowing are R scripts, which can be directly used on
the console:
Table 47. Methods used to evaluate the vitamin content of bee pollen.
Method Vitamins analyzed Country Reference Vitamin content (mean)
Open column
chromatography and
microfluorimetric
method
C, beta-carotene as
provitamin A
Brazil Almeida-Muradian
et al., 2005
Absence of vitamin C and
beta-carotene (dried
bee pollen)
Open column
chromatography and
titrimetric method
E, C and beta-carotene
as provitamin A
Brazil Oliveira et al., 2009 13,5mg/g (E);
56,3–198,9 mg/g (beta-
carotene);
273,9–560,3 mg/g (C)
(fresh bee pollen)
HPLC, open column
chromatography and
titrimetric method
E, C and beta-carotene
as provitamin A
Brazil Melo et al., 2009 114–340mg/g (C);
16,27–38,64 mg/g (E);
3,14–77,88 mg/g (beta
carotene)
(dried bee pollen)
HPLC, open column
chromatography and
titrimetric method
E, C and beta-carotene
as provitamin A
Brazil Melo & Almeida-
Muradian, 2010
14–119 ug/g (C);
19.43–37.19 mg/g (E);
5.95–99.27 mg/g (beta-
carotene)
(fresh bee pollen)
HPLC method with
fluorescence detection.
B1, B2, B6 and PP Brazil Arruda et al., 2013a;
Arruda et al., 2013b
0.64mg/100 g (B1);
0.60mg/100 g (B2);
12.18mg/100 g (PP);
0.55mg/100 g (B6)
HPLC and
titrimetric method
E (alpha, beta, gamma
and delta tocoperol),
C (ascorbic acid) and
provitamin A (alpha
and beta-carotene)
Brazil Sattler, 2013 60–797 mg/g (C);
0.57–11.7mg/100g (E);
0–179.21 mg/g of beta-
carotene; 0–828.43 mg/g
of alpha-carotene
HPLC method with
fluorescence detection.
Brazil Souza, 2014 0.46–1.57mg/100g (B1);
0.40–1.86mg/100g (B2);
2.67–7.74mg/100g (PP
or B3);0.78–7.13mg/
100g (B6)
Table 48. Calibration curves obtained for echimidine, lycopsamine and intermedine.
Echimidine or Lycopsamine or Intermedine
(mg/ml)
Echimidine
(intensity)
Intermedine
(intensity)
Lycopsamine
(intensity)
0.02 2950 3130 3670
0.1 13660 15540 16200
0.5 63940 62050 71200
2 243800 200400 277000
Calibration curve
 y¼ 122755x þ 686 y¼ 104410x þ 1900 y¼ 138740x þ 1143

Weighted by 1/x.
Standard methods for pollen research 63
1. Load packages and data:
library(phyloseq) # load the package
setwd("<PATH-TO-DATA>") # data folder
tax <- tax_table(as.matrix(read.table("taxonomy.data",
header¼T,row.names ¼ 1,fill¼T,sep¼",")))
otu <- otu_table(read.table("zotu_table.txt", header¼T, row.names ¼ 1),
taxa_are_rows¼T)
map <- sample_data(read.table("microbiome_DATA_SamBoff.csv", sep¼",",
header¼T, row.names ¼ 1,fill¼T))
data <- merge_phyloseq(otu, tax, map) # create phyloseq object
2. Adjust for missing taxa names in hierarchical classification:
tax_table(data)[tax_table(data)[,"phylum"]¼¼"","phylum"]<- paste(
tax_table(data)[tax_table(data)[,"phylum"]¼¼"","kingdom"],
"_spc",sep¼"")
tax_table(data)[tax_table(data)[,"order"]¼¼"","order"]<- paste(tax_table(
data)[tax_table(data)[,"order"]¼¼"","phylum"],
"_spc",sep¼"")
tax_table(data)[tax_table(data)[,"family"]¼¼"","family"]<- paste(
tax_table(data)[tax_table(data)[,"family"]¼¼"","order"],
"_spc",sep¼"")
tax_table(data)[tax_table(data)[,"genus"]¼¼"","genus"]<- paste(
tax_table(data)[tax_table(data)[,"genus"]¼¼"","family"],
"_spc",sep¼"")
3. Combine Amplicon Sequence Variants to species or genus level:
(data.genus <- tax_glom(dataset.comp, taxrank¼"genus"))
(data.species <- tax_glom(dataset.comp, taxrank¼"species"))
4. Transform to relative abundances and filter low abundance taxa:
data.genus.rel¼transform_sample_counts(data.genus,
function(x) x/sum(x))
otu_table(data.genus.rel)[otu_table(data.genus.rel)<0.01 ]<-0
Table 49. Retention and mass characteristics of known Echium-type and Eupatorium-type PAs (free bases/-N-oxides and isomers).
PA peak n Compound Name
RT
(min) Chemical formula [MþH]þ experimental [MþH]þ calculated Error (mDa)
 Echium vulgare
1 Echimidine 2.28 C20H31NO7 398.2184 398.2179 0.5
Echimidine-N-oxide 2.27 C20H31NO8 414.2126 414.2128 0.2
Acetylechimidine 2.74 C22H33NO8 440.2282 440.2284 0.2
3 Acetylechimidine-N-oxide 2.71 C22H33NO9 456.2234 456.2234 0.0
Echivulgarine 3.81 C25H37NO8 480.2575 480.2575 0.0
5 Echivulgarine-N-oxide 3.83 C25H37NO9 496.3402 496.3405 0.3
Vulgarine 2.33 C20H31NO7 398.2174 398.2179 0.5
2 Vulgarine-N-oxide 2.43 C20H31NO8 414.2133 414.2128 0.5
Acetylvulgarine 2.88 C22H33NO9 440.2285 440.2284 0.1
4 Acetylvulgarine-N-oxide 2.87 C22H33NO9 456.2243 456.2234 0.9
 Eupatorium cannabinum
6 Lycopsamine 1.12 C15H25NO5 300.1807 300.1811 0.4
7 Intermedine 1.16 C15H25NO5 300.1810 300.1811 0.1
8 Lycopsamine-N-oxide or
Intermedine-N-oxide

1.27 C15H25NO6 316.1756 316.1760 0.4
9 Lycopsamine-N-oxide or
Intermedine-N-oxide

1.35 C15H25NO6 316.1758 316.1760 0.2

Distinction between the N-oxides of intermedine and lycopsamine not possible due to the lack of standards.
64 M. G. Campos et al.
After completion of the tasks above, the dataset is in
a condition where individual analyses can be started.
The tutorials at the repository of phyloseq ((McMurdie
& Holmes, 2013); https://joey711.github.io/phyloseq)
provide a good starting point for this.
8.3.4. Recommended packages for further analysis
Whilst phyloseq provides basic tools suited for most
purposes, the modularity of R packages allows a variety
of more and deeper analyses. It is not possible to dis-
cuss all the features here, yet we provide a list some of
the major packages relevant for community ecology and
pollination studies:
a. vegan: comprehensive community ecology package;
b. picante: phylogenetic diversity indices;
c. bipartite: interaction network ecology;
d. edgeR: sequencing library size correction.
9. Standard method for the determination of
protein content of pollen using the
Kjeldahl method
9.1. Introduction
The knowledge of bee pollen chemical composition
increases the commercial value of the product.
Especially, among the chemical constituents of bee pol-
len, the proteins’ concentration has been the most
studied (Table 37). There seems to be a considerable
variation among the results regarding the protein con-
centration, probably because of the different number of
the analyzed samples and whether the pollen samples
are unifloral or polyfloral.
For the determination of the total protein content,
the majority of authors use the Kjeldahl method modi-
fied for bee pollen analysis by Rabie et al. (1983).
However, the method of analysis has not been investi-
gated. Regarding the calculation of the nitrogen content
in this type of samples, several authors used different
factors (6.25 or 5.60). Indeed, Rabie et al. (1983), rec-
ommended the factor 5.60, while Serra-Bonveh
ı and
Escol
a Jord
a (1997) stated that the use of factor 6.25
would overestimate the protein content of analyzed
samples. The follow methods elucidate the optimum
conditions, in order to minimize the reagents’ consump-
tion and the sample quantity.
9.2. Determination of protein content by
Kjeldahl method
1. Weigh 1 g of bee pollen.
2. Heat the sample with 20ml of sulfuric acid (95–97%)
at the presence of a catalyst (potassium sulfate, cop-
per sulfate) for about 4 h until the solution becomes
clear and blue-green in color.
3. Neutralize the digested sample with 90ml NaOH
(30%) and distillate the solution for 2min using for
the trapping of NH3, 30ml of H3BO3 solution (4%).
4. Titrate with HCl solution (0.1 M).
The nitrogen content is determined by the consumption of
the HCl solution (0.1M) during the titration. The crude pro-
tein content is estimated using the most suitable factor 5.60,
which was calculated and suggested by Rabie et al. (1983)
after amino-acids analysis in different pollen species.
9.3. Development of the method
For the development and validation of the suggested
method, bee pollen of Papaver rhoeas was used. In order
to find the optimum conditions for the determination of
the protein content of bee pollen tests were applied
regarding the volumes of NaOH, and H3BO3 solutions,
the quantity of sample and the distillation time.
9.3.1. Volume of NaOH solution
During the distillation, the NaOH solution converts any
nitrogen in the bee pollen into NH3, while the appear-
ance of the blue color demonstrates a sufficient volume
of the solution for this formation. Volumes of 50ml,
60ml, 70ml, 80ml, 100ml and 110ml were tested. No
change of color was observed in volumes of 50ml to
80ml, while in volume of 90ml, the color turned from
light green into dark blue indicating the end of the reac-
tion. The volumes of 100ml and 110ml were rejected,
to avoid the unnecessary consumption and to reduce
the cost of the analysis. Thus, the suitable volume of
the NaOH solution (30%) was determined at 90ml.
Table 50 MS/MS fragment ions for PA-N-oxides from E. vulgare or E. cannabinum.
Pyrrolizidine alkaloid [MþH]þ MS/MS fragment ions
 Echium vulgare
Echimidine-N-oxide 414.2126 396.2029, 352.1764, 338.1608, 254.1403, 220.1350, 120.0817
Acetylechimidine-N-oxide 456.2234 438.2131, 396.2025, 338.1610, 254.1393, 220.1339
Echivulgarine-N-oxide 496.3402 478.2448, 396.2030, 338.1613, 254.1397, 220.1341, 120.0816
Vulgarine-N-oxide 414.2133 396.2021, 314.1612, 256.1189, 172.0976, 138.0923, 136.0764
Acetylvulgarine-N-oxide 456.2243 438.2132, 356.1714, 298.1295, 214.1081, 180.1030, 120.0818
 Eupatorium cannabinum
Lycopsamine/Intermedine-N-oxide 316.1756 172.0973, 155.0947, 138.0917, 111.0684
Standard methods for pollen research 65
9.3.2. Volume of H3BO3 solution
Boric acid is used for the capture of NH3, during distil-
lation. Volumes of 30ml and 40ml were tested without
statistically significant changes to be observed between
the results (Duncan’s multiple range test, a¼ 0.05)
(Table 38). Eventually, the volume of 30ml of the
H3BO3 solution (4%) was chosen for the analysis.
9.3.3. Distillation time
Different distillation periods of time (2, 3, 4min) were used
in order to choose the most appropriate. In all cases, 1 g of
sample, 90ml NaOH and 30ml H3BO3 were used.
According to the statistical process of the results, no signifi-
cant changes were observed in the protein content regard-
ing to the distillation time (Table 39). For this reason, the
shortest time of 2min was chosen for the analysis.
9.4. Sample preparation
For the chemical analysis of bee pollen of a single plant,
the pollen pellets should be separated from the mixture
according to its color, shape and texture and this is a
tedious and time consuming process. Consequently, the
minimum amount of pollen for chemical analysis will
reduce the needed time and facilitate the analysis.
1. Different quantities of pollen 0.5 g, 1 g and 2 g were
tested with five replications in each case, maintaining
the other parameters constant.
2. The volumes of 90ml NaOH and 30ml H3BO3 were
used, while the distillation time was 2min.
3. Statistically significant changes were observed regard-
ing to the protein content (Table 43). The minimum
quantity of 0.5 g gave significantly less amount of pro-
teins, while no differences were found when one or
two grams were used.
4. After that, we propose the use of one gram sample
for the analysis of proteins in bee pollen samples.
9.5. Validation of the method
For the determination of the repeatability of the method:
1. eight pollen samples of Papaver rhoeas of the same
quantity (1 g) were analyzed;
2. samples were treated equally during the three steps
of the Kjeldahl method, as described previously;
3. average protein content of Papaver rhoeas pollen was
found 20.6%, while the low value of RSD (0.81%) indi-
cated the very good repeatability of the method.
9.6. Application of the method
Pollen from different botanical species has different nutritional
value for honey bees, affecting the development and the prod-
uctivity of the bee colony. To investigate qualitative and quanti-
tative differences on pollen composition we collected bee
pollen samples from the apiary of Aristotle University in con-
secutive years during the full season and samples from bee-
keepers all over Greece. The samples were analyzed as a
mixture or as separate pollen pellets that are classified accord-
ing to their color, shape and texture. For the identification of
the pollen microscopic method is applied. Monofloral and mul-
tifloral bee pollen samples were analyzed using the method
described above. Some of the results regarding the differences
based on plant origin are presented below, in Figure 43. The
lowest value was observed for Actinidia chinensis (9.7%) and the
highest for Crocus sativa (26.1%). The range of 16.4%, confirms
the large variation among pollen from different plants. The
protein content of mixed bee pollen samples was also exam-
ined during the whole beekeeping season, in order to find out
differences based on the harvesting period (Figure 44).
The mixed pollen samples collected from March till the
middle of June indicated that the nutritional requirements
of bees are very well met since the protein content ranged
from 19.7% to 26.6%. Herbert (1992) mentions that a pro-
tein content of 20–23% in pollen substitutes is ideal for the
dietary requirements of honey bees. As opposed to that,
pollen collected from the middle of June till the middle of
September had average protein content below 20%, includ-
ing the lowest observed value (14.8%) in the end of August.
During autumn, the protein content of pollen increased
again and exceeded 20%. According to these results, it is
obvious that the protein content of mixed pollen ranges
according to the season and this could affect the colony’s
development and the quality of the bee product for sale.
10. Standard method for determination of lipid
content of pollen using Soxhlet extraction
10.1. Introduction
The lipid fraction is one of the basic constituents of bee pol-
len and varies significantly among the different botanical spe-
cies (Somerville, 2001; Yang et al., 2014). Several authors
have studied the lipid concentration in pollen, whose results
are referred in Table 41. The variations among the research
observed, could be attributed to the different number of the
analyzed samples and to variable botanical origin of the sam-
ples. The most common method for the determination of
the total lipid content of foods is the extraction method
described by Soxhlet. According to this procedure, oil and
fat are extracted by repeated elution with an organic solvent
(extraction cycle). Nevertheless, in case of the extraction of
lipids in pollen, there is vagueness about the optimum meth-
odology concerning the reagent’s consumption, the sample
quantity and the elution time. The following methods clarify
these parameters and provide experimental results that
should be implemented to produce reliable analytical data
on lipids analysis of pollen.
10.2. Methodology for total lipid content
determination
The total lipid content is determined as crude fat from
a solvent extract.
66 M. G. Campos et al.
1. weigh 2 g of bee pollen and place in a porous cellulose
thimble (25	 100mm).
2. place the thimble in an extraction chamber, which is
suspend above a round bottom flask containing petrol-
eum ether as solvent and below a condenser.
3. heat the flask – evaporate the solvent and move up
into the condenser where it is converted into a liquid
that trickle into the extraction chamber containing the
sample. This extraction cycle is repeated many times
for 5 hours. In order to remove the excess solvent.
4. fat residue is dried in an oven at 80 C for 45min until
to constant weight.
5. place the flask for 15minutes in a desiccator which
contains CaCl2, in order to cool down without
absorbing moisture.
6. estimate the fat content as the difference in weight of
the flask before and after the extraction. Express as
percent of dry matter.
10.3. Development of the method
For the development and validation of the previous
method, pollen of Papaver rhoeas was chosen.
1. Pollen samples: collected from traps placed in bee
hives (in this example they were located in the sur-
rounding area of Thessaloniki). The specific plant taxon
was chosen, because of its abundance in the area
where the bee hives were located and its regular for-
aging by the bees. Moreover, its pollen loads possess
a distinct black color, so they can be separated easily
from mixed pollen samples.
2. Determination of the optimum quantities of the mate-
rials and the appropriate analysis’ conditions of the
method – tests implemented regarding the pollen
amount (1 g, 2 g, 3 g), extraction time (3 h, 5 h, 7 h)
and the type of the solvent (acetone, petroleum ether,
hexane). Statistically significant differences were found
in total lipid content (Duncan’s multiple range test,
a¼ 0.05), using different amounts of pollen (Table 42)
and different elution times (Table 43), while no signifi-
cant differences were observed by using different sol-
vents (Table 44). Note: Considering that the
separation of pollen samples based on color is a tedi-
ous and time consuming process, the quantity of pol-
len should minimize as possible. As a consequence,
according to the Table 2, the amount of 2 g was
chosen for the analysis.
3. Elution time the period of 5 h was chosen. The time
of 7 h was rejected in order to reduce the time con-
sumption, while when 3 h elution was applied statistic-
ally significant differences were observed in the total
lipid content.
4. Solvent extraction – no statistically significant differen-
ces were found among the solvents and petroleum
ether is widely referred in bibliography for the
extraction of the lipids in pollen this liquid was chosen
as the appropriate solvent.
10.4. Validation of the method
For the determination of the repeatability of the
method, Papaver rhoeas pollen was separated, homogen-
ized and eight subsamples of 2 g analyzed under the
same conditions (extraction time: 5 h, solvent: petrol-
eum ether). The average liquid content of Papaver
rhoeas was found 2.47%, while the relative standard
deviation (RSD) was 4.66% indicating a satisfac-
tory precision.
10.5. Application of the method
The method described above was applied in monofloral
and polyfloral pollen samples collected from University
apiary. Some of the results regarding the differences
based on plant origin are presented in Figure 45. The
identification of pollen taxa was based on the micro-
scopic characteristics (Lau et al., 2018; Louveaux et al.,
1978). The lowest value was observed for Erica manipuli-
flora (0.40%) and the highest for Echinops ritro (13.60%).
The average of the samples examined was 4.61% and
the range was 13.2%.
The lipid content of mixed pollen samples was also
examined during the whole beekeeping season, in order
to find differences based on the harvesting time (Figure
46). The higher lipid content of mixed pollen samples
was observed in samples collected between 19/9 until
15/10. Pollen collected early in autumn (13/9) has very
low lipid content (2.15%) but after that, the lipid con-
centration increases considerably to reach the high level
of 6.22% on the day 15/10. One factor that may con-
tribute to this increase is the second period of bloom-
ing of Sisymbrium irio which is very reach in lipids
(7.22%). The fluctuation of lipid curve of pollen mix-
tures collected in different period of year (Figure 46)
indicates the effect of various blossoms to the final
chemical composition of pollen.
11. Determination of sugar content in pollen
using HPLC-RID
11.1. Introduction
Carbohydrates are the sugars, dietary fiber, and starch
which constitute the main fraction of pollen dry matter
and represent important components from the nutri-
tional and energy point of view. Collected pollen pro-
vides bees a valuable source of simple sugars and other
carbohydrates which are essential for proper nutrition
and advisable for certain physiological conditions. The
generally accepted calculation for total carbohydrate
content in pollen is: 100 – (g fatþ g proteinþ g ashþ g
water) (Bogdanov, 2015; Feas et al., 2012; Human &
Nicolson, 2006; Kosti
c et al., 2015; Nogueira et al.,
2012; Yang et al., 2014) (Table 45). This calculation is
Standard methods for pollen research 67
much higher than results received through HPLC or
GC methods because they do not determine such com-
ponents as dietary fiber, starch, and cell wall material.
Studies dealing with the composition of sugars in pollen
were concerned with the total content of reducing and
non-reducing sugars (Cirnu et al., 1969; Rosenthal et al.,
1969; Solberg & Remedios, 1980; SzczeRsna et al., 1995a,
1995b; Youssef et al., 1978).
Advances in chromatographic techniques with differ-
ent detections (TLC, HPLC-RID, HPTLC, HPLEC-PAD,
HPLEC coupled to LC-ESI-MS, GC-FID; GC-MS) made
it possible to analyze, in detail, the composition of sug-
ars in bee pollen (Aratchi et al., 2018; Bobis et al.,
2010; Echigo et al., 1986; Margaoan et al., 2012; Martins
et al., 2011; Qian et al., 2008; Serra-Bonveh
ı et al.,
1986; Serra-Bonveh
ı & Escol
a Jord
a, 1997; SzczeRsna,
2007; SzczeRsna et al., 2002; Taha, 2015). The quality and
quantity composition of pollen sugars was found to be
closely related to the plant species of origin as well as
to the methods of analysis. Researchers identified
fourteen different sugars in pollen of which fructose,
glucose, and sucrose occurred in the greatest amounts,
followed by maltose (Table 46). The following compos-
ition of basic sugars in bee pollen was determined to
be: fructose 44–46%, glucose 35–37%, sucrose
8–30%,and maltose 5–10% of the total sugar content.
Mono-, di- and trisaccharides furanose, arabinose,
ribose, trehalose, isomaltose, turanose, kojibiose, gen-
tiobiose, melibiose and melezitose accounted for about
1% (Aratchi et al., 2018; Serra-Bonveh
ı et al., 1986;
Serra-Bonveh
ı & Escol
a Jord
a, 1997; SzczeRsna, 2007;
SzczeRsna et al., 2002). The metabolome analysis of bee-
collected pollen, especially myo-inositol (predominant
physiologica form of six-carbon sugar alcohol) and pen-
tose sugars (furanose) is useful in the study of plant-pol-
linators mutualism (Aratchi et al., 2018).
Tetrasaccharide stachyose is also determined in pollen
samples (Quian et al., 2008). The fructose-to-glucose
ratio (F/G) of bee-collected pollen samples varies con-
siderably from 1.01 in Brazil (Martins et al., 2011) to
Figure 43. Crude protein content (% D.M) of bee pollen from selected botanical origins.
Figure 44. Protein content (% D.M) of mixed pollen samples harvested during the beekeeping season of 2011.
68 M. G. Campos et al.
Figure 45. Crude lipid content (% D.M) of bee pollen from selected botanical origins.
Figure 46. Lipid content (% D.M) of mixed pollen samples harvested during the beekeeping season of 2013.
Figure 47. (a) Structure of PAs: Necinic acids are connected to the necine base by ester bond in position 1 and/or 7. Necinic acids
can be connected to form a macrocyclic ester; (b) Example of PA free base and N-oxide macrocyclic esters (Hartmann &
Witte, 1995).
Standard methods for pollen research 69
2.51 in Poland (SzczeRsna et al., 2002) (Table 46). The
sugar content of pollen has not been normalized, and in
fact only Poland, Switzerland and Argentina currently
have a standard for it. The International Honey
Commission proposes 4 g/100 g minimum sugar content
for dried bee pollen (Campos et al., 2008). The HPLC
technique with refractive index detection (RID) has gen-
erally been used for the determination of sugars in bee-
collected pollen (Bobis et al., 2010; Margaoan et al.,
2012; Martins et al., 2011; SzczeRsna, 2007; Taha, 2015).
11.2. Method for identification and quantification of
sugars by HPLC-RID
11.2.1. Principle
This method is based on the procedure elaborated by
Bogdanov et al. (1997) for the determination of honey
sugars and adopted by SzczeRsna (2007). Briefly, the sugars
are extracted from pollen samples and then, determined
by HPLC with a refractive index detector (RID) using a
normal-phase column containing amine-modified silica gel
with a 5–7mm particle size. The analyzed sugars are: fruc-
tose, glucose, saccharose, maltose, turanose, trehalose,
izomaltose, erlose, melezitose, and rafinose. The sugars
are identified through the comparison of individual sugars'
retention times of the reference and the analyzed solu-
tion. The quantitative assays are done by the external
standard method comparing those sugars' peak surfaces.
11.2.2. Reagents
a. deionized water (18.3 MX).
b. methanol, acetonitrile for HPLC (J.T. Baker or com-
parable purity from another company).
c. analytical standards of: fructose, glucose, saccharose,
maltose, turanose, trehalose, izomaltose, erlose, mele-
zitose, rafinose (SIGMA-ALDRICH or comparable pur-
ity from another company).
d. Carrez solution I: 15 g of potassium hexacyanoferrate
(II) trihydrate, K4Fe(CN)6 3H2O dissolved in 100ml
of deionized water.
e. Carrez solution II: 30 g of zinc acetate dihydrate,
Zn(CH3COO)22H2O dissolved in 100ml of deion-
ized water.
11.2.3. Equipment
a. High Performance Liquid Chromatograph operated by
computer software and equipped with: pump, degas-
ser, autosampler, column-oven, and refractive index
detector (RID).
b. amine-modified silica gel HPLC column with 5–7 mm
particle size (250	 4.6mm 	 5 mm) (Phenomenex or
another company).
c. analytical balance.
d. mechanical shaker.
e. automatic dispenser.
f. analytical grade filter paper.
g. PVDF membrane filter, pore size 0.45 mm (Karl Roth
or another company).
h. volumetric flask, 100ml.
i. Eylenmeyer flask, 100ml.
j. beaker, 50ml.
k. autosampler vial.
11.2.4. Preparation of stock standard solution and working
standard solutions of sugars
1. weigh each sugar to prepare a standard solution, with
an accuracy of 0.001 g. Weights of individual sugars:
(a) fructose: 1.000 g, (b) glucose: 0.750 g, (c) sacchar-
ose, maltose: 0.500 g, (d) turanose, trehalose, izomal-
tose, erlose, melezitose, rafinose: 0.075g.
2. dissolve the standard sugars in a small amount of
deionized water and transferred quantitatively to a
volumetric 100ml flask. Add 25ml of methanol with
the use of an automatic dispenser, and poured into a
volumetric flask up to the mark with deionized water.
3. filter the standard solution with 0.45 mm membrane filter
and stored at 18 C. A stock standard sugar solution
kept at this temperature is stable for six months.
Note: a calibration curve is used for the external standard
method of the quantitative analysis of sugar in pollen. The
calibration curve is prepared for three different concentra-
tions of each analyzed sugar. For this purpose, the working
standard solutions of those sugars are prepared on the day
of analysis from the stock standard solution.
11.2.5. Sample preparation
1. weigh a pollen sample of 5 g ± 0.001 g and put into a
100ml Eylenmeyer flask.
2. add about 40ml of deionized water, 5ml of Carrez
solution I, and 5ml of Carrez solution II and mix for
2 h at 30 C using a mechanical shaker.
3. transfer the solution quantitatively to a 100ml volu-
metric flask, add 25ml of methanol and filled up to
the notch, with deionized water.
4. filter the sample solution through qualitative filter
paper into a beaker. The first 10ml of the filtrate are
discarded. The solution is passed through a 0.45 lm
membrane filter into an autosampler vial and the fil-
trate injected in the HPLC-RID system.
11.2.6. HPLC-RID analysis
The HPLC system is run using computer software in
the following conditions:
1. flow rate: 1.5ml/min.
2. mobile phase: acetonitrile: water (80:20 V/V).
3. column and detector temperature: 30 C.
4. injection volume: 20 ml.
After the HPLC System is stabilized, the working
standard solutions of the sugars and then pollen sample
70 M. G. Campos et al.
are injected and analyzed under the same conditions.
The external standard method is used to calculate the
results using a calibration curve determine for three dif-
ferent concentrations of each sugar standard. The
results are express as g/100 g to one decimal place.
11.3. Validation of the method
The method described above can be applied for routine
analysis of fructose, glucose, saccharose, maltose, turanose,
trehalose, izomaltose, erlose, melezitose, and rafinose in
pollen samples. The detection limit, working range, linear-
ity, repeatability and within-laboratory reproducibility
were determined for this method. The detection limit was
set at 0.2 g/100 g DM for fructose and glucose and at 0.5 g/
100 g DM for saccharose, maltose, turanose, trehalose, izo-
maltose, erlose, melezitose, and rafinose. This procedure is
used for the routine determination of sugars in pollen, in
the range of: 5–40 g/100 g for fructose and glucose;
0.5–30 g/100 g for saccharose, maltose, trehalose, tura-
nose, and isomaltose; 0.5–10 g/100 g for erlose, rafinose
and melecytoze. The correlation coefficient of the linear
dependence of sugar concentration and peak area in the
working range, reached the value of over 0.995 for all the
analyzed sugars. A series analysis for three pollen samples
showed that the variation coefficient of the repeatability
and the within-laboratory reproducibility of the elaborated
method did not exceed 10% for fructose and glucose, and
20% for both other analyzed disaccharides (saccharose,
maltose, turanose, trehalose, izomaltose) and trisacchar-
ides (erloze, melezitose). Thus, sufficient accuracy and pre-
cision of the methods were demonstrated.
11.4. Application of the method
The results on the content of the sugars analyzed by the
HPLC-RID method, in multifloral pollen samples originating
from Poland, South Korea and China are presented
SzczeRsna, (2007) in Table 46 and show that the sugar con-
tent of pollen dry matter averages 40%. The fructose ranged
from 9.74 to 22.06DM, and the average was 15.94 g/100 g
DM. The glucose content was lower by several g/100 g and
ranged from 8.45 to 20.06 g/100 g; the average being 14.02 g/
100 g DM. Of the tested disaccharides, saccharose occurred
in the highest concentrations, ranging from 1.13 to 4.84 g/
100 g DM; the average being 3.20 g/100 g DMmaltose fol-
lowed ranging from 1.07 to 3.51 g/100 g DM; the average
being 2.47 g/100 g DM. Those four sugars varied substantially
within each of the samples. The variation was related to
their botanical origin. In the majority of samples, the content
of other such analyzed disaccharides as trehalose and tura-
nose was at the detection limit of 0.2 g/100 g DM.
Bobis et al. (2010), Martins et al. (2011), Margaoan et al.
(2012) and Taha (2015) used the same method as SzczeRsna
(2007) to quantify the sugar content of sugars and con-
firmed the predominance of fructose followed by glucose
in pollen samples (Table 48). Bobis et al. (2010) found fruc-
tose and glucose to be on average at concentrations of
19.31 and 17.86 g/100 g, respectively. They investigated dis-
accharides were in similar quantities as those investigated
by SzczeRsna (2007). Martins et al. (2011) determined the
fructose concentration to be ranged between12.59 and
23.62 g/100 g and the average 18.99 g/100 g, while the glu-
cose concentration between 6.99 and 21.85 g/100 g and the
average 14.89 g/100 g. According to Margaoan et al. (2012),
fructose content ranged between 8.44 and 15.39 g/100 g,
and glucose content – between 4.37 and 16.14 g/100 g. The
results received by Taha (2015) for a few monofloral pol-
lens (Cucurbita pepo Thunb, Phoenix dactylifera L., Helianthus
annus L., Medicago sativa L., Brasica napus L.), showed that
fructose content ranged from 17.13 g/100 g for rape
(Brassica napus L.) pollen to 21.30 g/100 g for date palm pol-
len (Phoenix dacttylifera L.), and glucose content from
15.44 g/100 g for sunflower (Helianthus annus L.) pollen to
17.06 g/100 g for date palm (Phoenix dacttylifera L.) pollen.
With the above-described method, the Bee Products
Quality Testing Laboratory, the Apiculture Division of
the National Institute of Horticultural Research, Poland
produced results for multifloral pollen samples were in
the same range as the results received by the above-
mentioned authors. Fructose content ranged between
15.40 and 22.54 g/100 g DM, 17.01 g/100 g DM, on aver-
age. Glucose content ranged between 10.42 and18.54 g/
100 g DM, 14.01 g/100 g DM, on average. Saccharose
content ranged between 2.80 and 7.40 g/100 g DM, with
an average value of 5.70 g/100 g DM.
12. Standard method for vitamin analysis in
bee pollen
12.1. Introduction
Bee pollen contains vitamins of B complex, vitamin C, E
and carotenes with provitamin A function (alpha and beta-
carotene) (Arruda et al., 2013a, 2013b; Melo et al., 2009;
Melo & Almeida-Muradian, 2010; Oliveira et al., 2009;
Sattler et al., 2015; Souza et al., 2018). Vitamins are classi-
fied in terms of solubility. The lipid-soluble vitamins (vita-
mins A, D, E and K) are a group of chemical substances
with different structure which are soluble in organic sol-
vents and are stored in the body fat, but can be toxic when
consumed in excess. Water-soluble vitamins (vitamins B1,
B2, B6, B12, folic acid, pantothenic acid, niacin, biotin and
vitamin C) are not normally stored in significant amounts in
the body, which leads to the need for a daily supply of these
vitamins (Ball, 1998). Methods used to evaluate the vitamin
content can be seen in Table 47.
12.2. B complex vitamins
12.2.1. Methods for B complex vitamins
Before vitamin analysis, it is necessary to determine the
moisture content, so the results can be expressed in
dry basis. From the current methods for moisture
determination in dried bee pollen samples, the drying
out process with infrared light at 85 C and
Standard methods for pollen research 71
lyophilization is recommended (Melo & Almeida-
Muradian, 2010, 2011). The following procedures are
based mainly in the on the work of Arruda et al.
(2013a) and Arruda et al. (2013b) and Souza
et al. (2018).
12.2.1.1. Simultaneous extraction of vitamins B1, B2,
vitamers from vitamin B6 and niacin. For the simul-
taneous extraction of B vitamins, it is used an acid
hydrolysis followed by enzymatic treatment with fungal
diastase like this:
In an Erlenmeyer flask, approximately 5 g of dried bee
pollen is mixed with 50ml of HCl 0.1mol/l and kept in
boiling water bath for 30min. After cooling to room tem-
perature, the solutions have their pH adjusted to 4.6 with
sodium acetate 2.5M. After adding 0.5 g of the diastase
(from fungi), the solutions are incubated for 2 h at 42 C
in a water bath. After the enzymatic treatment and cool-
ing at room temperature, the solutions are transferred to
100ml volumetric flasks; the volumes should be com-
pleted with deionized water, homogenized and filtered
through filter paper and in 0.45-mm membrane for injec-
tion into the chromatographic system.
12.2.1.2. Chromatographic conditions for the analysis
of the B complex vitamins in dried bee pollen. The
chromatographic conditions vary for each vitamin.
Vitamin B1 and B2 are analyzed under the same condi-
tions, while vitamin B6 and PP are quantified in distinct
conditions. Vitamins B2 and B6 are detected directly by
fluorescence, while vitamins B1 and PP underwent pre-
and post-column reaction, respectively, to be also
detected by fluorescence.
12.2.1.3. Determination of thiamin (vitamin B1), with
pre-column reaction. For the reaction of thiamin into
thiochrome, 1ml of standard solution is diluted from
the sample extract is pipetted into 10ml amber volu-
metric flasks, 2ml of deionized water and 3ml of freshly
prepared alkaline potassium ferricyanide solution,
[0.25% (w/v) in 15% NaOH]. The mixture is homogen-
ized and put to rest for 2min to allow the reaction.
After this period, 450ml of 85% orthophosphoric acid
is added. The solution is then cooled, and the final vol-
ume is completed with deionized water. These solutions
are filtered in a 0.45-mm membrane and injected into
the chromatographic system immediately after prepar-
ation. For the separation, 20ml is injected into a C18
reversed-phase column (RP-18 spherical 5 lmm/
250	 4.6mm) with pre-column (5 lm/10	 4.6mm).
The mobile phase is composed of phosphate buffer pH
7.2 (10mM KH2PO4.3H2O) and dimethylformamide
(85:15); flow rate: 1ml/min and detection by fluores-
cence (Ex 368 nm; Em 440 nm).
12.2.1.4. Determination of riboflavin (vitamin B2).
20ml is injected into C18 reversed-phase column (RP-18
spherical 5lm/250	 4.6mm ) with pre-column lm/
10	 4.6mm). The mobile phase is composed of phos-
phate buffer pH 7.2 (10mM KH2PO4.3H2O) and dime-
thylformamide (85:15); flow rate of 1ml/min and
detection by fluorescence (Ex 450 nm; Em 530 nm).
12.2.1.5. Determination of vitamin PP through niacin
and niacinamide vitamers. Vitamin PP is detected by
fluorescence after reaction under ultraviolet (UV). For
the separation, 20ml is injected into C18 reversed-phase
column (Luna C18, spherical 5 mm/250	 4.6mm) with
pre-column (5 mm/2	 8mm). The mobile phase is com-
posed of 0.07M potassium dihydrogen phosphate,
KH2PO4 solution (pH 4.5) of phosphate buffer of
75mM hydrogen peroxide and 0.1% of 5mM copper
sulfate solution; flow: 1.5ml/min; reactor composed of
12m of tetrafluoroethylene (TFE) tuve (0.5mm internal
diameter, wrapped around a black lamp) that emits radi-
ation from 300 to 400 nm; fluorescence detection (Ex
322 nm; Em 380 nm).
12.2.1.6. Determination of vitamin B6 through the
vitamers pyridoxol, pyridoxal and pyridoxamine. For
the separation, 20ml is injected into a C18 reversed-
phase column (LiChrospher1 100 RP-18 end-capped
5mm/250	 4.0mm), with pre-column (5 mm/4	 4mm
Lichrospher 100 RP-18). The mobile phase is composed
of phosphate buffer pH 2.5 (39mM KH2PO4) with ion
pair (0.63mM PIC 7) and acetonitrile (96:4); flow:
0.6ml/min; detection by fluorescence (Ex 296 nm;
Em 390 nm).
12.3. Methods recommended to analyze antioxidant
vitamins of bee pollen
The following procedures are based mainly in the on the
work of Melo and Almeida-Muradian (2010, 2011) and
Sattler et al. (2013).
12.3.1. Determination of vitamin C
The method is a titrimetric method, based on the
reduction of 2,6-dichlorophenol-indophenol (DCPIP) by
ascorbic acid (also recommended by AOAC, 1995).
12.3.2. Determination of vitamin E
This vitamin is analyzed using high-performance liquid
chromatography (HPLC) using silica column and fluores-
cence detector (Ex. 295; Em. 330); mobile phase: hex-
ane/isopropyl alcohol (99:1); flow rate: 1.5ml/min.
12.3.3. Determination of b-carotene (provitamin A)
This carotenoid, as well as other carotenoids can be
analyzed using both open column chromatography
(OCC) (Almeida-Muradian et al., 2005; Oliveira et al.,
2009) or HPLC, using C18 column (25 cm, 5 um), diode
72 M. G. Campos et al.
array detection (450 nm); mobile phase: methanol, ethyl
acetate, acetonitrile (70:20:10); flow rate: 1.2ml/min.
12.4. Concluding remarks
Note that the dehydration process, which is favorable
for the preservation of bee pollen and the increase of
shelf life of the product, do not interfere in the content
of complex B vitamins (p< 0.05). In general, it can be
stated that the concentration of vitamins is dependent
on storage time and not on the conditions in which the
dry pollen is stored (p< 0.05). When stored at room
temperature, vitamin C is rapidly lost, probably due to
oscillations in temperature, and, after this initial sharp
loss, it becomes quite stable during the storage process.
In relation to vitamin E (Melo & Almeida-Muradian,
2010), the storage conditions do not account for drastic
losses in its concentrations.
13. Standard methods for mineral and
trace elements
13.1. Introduction
Insufficient data are available on the different compos-
ition and variability of minerals and trace elements in
bee pollen between plant species. In each region, the
collected bee pollen generally came from just a few
spontaneous plant species and agricultural crops in each
region play also an important role as pollen sources.
Among the micronutrients vitamins, polyphenols and
minerals are the more important for human well-being.
Nevertheless, some minerals can be toxic if the intake
is higher or over a long period, (e.g., Zn, Se, Mn and
Mo). Others at excessive levels may exhibit high toxicity
(Villanueva & Marquina, 2001; Serra-Bonveh
ı & Escol
a
Jord
a, 1997; Yang et al, (2014). According FAO (2001)
the mineral elements can be classified as “nutritionally
essential major elements” (e.g., Ca, Cl, K, Mg, N, Na, P
and S), “nutritionally essential minor elements” (e.g., Fe,
B, Br, I and Si) and those regarded as “toxic” at exces-
sive levels and may exhibit toxicity (e.g., Zn, Se, Mn,
Mo). Given the importance and variability of different
minerals founded in bee pollen, Campos et al. (2008)
proposed the inclusion in label of bee pollen, for sale,
additional information related to different constituents
and also minerals, to complement the label and improve
the value of the product.
13.2. Materials and methods
Usually the determination of mineral and trace elements
on bee pollen is carried out on pollen ash, most fre-
quently by atomic absorption. The main mineral is K
(about 60% of total mineral content), Mg constitutes
about 20%, and Na and Ca 10% (SzczeRsna & Rybak-
Chmielewska, 1998).
13.2.1. Sample preparation
According to the analytical method used, different sam-
ple preparations methodologies can be performed.
Method 1: for Flame and Electrothermal Atomic
Absorption Spectrometry (Stanciu, Marghitas, Dezmirean
and Campos, 2012).
1. weigh 2 g of pollen;
2. ash in a muffle furnace at 450 C overnight;
3. digest by treating with hydrochloric acid 6 M (5ml)
and H2O2 (3ml) and then evaporate;
4. dissolve with 25ml nitric acid 0.1 M. A blank digest
was also carried out;
5. Lanthanum nitrate [La(NO3)3x6H2O] were used as
matrix modifier for the determination of potassium,
calcium and magnesium.
Method 2: for flame atomic absorption spectrometry and
analyzing honey and pollen (Grembecka & Szefer, 2013):
1. weigh 1 g (±0.0001 g);
2. treat with 9ml 65% HNO3 and digest in an automatic
microwave digestion system: 250 W, 48 s; 0 W, 48 s;
250W, 6min 24 s; 400 W, 4min; 650 W, 4min;
3. cycle consists of five food samples and one blank sam-
ple (9ml 65% HNO3);
4. every digested sample is dissolved in up to 10ml
deionized water.
Method 3 (Loper et al., 1980):
1. weigh 1g of pollen;
2. digest with alternate aliquots of concentrated nitric
acid and hydrogen peroxide;
3. boil and dry at 130 C until the residual ash will be
white or light gray. Three cycles were sufficient;
4. taken up residues in dilute HCl and determine directly
for Zn and Pb. For Mg, Ca and K analyses, samples
are diluted 50–100 times to provide concentrations in
the proper absorption range.
Method 4: for Inductively coupled argon plasma-atomic
emission spectrometry (ICP-AES): Paulo et al. (2012, 2014)
1. 200mg of dry bee pollen digest with a mixture of
nitric acid (10% v/v) and hydrogen peroxide (30% v/v)
at 100 C in a block digestion system (Digiprep MS)
until complete digestion (12 h).
2. final residue is dilute to 50ml with nitric acid (10% v/
v) and filter.
Method 5: Yang et al. (2014)
1. 1.0 g of digested bee pollen with 10ml of concentrate
nitric acid (65%) and 1ml of hydrogen peroxide (30%)
in a close polytetrafluoroethylene-stopper vessel for
8 h at 170 C.
Standard methods for pollen research 73
13.2.2. Methods of analysis
The methods that are usually used for mineral compos-
ition determination in food products were the Flame
and Electrothermal Atomic Absorption Spectrometry
and the inductively coupled plasma atomic emission
spectroscopy with or without mass detector. However,
some easier methods could be use if only the determin-
ation of calcium will be required. Kelina et al. (2013)
determine the concentration of calcium in bee pollen
samples from Lambayeque – Peru using complexometric
method. In this method 0.1 g of ash were used to add-
ing a few drops of hydrochloric acid, the results being
expressed as percentage on dry basis.
13.2.2.1. Flame and electrothermal atomic absorption
spectrometry. Flame and Electrothermal Atomic
Absorption Spectrometry is a very common method of
food mineral analyses applied mostly for the determin-
ation of alkali and alkaline metals.
Equipment
(1) Device equipped with deuterium lamp for back-
ground correction and hallow-cathode lamps for each
of the elements studied and air/acetylene flame (Stanciu
et al., 2011, 2012). (2) Flame atomic spectrometric
measurements (F-AES/AAS) using the optimal instru-
mental parameters for each element with the following
wavelengths: 766.5 nm (K); 422.7 nm (Ca); 285.2 nm
(Mg); 248.3 nm (Fe) and 213.9 nm (Zn). Good linearity
for the method is observed and the calibration curves
present a very good accuracy with an r2 ranged
between 0.9990 and 0.9999.
Grembecka and Szefer (2013) analyzed the concen-
tration of elements (Mg, Ca, K, Na, Zn, Cu, Fe, Cr, Co,
Ni, Mn, Pb and Cd) in pollen and honey in an air–acety-
lene flame by AAS method using deuterium-background
correction. These authors added Cs to samples, for Na
and K determinations and standards as an ionization
buffer at a concentration of 0.2%w/v. In the case of Ca
and Mg measurements, La was used as a releasing agent
at a concentration of 0.4%w/v.
13.2.2.2. Inductively coupled plasma atomic emission
spectroscopy (ICP-AES) methods. Method I (Paulo
et al., 2012, 2014):
1. All experiments were carried out using an inductively
coupled plasma-optical emission spectrometry (ICP-
OES- Activa M, Horiba Jobin Yvon), operating at 1000
W plasma power;
2. 15 l min1 plasma gas flow, 0.02 l min1 nebulizer Air
flow and 1.0 bar air pressure;
3. The analytical wavelengths (nm) were set of the fol-
lowing: Cd (228.802), Cr (205.571), Cu (327.395), Fe
(259.940), Mn (257.611), Pb (283.305) and
Zn (213.857);
4. Results expressed in mg.kg1 on a dry weight basis.
Method II (Yang et al., 2014):
1. Plasma gas flow rate (Ar), 16 l.min1;
2. auxiliary flux, 1.0 l.min1;
3. nebulized pressure, 25 psi; sample flush time, 30 s;
delay time, 30 s; solution uptake rate, 1.60ml.min1;
4. radio frequency, 27.12MHz; power, 1.05 kW;
5. absorbance data for each element recorded at the fol-
lowing wavelengths: Al (k ¼ 308.2 nm), As (k ¼
189.0 nm), Ca (k ¼ 317.9 nm), Cd (k ¼ 228.8 nm), Co
(k ¼ 228.6 nm), Cr (k ¼ 283.5 nm), Cu (k ¼
324.7 nm), Fe (k ¼ 259.9 nm), Ge (k ¼ 265.1 nm), Hg
(k ¼ 184.9 nm), K (k ¼ 766.4 nm), Mg (k ¼
279.5 nm), Mn (k ¼ 257.6 nm), Mo (k ¼ 202.0 nm),
Na (k ¼ 589.5 nm), Ni (k ¼ 231.6 nm), P (k
¼213.6 nm), Pb (k ¼ 220.3 nm), Se (k ¼ 196.0 nm)
and Zn (k ¼ 213.8 nm);
6. Results expressed in micrograms per gram on a dry
weight basis.
To be underlining that ICP-MS is a powerful, but
expensive technique for determination of trace mineral.
It is also used for other bee products too as honey, for
example Caroli et al. (1999), Madejczyk and
Baralkiewicz (2008) and Chudzinska and
Baralkiewicz (2011).
13.3. Concluding remarks
Some preliminary results from the composition of min-
erals in bee pollen can give a good idea of the amounts
and the identification of the main of these compounds
on this crude material. Still the information remains
scarce relatively to different floral sources and in special
correlating minerals and regions of the globe. The coun-
tries should do similar assays to collect the most data
possible to validate the amounts and the contaminants
as this content could be a bio-indicator of the environ-
mental condition or pollutants.
The amount of minerals in bee pollen from the vari-
ous data collected in the published papers can be found
in Table 49. Stanciu et al. (2012) analyzed pollen loads
samples obtained by pollen traps installed in Apis melli-
fera (L.) hives located in the northwest and central
counties of Transylvania area (Romania). These authors
confirmed that the examined bee pollen is a natural
source of nutritionally essential minerals and can con-
tribute for a better balanced diet or for special thera-
peutic applications. In their work, the authors reported
that the main macro elements were potassium
(2483–7620mg/kg), calcium (553–2798mg/kg) and mag-
nesium (3555–205mg/kg), while the main microele-
ments were iron (18.8–135mg/kg) and zinc
18.8–60.5mg.kg1. Variations in the mineral levels of
the analyzed monofloral bee pollens were due to differ-
ences in the floral origin.
Loper et al. (1980) use similar methodology, with
flame absorption spectrometry for mineral
74 M. G. Campos et al.
determination (e.g., Mg, Ca, K, Zn and Pb) in fresh pol-
len collected by hand from flowers, bee pollen removed
from the bees’ legs and collected from the pollen traps
from California. Potassium is the main compound in bee
pollen. Highest values of potassium were reported by
Stanciu et al. (2011) in Salix sp. pollen (8686mg/kg), in
Helianthus annuus L. pollen (5421mg/kg) and by Stanciu
et al. (2012) in Knautia arvensis (L.) (7619.57mg/kg), in
Onobrychis viciifolia Scop. (6748.72mg/kg), Crataegus
monogyna Jacq. (6224.17mg/kg) and Inula helenium L.
(6078.28mg/kg), Taraxacum officinale Web. (2483.51mg/
kg), Carduus sp. (3865.27mg/kg) and among others pol-
len samples analyzed.
Different authors reported different values for pollen
potassium concentration, namely: 4000mg/kg (Villanueva
& Marquina, 2001), 5530mg/kg (Somerville & Nicol,
2002), 2843–5976mg/kg (SzczeRsna, 2007) and
4950–5131mg/kg (Salamanca et al., 2008) (Table 49).
These results confirm the higher variability of pollen
mineral composition related to the different origin of
pollen. Calcium followed by magnesium were the next
highest mineral content in pollen. All authors founded
significant differences related to the higher concentra-
tion of potassium compared with the other minerals.
Yang et al. (2014) analyzed twelve common varieties of
monofloral bee pollen collected from China’s main pro-
ducing regions. On average the values obtained were
the following: P 5946mg/kg, K 5324mg/kg, Ca 2068mg/
kg, Mg 1449mg/kg, sodium 483mg/kg, aluminum
129mg/kg, iron 119mg/kg, magnesium 70mg/kg, zinc
45mg/kg and cupper 17mg/kg.
Paulo et al. (2012) analyzing Cistus ladanifer L., Rubus
ulmifolius Schott and Calluna vulgaris (L.) found different
composition of micronutrients (Cu, Fe, Mn and Zn) in
pollen related to different botanical origins but not
from the same species in different region. These
authors concluded that the determination of micronu-
trients could be suitable for the identification of botan-
ical species. The contents of Cd, Pb, As and Hg were
usually lower or not detected in pollen samples. In
Campos et al. (2008), the limits for this trace elements
were Cd  0.1mg/kg, Pb  0.5mg/kg, As  0.5mg/kg
and Hg  0.03mg/kg.
14. Standard methods for pyrrolizidine alkaloid
profiling by UHPLC-HRMS
14.1. Introduction
Pyrrolizidine alkaloids (PAs) are secondary metabolites
produced by some plants as defense mechanism against
herbivores and phytophagous insects. They can occur as
N-oxides or free-bases/tertiary PAs. In plants, the N-
oxides are often found in higher concentrations as com-
pared to the corresponding free-bases/tertiary PAs
(Hartmann & Toppel, 1987; Hartmann & Witte, 1995).
PA containing plant species mainly belong to the families
of Boraginaceae (all genera), Asteraceae (mainly genera
of Senecioneae and Eupatorieae), Fabaceae (mainly
genus of Crotalaria) and Apocynaceae (Hartmann &
Witte, 1995). PAs can get into bee products, such as
honey and pollen, when bees collect nectar or pollen of
PA containing plants. PAs have been detected in plant
pollen (Boppr
e et al., 2005; Kempf et al., 2010). High
concentrations of PAs have also previously been
reported in bee collected pollen (D€ubecke et al., 2011;
Kempf et al., 2010) presenting a food safety concern for
the consumers of pollen as a nutritional supplement.
PAs are classified as esters of hydroxylated methyl pyr-
rolizidines, consisting of a necine base (1,2-saturated or
unsaturated) and one or more necinic acids (Hartmann
& Witte, 1995) (Figure 47a,b). Esters of 1,2 unsaturated
retronecine- and otonecine type PAs are toxic for
humans and animals (Prakash et al., 1999; R€osemann
et al., 2014). The structural diversity of more than 400
known PAs (Boppr
e, 2011; Kempf et al., 2011) repre-
sents an analytical challenge and no standardized
method has been established so far to determine the
PA content in bee products. Many PAs are difficult to
quantify, due to the lack of standard materials.
Two main analytical approaches are commonly used:
a sum parameter GC-MS method (Kempf et al., 2008,
2010, 2011) and a targeted LC-MS method allowing the
individual identification of PAs and PA-N-oxides
(Betteridge et al., 2005; Boppre et al., 2005; Kempf
et al., 2011). The GC-MS method covers most PAs,
except the otonecine-type PAs, but derivatization is
required and structural information of the original PAs
is lost. On the other hand, the targeted LC-MS method
identifies individual PAs, but unknown PAs are not
detected and the chromatographic separation of many
PA isomers is also not easily obtained. To at least par-
tially overcome these limitations, ultra-high pressure
liquid chromatography coupled to high resolution mass
spectrometry (UHPLC-HRMS) as a way to profile both
known and unknown PAs can be used (Avula et al.,
2015). Compared to conventional HPLC, UHPLC ena-
bles a more rapid separation with higher resolution and
is more efficient at separating closely related molecules,
e.g., positional isomers (Plumb et al., 2004). UHPLC
relies on the use of smaller particles (sub-2mm) and
instrumentations able to resist higher pressure, typically
up to 1000 or 1500 bars. HRMS systems, such as time-
of-flights (TOF) or electrostatic trap (OrbitrapTM) rep-
resent an attractive way to record data from UHPLC
separation at high frequency and in a non-targeted man-
ner. HRMS measures ions with high mass and spectral
accuracies and allows for the determination of elemen-
tal compositions, which in turn may assist the identifica-
tion of unknowns. Furthermore, recent instruments are
fully compatible with quantitative or semi-quantitative
analysis (Rochat, 2012). Here we describe a UHPLC-
HRMS-based method for the profiling of PAs in plant
pollen from Echium and Eupatorium species, which may
be applied to other plant genera with minor adaptation.
Standard methods for pollen research 75
14.2. Materials and methods
14.2.1. Equipment and chemicals
a. Steel-made forceps from A. Dumont (#0208-55-PO)
b. 1.5ml and 2ml tubes from Eppendorf
(#0030120086, #0030120094)
c. 2ml glass LC-MS tubes (screw top) from
Kinesis (#KVP6112)
d. polystyrene Petri dishes (Ø 10 cm) from
Greiner (#663102)
e. glass beads (Ø 2mm) from Sigma-Aldrich
f. nitrile gloves (powder free) from Kimberly-
Clark (#52003M)
g. Solvents: cyclohexane SupraSolv (#102817), methanol
SupraSolv (#106011) and formic acid for Analysis
Emsure (#100264) from Merck, Acetonitrile ULC/MS
(#012041) and Water ULC/MS (#232141)
from Biosolve
h. Standards: echimidine (CAS #520-68-3) and lycops-
amine (CAS #10285-07-1) from Phytolab
(#89553, #89726).
14.2.2. Collection of plant pollen from PA-containing plants
Two examples: collection of pollen from E. vulgare
(Boraginaceae) and from E. cannabinum (Asteraceae).
14.2.2.1. Pollen collection from E. vulgare. Plant pollen
from E. vulgare, located as blue/violet dust on the anthers
of the flower can be collected using steel-made surgery
forceps or using pure cyclohexane (Figure 48a,b and
Figure 49). Rainy days are not recommended for collec-
tion, since pollen can be washed out from the anthers.
1. Verify that most of the flowers are well developed
and fully open.
2. Tightly bag the plants on the field with fine mesh nets
to avoid the pollen harvest from insects (at least one
day before the collection).
3. Spread a layer of paraffin around the lower part of
the stem to avoid that non-flying insects are climbing
the plant to collect pollen.
4. Plant pollen can be collected two days after bagging
the plant and is stored at 20 C until extraction. It is
suggested to bring a cooler box container filled with
dry ice on the field in order to place the freshly col-
lected samples immediately on dry ice.
14.2.2.1.1. Collection with forceps.
1. Wear nitrile gloves during the collection.
2. Using forceps, scratch the pollen from the surface of
the anthers of the flowers, carefully avoiding PA con-
tamination by the stamen tissues.
3. Collect the pollen using pre-weighed 1.5ml tubes.
Collect 1mg of pollen in each tube.
4. Insert the forceps carefully inside the tube and release
the pollen on the walls of the vial. Try to avoid any
electrostatic charge of the tube that could comprom-
ise the collection.
14.2.2.1.2. Collection with cyclohexane.
1. Remove the stamens from the flower.
2. Immerge the stamens in a pre-weighed LC-MS glass
tube containing cyclohexane.
3. Shake the stamens delicately into the cyclohexane to
release the pollen.
4. If possible, bring the tubes back to the laboratory to
evaporate the cyclohexane under laminar-flow hood.
Determine the weight of the dry pollen.
5. Store the dry pollen at 80 C until extraction.
Pros: This technique allows a faster collection of pollen
when compared to the method by forceps. Both collec-
tion methods give similar concentrations of PAs as we
have determined in our laboratory.
14.2.3. Pollen collection from E. cannabinum
Plant pollen from E. cannabinum (Figures 50 and 51),
located along the stamens, can’t be collected using for-
ceps due to the small dimensions of the tubular floral
units (approximately 7mm).
Figure 48. (a and b) Structure of a flower of E. vulgare.
76 M. G. Campos et al.
Collection on Petri dishes
1. Bag the plant and apply the paraffin layer on the lower
part of E. cannabinum.
2. After two days, shake the entire floral head of E. can-
nabinum over a Petri dish in order to collect the pol-
len that is released by this procedure.
3. Remove the impurities from the dish using forceps.
4. Collect the pollen into a pre-weighted 1.5ml tube.
5. Store the tube at 80 C until extraction.
Pros: This method allows to collect rapidly a high
amount of pollen from E. cannabinum.
14.3. Extraction of PAs from plant pollen
PAs are polar organic compounds with basic character-
istics. They are soluble in polar organic solvents or in
mixtures of solvents and acidified water.
1. Transfer 1mg of plant pollen in a pre-weighed 2ml
microcentrifuge tube using a metal spatula.
2. Dissolve the pollen in 100 ml of extraction solvent
consisting of 70% methanol, 29.5% ultra-pure water
and 0.5% formic acid.
3. Add 4–8 glass beads to the tube.
4. Shake the tube at 30.0Hz for 4minutes.
5. Centrifuge the tube at 18,407 g for 4minutes.
6. Collect the supernatant (extract) in a new tube.
7. Transfer with a pipette 5 ml of the extract and dilute
5–20 times with the extraction solvent into a glass
LC-MS vial containing a conical glass insert.
Pros: This technique can be used for extracting
plant pollen from many plant genera.
14.4. Profiling of PAs in plant pollen with
UHPLC-HRMS
Non-targeted analysis using LC-HR-MS system allows
the detection of alkaloids found in Echium vulgare (echi-
midine/-N-oxide, vulgarine/-N-oxide, acetyl-echimidine/-
N-oxide, acetylvulgarine/-N-oxide and echivulgarine/-N-
oxide) and alkaloids found in Eupatorium cannabinum
(intermedine/-N-oxide, lycopsamine/-N-oxide).
Separation of the alkaloids is performed using an
Acquity UPLC (Waters). The UPLC system is coupled
to a Synapt G2 QTOF mass spectrometer (Waters). An
Acquity BEH C18 column (50	 2.1mm i.d., 1.7 mm par-
ticle size, Waters) fitted with guard column
(5	 2.1mm) of identical phase is utilized. The column is
maintained at 30 C and a binary gradient of separation
Figure 49. An E. vulgare plant bagged with a net and supported
by a metal structure.
Figure 50. Structure of a floret and a floral head of
E. cannabinum.
Standard methods for pollen research 77
is performed at flow rate of 0.4ml min1. The mobile
phase consists of 0.05% formic acid in water (solvent A)
and 0.05% formic acid in acetonitrile (solvent B). The
gradient program used is: 0–4min 5–40% B, 4–6min
40–100% B, 6–9min 100% B, 9.1–10.5min 5% B.
Injection volume is 1ml. The autosampler needle is
washed with methanol/acetonitrile/isopropanol (1:1:1)
followed by 0.05% formic acid in water between injec-
tions to eliminate carryover. The QTOF operates in
electrospray positive mode over a mass range of
50–600Da. MS conditions are: Capillary voltage
þ2800V, cone voltage þ30 V, source temperature
120 C, desolvation gas temperature and flow 350 C
and 800 l/h, respectively, scan time 0.4 s. A leucine-enke-
phaline solution at 400 ng/ml is infused throughout the
analysis to ensure high mass accuracy (<2 ppm).
Fragmentation spectra are recorded in separate analyses
in MSe mode using a collision energy ramp of 10–30 eV.
Data are recorded using Masslynx 4.1 and PAs are iden-
tified on the basis of their retention times, exact mass
fragmentation characteristics and comparison with the
existing literature and databases containing information
on PAs known in Echium and Eupatorium genus. External
calibration for the quantification of the PAs is made
using echimidine, intermedine and lycopsamine as stand-
ards. Five calibration points are made: 0.02, 0.1, 0.5, 2,
and 10 mg/ml. Linear responses are obtained from 0.02
to 2mg/ml (Table 48). The limit of detection (LOD) and
limit of quantitation (LOQ) for echimidine, lycopsamine
and intermedine are 1.5 ng/ml and 3 ng/ml, respectively.
As only a limited number of reference standards are
available, a number of other PAs and PA-N-oxides com-
monly found in Echium (echimidine-N-oxide, vulgarine/-
N-oxide, acetyl-echimidine/-N-oxide, acetylvulgarine/-N-
oxide, echivulgarine/-N-oxide) are indirectly quantified
as echimidine equivalents. PAs and their PA-N-oxides
and isomers in pollen of E. vulgare and E. cannabinum
determined with LC-MS analysis are reported in
Table 49. Two examples of extracted ion chromato-
grams of PAs are shown in Figure 52.
Pollen samples of E. vulgare and E. cannabinum ana-
lyzed contain mainly PA-N-oxides while free bases were
only present as traces. Table 50 shows the MS/MS frag-
ments of PA-N-oxides detected in pollen samples. In
addition, PAs not previously reported to be present in
pollen of E. vulgare and E. cannabinum can also be
detected by UHPLC-HRMS. They are tentatively identi-
fied through the determination of their molecular for-
mula based on high mass and spectral accuracy
measurements by the high resolution QTOF mass spec-
trometer (Table 51).
14.5. Data processing
The following procedure is used for the detection and
quantification of PAs. Provided that the identity of the
PA is known, this procedure may also be applied to
other types of alkaloids from other plant species.
1. To identify the different alkaloids, generate an
extracted ion chromatogram (EIC) using a mass win-
dow of ± 0.01Da around the values provided in
Table 49.
2. Verify the mass accuracy of the ions compared to the
theoretical values.
3. In the EICs integrate the peaks and calculate concen-
trations in ng/g or mg/g of pollen using your ownFigure 51. An E. cannabinum plant bagged with a net.
Table 51. PAs which have not been previously reported from E. vulgare and E. cannabinum.
Compound name
RT
(min) Chemical formula [MþH]þ experimental [MþH]þ calculated Error (mDa)
 Echium vulgare
Curassavine-N-oxide
 1.54 C16H29NO5 316.2125 316.2124 0.1
 Eupatorium cannabinum
Leptantine
 1.41 C15H27NO6 318.1922 318.1917 0.5
Amabiline
 1.48 C15H25NO4 284.1865 284.1862 0.3
Uplandicine
 1.71 C17H27NO7 358.1868 358.1866 0.2

Tentative identification.
78 M. G. Campos et al.
calibration curves. Examples of equations are provided
in Table 48.
In this study we obtained a total PA concentration of
7.43 (± 2.45) mg/g in pollen of E. vulgare and 5.32 (±
1.98) mg/g in pollen of E. cannabinum. Pollen samples
were collected in Switzerland in 2013.
14.6. Concluding remarks
Liquid chromatography approaches are frequently
chosen for the identification and quantification of PAs in
plants and for food analysis, allowing the determination
of individual PAs and PA-N-oxides. This is particularly
important for determining the exact type of PAs which
are present in a given plant species and also in food
analysis, where the spectrum of PAs may help deduce
the plant species that contribute to PA contamination.
Additionally, the sample preparation is faster and does
not need derivatization as required for gas chromatog-
raphy analysis. However, many types of PAs are still not
available on the market as reference substances.
Therefore, a number of PAs are quantified using a refer-
ence substance that is closely related but not identical
to the analyzed PAs. Hence, quantification may not be
entirely accurate, since the response factor of the
Table 53. Statistics of the 134 sample of bee pollen analyzed.
Parameter Mean ± r Minmax Coefficient of variation
Total phenolic compounds 23.9 ± 4.16 16.1–35.5 17.4
Total flavonoids 4.1 ± 0.96 2.2–5.9 23.2
EC50 3.5 ± 1.05 2.0–6.4 39.1
ABTS 244.9 ± 24.40 203.8–298.6 10.0
Table 54. Results for calibration and cross-validation.
Calibration
Parameter
Spectral range
(cm1) Pre-processing Rk r2 RMSEE RPD
Total phenolic compounds 2787 2459þ 2135 1809 2ndDer 8 92.6 1.16 3.7
Total flavonoids 2858 2509þ 2164 1470þ 777 430 VecNor 9 85.9 0.373 2.7
EC50 3055 2731þ 2087 1119þ 799 474 VecNor 10 89.7 0.351 3.1
ABTS 2557 1838þ 1479 399 1stDer 10 91.3 7.57 3.4
Cross-validation
Parameter Spectral range(cm21) Pre-processing Bias r2 RMSECV RPD
Total phenolic compounds 2787 2459þ 2135 1809 2ndDer 0.0044 79.3 1.88 2.2
Total flavonoids 2858 2509þ 2164 1470þ 777 430 VecNor 0.0020 76.0 0.469 2.0
EC50 3055 2731þ 2087 1119þ 799 474 VecNor 0.0185 77.9 0.492 2.1
ABTS 2557 1838þ 1479 399 1stDer 0.2470 72.5 12.7 1.9
2ndDer – seconded derivative; VecNor – vector normalization; 1stDer – first derivative; r2 – coefficient of determination; Rk – number of PLS
components; RMSEE – root mean square error of calibration; RPD – residual prediction deviation; Bias – mean value of deviation, also called sys-
tematic error; RMSECV – root mean square error of cross-validation.
Table 52. Frequency of each plant genus as predominant pollen, secondary pollen, important minor pollen and minor pollen in the
134 pollen samples mixture analyzed.
Genus
Predominant pollen
(>45%)
Secondary pollen
(15–45%)
Important minor pollen
(3–15%) Minor pollen (<3%)
Eucalyptus spp. 13 21 21 21
Prunus spp. 8 4 4 4
Rubus spp. 8 11 8 8
Castanea spp. 15 34 23 8
Trifolium spp. 0 6 11 19
Cistus spp. 53 17 6 4
Echium spp. 21 17 11 4
Kiwi spp. 2 0 4 11
Leontondon spp. 0 2 17 11
Quercus spp. 0 6 11 15
Erica spp. 2 4 6 2
Thymus spp. 2 0 11 2
Genista spp. 0 6 6 2
Lavandula spp. 0 4 13 0
Cytisus spp. 11 23 11 2
Standard methods for pollen research 79
detector may differ between PAs, even when they are
closely related to each other. Furthermore, positional
isomers, such as lycopsamine and intermedine, are also
not easy to separate. Finally, while some plant species
have been relatively well characterized regarding their
PA content (e.g., E. vulgare), little is known for many
other PA-producing plant species. Recently, ultra-high
performance liquid chromatography coupled with high
resolution mass spectrometry has been described as a
selective and sensitive method for the analysis of PAs in
plants (Avula et al., 2015; Crews et al., 2009, 2010;
Xiong et al., 2009). This method provides an accurate
mass information based on which unknown PAs may
also be detected in an untargeted manner without the
need for reference standards for each PA, as has been
shown in the present study. In the future we anticipate
that UHPLC-HRMS will become the method of choice
for the profiling of both known and unknown PAs
in plants.
15. Standard methods for total phenolic
compounds, total flavonoids and antioxidative
capacity bee pollen evaluation by FTIR-ATR
spectroscopy
15.1. Introduction
FTIR-ATR techniques have been applied in several dif-
ferent crud materials in order to simplify the analysis
and have results quickly. Gok et al. (2015) used FTIR-
ATR spectroscopy in automated and highly sensitive
botanical origin estimation of honey. According Anjos
et al. (2017) FTIR-ATR spectroscopy technique in com-
bination with partial least squares regression (PLS-R)
can be used to evaluate the ability to quantify total
phenolic compounds, total flavonoids, free radical scav-
enging (EC50) and antioxidant capacity (ABTS) in bee
pollen. The method described in this chapter is based in
this previously paper was PLS-Regression analysis was
used to predicted for total phenolic compounds, total
flavonoids, EC50 and ABTS. The calibrations model
founded present a good accuracy for total phenolic
compounds (r2 ¼ 92.6%; RPD ¼ 3.7), total flavonoids
(r2 ¼ 85.9%; RPD ¼ 2.7); EC50 (r2 ¼ 89.7%; RPD ¼
3.1) and ABTS (r2 ¼ 91.3%; RPD ¼ 3.4) of bee pollen).
Figure 52. Extracted ion chromatograms of PAs identified in plant pollen of E. vulgare (A) and E. cannabinum (B). Peak numbers cor-
respond to the PAs described in Table 51 (1: echimidine-N-oxide; 2: vulgarine-N-oxide; 3: acetylechimidine-N-oxide; 4: acetylvulgar-
ine-N-oxide; 5: echivulgarine-N-oxide; 6: lycopsamine; 7: intermedine; 8-9: lycopsamine-N-oxide or intermedine-N-oxide).
Figure 53. (a–d) Scheme of sample preparation for analysis in
FTIR-ATR equipment.
80 M. G. Campos et al.
T
ab
le
55
.
C
on
di
tio
n
va
ri
at
io
ns
of
th
e
D
PP
H

 a
ss
ay
on
be
e
po
lle
n
ac
co
rd
in
g
to
di
ffe
re
nt
au
th
or
s.
N
um
be
r
of re
po
rt
R
ef
er
en
ce
In
iti
al
D
PP
H

 c
on
ce
nt
ra
tio
n
R
ea
ct
io
n
tim
e
Ex
pr
es
si
on
of
ac
tiv
ity
D
PP
H

 v
ol
um
e/
ex
tr
ac
t
vo
lu
m
e
1
C
am
po
s
et
al
.(
19
97
)
5.
91
m
g/
25
0
m
li
n
et
ha
no
l
10
m
in
EC
50
2.
5
m
l-D
PP
H

 s
ol
ut
io
n
þ
1
to
10
ml
ex
tr
ac
t
2
K
ro
ye
r
an
d
H
eg
ed
us
(2
00
1)
6
	
10

5
M
in
m
et
ha
no
l
60
m
in
%
in
hi
bi
tio
n
2.
7
m
l-D
PP
H

 s
ol
ut
io
n
þ
0.
3
m
le
xt
ra
ct
3
N
ag
ai
et
al
.(
20
02
)
1
m
M
N
I
%
ac
tiv
ity
0.
3
m
l-D
PP
H

 s
ol
ut
io
n
þ
2.
4
m
le
th
an
ol
þ
0.
3
m
le
xt
ra
ct
4
C
am
po
s
et
al
.(
20
03
)
6
	
10

5
M
in
m
et
ha
no
l
10
m
in
EC
50
2
m
l-D
PP
H

 s
ol
ut
io
n
þ
ap
pr
op
ri
at
e
am
ou
nt
of
ex
tr
ac
t
5
A
lm
ar
az
-A
ba
rc
a
et
al
.(
20
04
)
3.
4
mg
/m
li
n
50
%
et
an
ol
(v
/v
)
10
m
in
EC
50
16
00
to
19
90
ml
-D
PP
H

 s
ol
ut
io
n
þ
10
to
40
0
ml
ex
tr
ac
t.
Fi
na
lv
ol
um
e:
2
m
l
6
N
ag
ai
et
al
.(
20
05
)
1
m
M
N
I
%
ac
tiv
ity
0.
3
m
l-D
PP
H

 s
ol
ut
io
n
þ
2.
4
m
le
ta
no
lþ
0.
3
m
le
xt
ra
ct
7
Si
lv
a
et
al
.(
20
06
)
6
	
10

5
M
in
et
ha
no
l
30
m
in
EC
50
2
m
l-D
PP
H

 s
ol
ut
io
n
þ
ap
pr
op
ri
at
e
am
ou
nt
of
ex
tr
ac
t
8
Le
ja
et
al
.(
20
07
)
0.
1
m
M
in
et
ha
no
l
30
m
in
%
in
hi
bi
tio
n
0.
08
3%
po
lle
n
in
th
e
re
ac
tio
n
m
ix
tu
re
9
Ba
ltr
u s
ai
ty
t_ e
et
al
.(
20
07
)
6.
5
	
10

5
M
in
m
et
ha
no
l
16
m
in
%
in
hi
bi
tio
n
2
m
l-D
PP
H

 s
ol
ut
io
n
þ
50
ml
ex
tr
ac
t
10
C
ar
pe
s
et
al
.(
20
08
)
0.
5
m
M
in
et
ha
no
l
10
0
m
in
%
in
hi
bi
tio
n,
EC
50
0.
3
m
l-D
PP
H

 s
ol
ut
io
n
þ
3
m
le
th
an
ol
þ
0.
5
m
le
xt
ra
ct
11
Si
lv
a
et
al
.(
20
09
)
10
0
mM
in
et
ha
no
l
30
m
in
EC
50
2
m
l-D
PP
H

 s
ol
ut
io
n
þ
ap
pr
op
ri
at
e
am
ou
nt
of
ex
tr
ac
t
12
Le
Bl
an
c
et
al
.(
20
09
)
51
m
g/
10
0
m
li
n
m
et
ha
no
l
60
m
in
%
ac
tiv
ity
,E
C
50
24
2.
5
ml
D
PP
H

 s
ol
ut
io
n
þ
28
0
ml
et
ha
no
lþ
72
ml
ex
tr
ac
t
13
Le
e
et
al
.(
20
09
)
10
m
M
10
m
in
%
sc
av
en
gi
ng
ef
fe
ct
1
m
l-D
PP
H

 s
ol
ut
io
n
þ
4
m
le
xt
ra
ct
14
M
 ar
gh
ita
s
et
al
.(
20
09
)
0.
02
m
g/
m
l
15
m
in
%
in
hi
bi
tio
n
20
0
ml
-D
PP
H

 s
ol
ut
io
n
þ
40
ml
ex
tr
ac
t
15
W
an
g
et
al
.(
20
10
)
0.
5
m
M
in
et
ha
no
l
30
m
in
%
ac
tiv
ity
0.
5
m
l-D
PP
H

 s
ol
ut
io
n
þ
0.
5
m
la
ce
ta
te
bu
ffe
r
(1
00
m
M
,
pH
5.
5)
þ
0.
5
m
le
xt
ra
ct
16
M
en
ez
es
et
al
.(
20
10
)
0.
5
m
M
in
et
ha
no
l
30
m
in
%
ac
tiv
ity
4
m
l-D
PP
H

 s
ol
ut
io
n
þ
10
0
ml
ex
tr
ac
t
17
M
or
ai
s
et
al
.(
20
11
)
6
	
10

5
M
in
m
et
ha
no
l
60
m
in
%
-D
PP
H

 d
is
co
lo
ra
tio
n,
EC
50
2.
7
m
l-D
PP
H

 s
ol
ut
io
n
þ
30
0
ml
ex
tr
ac
t
18
N
eg
ri
et
al
.(
20
11
)
20
mg
/m
li
n
m
et
ha
no
l
20
m
in
%
ac
tiv
ity
3.
9
m
l-D
PP
H

 s
ol
ut
io
n
þ
0.
1
m
le
xt
ra
ct
19
G
ra
ik
ou
et
al
.(
20
11
)
70
mM
in
m
et
ha
no
l
U
nt
il
a
pl
at
ea
u
w
as
re
ac
he
d
%
ac
tiv
ity
,E
C
50
1
m
l-D
PP
H

 s
ol
ut
io
n
þ
50
ml
ex
tr
ac
t
20
Fe

 as
et
al
.(
20
12
)
6
	
10

5
M
in
m
et
ha
no
l
60
m
in
%
ab
so
rb
an
ce
,E
C
50
2.
7
m
l-D
PP
H

 s
ol
ut
io
n
þ
0.
3
m
le
xt
ra
ct
21
C
ha
nt
ar
ud
ee
et
al
.(
20
12
)
0.
15
m
M
in
m
et
ha
no
l
60
m
in
%
ac
tiv
ity
,E
C
50
50
ml
-D
PP
H

 s
ol
ut
io
n
þ
50
ml
ex
tr
ac
t
22
M
ej

ıa
s
an
d
M
on
te
ne
gr
o
(2
01
2)
0.
02
m
g/
m
L
in
m
et
ha
no
l
N
I
mg
as
co
rb
ic
ac
id
eq
ui
va
le
nt
s/
g
sa
m
pl
e
1.
5
m
l-D
PP
H

 s
ol
ut
io
n
þ
75
0
ml
ex
tr
ac
t
23
M
el
o
et
al
.(
20
12
a)
N
I
N
I
%
ac
tiv
ity
N
I
24
M
el
o
et
al
.(
20
12
b)
N
I
N
I
%
ac
tiv
ity
N
I
25
M
el
o
et
al
.(
20
12
c)
N
I
30
m
in
%
in
hi
bi
tio
n
N
I
26
M
el
o
et
al
.(
20
12
d)
N
I
30
m
in
%
in
hi
bi
tio
n
N
I
29
A
rr
ud
a
(2
01
3)
6
	
10

5
M
in
et
ha
no
l
30
m
in
EC
50
1.
5
m
l-D
PP
H

 s
ol
ut
io
n
þ
2.
0
m
le
ta
no
lþ
0.
5
m
le
xt
ra
ct
A
rr
ud
a
et
al
.(
20
13
)
N
I
N
I
EC
50
N
I
27
M
on
te
ne
gr
o
et
al
.(
20
13
)
0.
02
m
g/
m
l
15
m
in
mg
as
co
rb
ic
ac
id
eq
ui
va
le
nt
s/
g
po
lle
n
1.
5
m
l-D
PP
H

 s
ol
ut
io
n
þ
0.
75
m
le
xt
ra
ct
28
C
he
ng
et
al
.(
20
13
)
0.
1
m
M
in
m
et
ha
no
l
30
m
in
%
in
hi
bi
tio
n
4
m
l-D
PP
H

 s
ol
ut
io
n
þ
va
ri
ab
le
vo
lu
m
es
of
ex
tr
ac
t.
Fi
na
lv
ol
um
e:
10
m
l
30
U
lu
so
y
an
d
K
ol
ay
li
(2
01
4)
0.
1
m
M
in
m
et
ha
no
l
50
m
in
SC
50
1.
5
m
l-D
PP
H

 s
ol
ut
io
n
þ
1.
5
m
le
xt
ra
ct
31
M
el
o
et
al
.(
20
13
)
N
I
30
m
in
IC
50
N
I
32
Ba
rr
ia
da
-B
er
na
le
t
al
.(
20
14
)
40
mg
/m
li
n
et
ha
no
l
30
m
in
EC
50
60
0
to
99
0
ml
-D
PP
H

 s
ol
ut
io
n
þ
10
to
40
0
ml
ex
tr
ac
t.
Fi
na
lv
ol
um
e:
1
m
N
I¼
N
ot
in
di
ca
te
d.
Standard methods for pollen research 81
T
ab
le
56
.
A
dv
an
ta
ge
s
an
d
di
sa
dv
an
ta
ge
s
of
so
m
e
te
ch
ni
qu
es
an
d
m
et
ho
ds
us
ed
fo
r
an
ti-
m
ic
ro
bi
al
an
al
ys
is
.
M
et
ho
ds
A
dv
an
ta
ge
s
D
is
ad
va
nt
ag
es
R
ef
er
en
ce
s
A
ga
r
D
iff
us
io
n

T
he
te
ch
ni
qu
e
w
or
ks
w
el
lw
ith
de
fin
ed
in
hi
bi
to
rs
;
Fl
ex
ib
ili
ty
in
se
le
ct
io
n
of
di
sk
s
fo
r
te
st
in
g;
It
s
si
m
pl
ic
ity
of
im
pl
em
en
ta
tio
n
an
d
it
is
th
e
le
as
t
co
st
ly
of
al
ls
us
ce
pt
ib
ili
ty
m
et
ho
ds
;
Fo
r
te
st
in
g
of
an
tif
un
ga
ls
us
ce
pt
ib
ili
ty
its
re
la
tiv
e
ea
se
;
D
oe
s
no
t
re
qu
ir
e
fo
r
sp
ec
ia
liz
ed
eq
ui
pm
en
t;
H
as
in
tr
in
si
c
fle
xi
bi
lit
y
by
be
in
g
ab
le
to
te
st
th
e
dr
ug
s
th
e
la
bo
ra
to
ry
ch
oo
se
s;
T
he
pr
ov
is
io
n
of
ca
te
go
ri
ca
lr
es
ul
ts
ea
si
ly
in
te
rp
re
te
d
by
al
lc
lin
ic
ia
ns
;
T
he
di
sk
te
st
ha
s
be
en
st
an
da
rd
iz
ed
fo
r
te
st
in
g
no
t
al
lf
as
tid
io
us
or
sl
ow
gr
ow
in
g
ba
ct
er
ia
bu
t,
fo
r
so
m
e
as
st
re
pt
oc
oc
ci
,
H
ae
m
op
hi
lu
s
in
flu
en
za
e,
an
d
N
.m
en
in
gi
tid
es
.

Le
ss
se
ns
iti
ve
th
an
m
ic
ro
-d
ilu
tio
n
te
ch
ni
qu
es
;
R
eq
ui
re
s
m
uc
h
qu
an
tit
y
of
sa
m
pl
e
fo
r
an
al
ys
is
;
T
he
an
tim
ic
ro
bi
al
ef
fe
ct
m
ay
be
in
hi
bi
te
d
or
in
cr
ea
se
d
by
ex
tr
in
si
c
fa
ct
or
s
or
co
nt
am
in
an
ts
as
th
e
ty
pe
of
ag
ar
an
d
hi
s
co
nc
en
tr
at
io
n,
sa
lt
co
nc
en
tr
at
io
n,
in
cu
ba
tio
n
te
m
pe
ra
tu
re
,a
nd
m
ol
ec
ul
ar
si
ze
of
th
e
an
tim
ic
ro
bi
al
co
m
po
ne
nt
;
T
he
re
ar
e
su
bs
ta
nc
es
in
th
e
cu
ltu
re
m
ed
iu
m
th
at
in
flu
en
ce
th
e
te
st
su
sc
ep
tib
ili
ty
as
th
ym
id
in
e
th
at
an
ta
go
ni
ze
s
th
e
ac
tiv
ity
of
su
lfo
na
m
id
es
an
d
tr
im
et
ro
pr
in
as
,l
ea
di
ng
to
a
fa
ls
e
re
su
lt
of
re
si
st
an
ce
to
su
ch
ag
en
ts
;
T
hi
s
te
ch
ni
qu
e
al
so
do
es
no
t
di
st
in
gu
is
h
be
tw
ee
n
ba
ct
er
ic
id
al
an
d
ba
ct
er
io
st
at
ic
ef
fe
ct
s;
T
he
pH
of
th
e
cu
ltu
re
m
ed
iu
m
sh
ou
ld
be
co
m
pa
tib
le
w
ith
th
e
m
ic
ro
bi
al
gr
ow
th
be
ca
us
e
th
e
in
cr
ea
se
in
ac
id
ity
of
th
e
m
ed
iu
m
de
cr
ea
se
s
th
e
an
tib
ac
te
ri
al
ac
tiv
ity
.B
y
an
ot
he
r
si
de
,i
nc
re
as
es
th
e
ac
tiv
ity
of
ac
id
ic
su
bs
ta
nc
es
su
ch
as
pe
ni
ci
lli
n;
T
he
m
in
im
um
in
hi
bi
to
ry
co
nc
en
tr
at
io
n
(M
IC
)
ca
nn
ot
be
de
te
rm
in
ed
;
Is
di
ffi
cu
lt
to
ex
am
in
e
th
e
su
sc
ep
tib
ili
ty
of
fa
st
id
io
us
an
d
sl
ow
-g
ro
w
in
g
ba
ct
er
ia
;
It
is
la
bo
r-
in
te
ns
iv
e
an
d
tim
e-
co
ns
um
in
g;
C
an
re
pr
es
en
t
an
ex
pe
ns
iv
e
ap
pr
oa
ch
if
m
or
e
th
an
a
fe
w
dr
ug
s
ar
e
te
st
ed
;
T
he
la
ck
of
m
ec
ha
ni
za
tio
n
or
au
to
m
at
io
n
of
th
e
te
st
.
H
ew
itt
an
d
V
in
ce
nt
(1
98
9)
;M
ar
sh
an
d
G
oo
de
(1
99
4)
;E
lo
ff
(1
99
8)
;S
co
rz
on
ie
t
al
.(
20
07
);
R
ex
et
al
.(
20
01
);
O
st
ro
sk
y
et
al
.(
20
08
);
Ba
rr
y
an
d
T
ho
rn
sb
er
ry
(1
99
1)
;K
la
nc
ni
k
et
al
.(
20
10
);
W
ilk
in
s
an
d
T
hi
el
(1
97
3)
;J
ia
ng
(2
01
1)
;
Jo
rg
en
se
n
an
d
Fe
rr
ar
o
(2
00
9)
.
M
ac
ro
di
lu
tio
n
-G
en
er
at
io
n
of
a
qu
an
tit
at
iv
e
re
su
lt
(i.
e.
,t
he
M
IC
);
-E
as
e
of
pe
rf
or
m
an
ce
;
-E
co
no
m
y;
-M
or
e
ra
pi
d
re
su
lts
,t
he
re
fo
re
th
ei
r
de
ve
lo
pm
en
t
co
ul
d
be
hi
gh
ly
de
si
ra
bl
e.

R
eq
ui
re
s
la
rg
e
vo
lu
m
es
of
an
tib
io
tic
;
It
is
no
t
re
co
m
m
en
d
us
in
g
fo
r
m
os
t
ro
ut
in
e
pr
oc
ed
ur
es
;
T
im
e
co
ns
um
in
g;
It
is
te
di
ou
s;
M
an
ua
lt
as
k
of
pr
ep
ar
in
g
th
e
an
tib
io
tic
so
lu
tio
ns
fo
r
ea
ch
te
st
;
T
he
po
ss
ib
ili
ty
of
er
ro
rs
in
pr
ep
ar
at
io
n
of
th
e
an
tib
io
tic
so
lu
tio
ns
;
R
el
at
iv
el
y
la
rg
e
am
ou
nt
of
re
ag
en
ts
an
d
sp
ac
e;
G
en
er
at
e
la
rg
e
am
ou
nt
s
of
w
as
te
.
Jo
rg
en
se
n
an
d
Fe
rr
ar
o
(2
00
9)
;C
LS
I
(2
00
8)
;
W
ie
ga
nd
et
al
.(
20
08
);
O
st
ro
sk
y
et
al
.(
20
08
).
M
ic
ro
-d
ilu
tio
n
in
m
icr
ot
ite
r
pl
at
e
(9
6
w
el
ls)

T
he
m
et
ho
d
is
ro
bu
st
,i
s
no
t
ex
pe
ns
iv
e;
G
iv
es
re
pr
od
uc
ib
le
re
su
lts
an
d
co
nv
en
ie
nc
e
of
ha
vi
ng
pr
e-
pr
ep
ar
ed
pa
ne
ls
;

T
he
la
ck
of
or
po
or
gr
ow
th
of
m
an
y
an
ae
ro
bi
c
m
ic
ro
or
ga
ni
sm
s;
T
es
tin
g
so
m
e
fa
st
id
io
us
an
ae
ro
be
s
gi
ve
s
El
of
f
(1
99
8)
;O
st
ro
sk
y
et
al
.(
20
08
);
C
lin
ic
al
an
d
La
bo
ra
to
ry
St
an
da
rd
s
In
st
itu
te
(C
LS
I)
(2
00
9)
;
Jia
ng
(2
01
1)
;J
or
ge
ns
en
an
d
Fe
rr
ar
o
(2
00
9)
.
(C
on
tin
ue
d)
82 M. G. Campos et al.
T
ab
le
56
.
(C
on
tin
ue
d)
.
M
et
ho
ds
A
dv
an
ta
ge
s
D
is
ad
va
nt
ag
es
R
ef
er
en
ce
s
Is
ca
.3
2
tim
es
m
or
e
se
ns
iti
ve
th
an
ot
he
r
m
et
ho
ds
us
ed
in
th
e
lit
er
at
ur
e;
R
eq
ui
re
s
a
sm
al
l/m
in
im
al
qu
an
tit
y
of
re
ag
en
ts
,
sa
m
pl
es
an
d
sp
ac
e
th
at
oc
cu
rs
du
e
to
th
e
m
in
ia
tu
ri
za
tio
n
of
th
e
te
st
;
C
an
be
us
ed
fo
r
la
rg
e
nu
m
be
rs
of
sa
m
pl
es
;
C
an
te
st
th
e
su
sc
ep
tib
ili
ty
of
m
ul
tip
le
an
tim
ic
ro
bi
al
s
at
th
e
sa
m
e
tim
e;
En
ab
le
s
a
la
rg
er
nu
m
be
r
of
re
pl
ic
at
es
,
in
cr
ea
si
ng
th
e
ab
ili
ty
of
co
nf
id
en
t
re
su
lts
;
D
oe
s
no
t
re
qu
ir
e
hi
gh
le
ve
ls
of
sk
ill
;
Le
av
es
a
pe
rm
an
en
t
re
co
rd
;
It
de
cr
ea
se
s
th
e
in
te
ns
iv
e
la
bo
r
an
d
tim
e
co
st
co
m
pa
re
d
w
ith
ag
ar
-b
as
ed
m
et
ho
d;
T
he
co
ns
id
er
ab
le
sa
vi
ng
s
in
m
ed
ia
us
ag
e;
A
ls
o
al
lo
w
as
si
st
an
ce
in
ge
ne
ra
tin
g
co
m
pu
te
ri
ze
d
re
po
rt
s
if
an
au
to
m
at
ed
pa
ne
l
re
ad
er
is
us
ed
;
W
or
ks
be
tt
er
w
ith
an
tib
io
tic
s
be
ca
us
e
no
pr
ec
ip
ita
te
fo
rm
s;
C
om
pa
re
d
w
ith
ag
ar
-b
as
ed
m
et
ho
d,
br
ot
h
m
ic
ro
di
lu
tio
n
ca
n
de
cr
ea
se
m
uc
h
la
bo
r
an
d
tim
e;
Is
co
nv
en
ie
nt
ly
us
ed
fo
r
ro
ut
in
e
an
tim
ic
ro
bi
al
su
sc
ep
tib
ili
ty
te
st
in
g
in
cl
in
ic
al
la
bo
ra
to
ry
;
Is
a
su
ita
bl
e
an
d
fa
st
sc
re
en
in
g
m
et
ho
d
fo
r
M
IC
de
te
rm
in
at
io
n
co
m
pa
re
d
to
ot
he
r
m
et
ho
ds
.
in
co
ns
is
te
nt
an
d
un
re
lia
bl
e
re
su
lts
be
ca
us
e
of
po
or
gr
ow
th
of
st
ra
in
s
du
e
to
ex
ce
ss
iv
e
ex
po
su
re
to
ox
yg
en
du
ri
ng
th
e
se
t-
up
pr
oc
ed
ur
e;
So
m
e
in
fle
xi
bi
lit
y
of
dr
ug
se
le
ct
io
ns
av
ai
la
bl
e
in
st
an
da
rd
co
m
m
er
ci
al
pa
ne
ls
.
Standard methods for pollen research 83
T
ab
le
57
.
Su
m
m
ar
y
of
so
m
e
an
tim
ic
ro
bi
al
ef
fe
ct
s
of
be
e
po
lle
n.
Ph
ar
m
ac
ol
og
ic
al
ac
tiv
ity
Su
sc
ep
tib
le
m
ic
ro
or
ga
ni
sm
s
M
et
ho
ds
of
de
te
rm
in
at
io
ns
R
ef
er
en
ce
s
E.
co
li,
S.
au
re
us
,B
.c
er
eu
s
an
d
P.
ae
ru
gi
no
sa
A
ga
r
w
el
ld
iff
us
io
n
A
bo
ud
a
et
al
.(
20
11
)
F
un
gi
:
As
pe
rg
illu
s
fu
m
ig
at
us
,A
sp
er
gi
llu
s
ni
ge
r
an
d
As
pe
rg
illu
s
fla
vu
s.
Y
ea
st
s:
Ca
nd
id
a
al
bi
ca
ns
(C
.a
lb
ica
ns
),
Ca
nd
id
a
gl
ab
ra
ta
,C
an
di
da
kr
us
ei
an
d
Rh
od
ot
or
ul
a
m
uc
ila
gi
no
sa
.
B
ac
te
ri
a:
E.
co
li,
S.
au
re
us
,S
al
m
on
el
la
en
t
 e
ric
a
(S
.e
nt

 er
ica
),
P.
ae
ru
gi
no
sa
an
d
Li
st
er
ia
m
on
oc
yt
og
en
es
(L
.m
on
oc
yt
og
en
es
)
“
K
ac

 an
io
v
 a
et
al
.(
20
12
)
L.
m
on
oc
yt
og
en
es
,S
.e
nt

 er
ica
,E
.c
ol
i,
S.
au
re
us
an
d
P.
ae
ru
gi
no
sa
“
Fa
tr
co
v
 a
- S
ra
m
ko
v
 a
et
al
.(
20
13
)
Cl
os
tr
id
iu
m
bu
ty
ric
um
,C
lo
st
rid
iu
m
pe
rf
rin
ge
ns
,C
lo
st
rid
iu
m
hy
st
ol
iti
cu
m
,C
lo
st
rid
iu
m
in
te
st
in
al
e
an
d
Cl
os
tr
id
iu
m
ra
m
os
um
“
K
ac

 an
io
v
 a
et
al
.(
20
14
)
Pa
en
ib
ac
illu
s
la
rv
ae
,B
ac
illu
s
su
bt
ilis
(B
.s
ub
til
is)
,B
.c
er
eu
s
,E
.c
ol
i
an
d
P.
ae
ru
gi
no
sa
“
Ba
rb
os
a
et
al
.(
20
06
)
S.
au
re
us
an
d
St
re
pt
oc
oc
cu
s
py
og
en
es
“
C
ab
re
ra
an
d
M
on
te
ne
gr
o
(2
01
3)
En
te
ro
ba
ct
er
ia
ce
ae
ge
ne
ra
“
H
le
ba
et
al
.(
20
13
)
S.
au
re
us
an
d
St
ap
hy
lo
co
cc
us
ep
id
er
m
id
is
“
Ba
ltr
u s
ai
ty
te
et
al
.(
20
07
)
Vi
rid
an
s
st
re
pt
oc
oc
ci
“
T
ic
hy
an
d
N
ov
ak
(2
00
0)
Ag
ro
ba
ct
er
iu
m
tu
m
ef
ac
ie
ns
,P
.s
yr
in
ga
e
pv
.t
om
at
o,
X
.
ax
on
op
od
is
pv
.V
es
ica
t
 o
ria
,E
.a
m
ylo
vo
ra
,P
.c
or
ru
ga
ta
,R
.
so
la
na
ce
ar
um
,X
.c
am
pe
st
ris
pv
.C
am
pe
st
ris
,A
.v
iti
s,
C.
m
ich
ig
an
en
sis
su
bs
p.
M
ich
ig
an
en
sis
,E
.c
ar
ot
ov
or
a
pv
.
ca
ro
to
vo
ra
,P
.s
av
as
ta
no
ip
v.
sa
va
st
an
oi
,P
.s
yr
in
ga
e
pv
.
Ph
as
eo
lic
ol
a,
P.
sy
rin
ga
e
pv
.s
yr
in
ga
e
Pa
pe
r
D
is
k
di
ffu
si
on
Ba
si
m
et
al
.(
20
06
)
S.
au
re
us
,B
.s
ub
til
is,
P.
ae
ru
gi
no
sa
,K
le
bs
ie
lla
sp
.a
nd
B.
ce
re
us
“
C
ar
pe
s
et
al
.(
20
07
)
S.
au
re
us
,B
.C
er
eu
s,
B.
Su
bt
illi
s,
P.
ae
ru
gi
no
sa
an
d
K.
Pn
eu
m
on
ia
e
“
Ba
su
ny
et
al
.(
20
13
)
E.
co
li,
S.
au
re
us
,B
.s
ub
til
is,
P.
ae
ru
gi
no
sa
,K
.p
ne
um
on
ia
,
Sh
ig
el
la
fle
xn
er
i,
Sa
lm
on
el
la
ty
ph
ia
nd
Pr
ot
eu
s
m
ira
bi
lis
“
Es
w
ar
an
an
d
Bh
ar
ga
va
(2
01
4)
S.
th
yp
hi
m
ur
iu
m
an
d
E.
co
li.
“
Y
er
lik
ay
a
(2
01
4)
S.
au
re
us
[A
T
C
C
29
21
3]
,E
.c
ol
i[
A
T
C
C
27
85
3]
,E
nt
er
oc
oc
cu
s
fa
ec
al
is
[A
T
C
C
29
21
2]
,o
r
P.
ae
ru
gi
no
sa
[A
T
C
C
25
92
2]
M
ic
ro
di
lu
tio
n
in
M
ic
ro
tit
er
pl
at
e
El
of
f
(1
99
8)
S.
au
re
us
,S
.e
pi
de
rm
id
is,
E.
co
li,
En
te
ro
ba
ct
er
clo
ac
ae
,K
le
bs
ie
lla
pn
eu
m
on
ia
,P
.a
er
ug
in
os
a,
C.
al
bi
ca
ns
,C
.t
ro
pi
ca
lis
an
d
C.
gl
ab
ra
ta
“
G
ra
ik
ou
et
al
.(
20
11
)
Sa
lm
on
el
la
ty
ph
i,
S.
au
re
us
,B
.c
er
eu
s,
E.
co
li,
Z
yg
os
ac
ha
ro
m
yc
es
m
el
lis
,Z
yg
os
ac
ha
ro
m
yc
es
ba
ilii
,Z
yg
os
ac
ha
ro
m
yc
es
ro
ux
ii
an
d
Ca
nd
id
a
m
ag
no
lia
“
M
or
ai
s
et
al
.(
20
11
)
S.
au
re
us
,P
.a
er
ug
in
os
a,
E.
co
li
an
d
C.
gl
ab
ra
ta
“
Pa
sc
oa
le
t
al
.(
20
14
)
84 M. G. Campos et al.
15.2. Materials and methods
15.2.1. Sample characterization and preparation
The pollen samples are checked for the botanical origin
based on the microscopic analysis, according the meth-
odology described in Section 5.
For FTIR-ATR analysis a representative part of each
samples are carefully homogenizing and reduce to a fine
powder (Figure 53). A small amount of sample is put in
the apparatus with diamond solid ATR crystal properly
mounter, and press against crystal (Anjos et al., 2017).
15.2.2. Reference analytical methods
15.2.2.1. Total phenolic and flavonoids content ana-
lysis. The total phenolic content of the extracts was
determined using the Folin–Ciocalteu method as
described by Moreira et al. (2008). Total phenols con-
tent is expressed as mg of Galic Acid equivalents per g
of bee pollen (GAEs).
For flavonoids contents the aluminum chloride
method was used. Total flavonoids content is express
as mg of catechin equivalents per g of bee pol-
len (CAEs).
15.2.2.2. Antioxidant activity analysis. The antioxidant
activity was assessed by calculating the 2,2-diphenyl-1-
picrylhydrazyl (DPPH
) free radical scavenging activity
the trolox equivalent antioxidant capacity (ABTS)
according the methodology proposed in Section 16.
The concentration of the extract that induced a 50%
inhibition (EC50) was calculated from the graphic of the
percentage effect of eliminating radicals as a function of
the concentration of the sample extract solution.
For trolox results were expressed in millimol of
equivalent trolox per gram of sample.
15.2.3. FTIR-ATR collection and data processing
1. The data acquisition and processing was described by
Anjos et al. (2017). Pollen spectra are acquiring in a
Bruker FTIR spectrometer with a resolution of
4 cm1 in the wavelength region 4000–400 cm1 (or
similar) using a diamond single reflection attenuated
total reflectance (ATR).
2. obtain a duplicate spectrum per sample, for example,
with 32 scans per spectrum and calculate the average
for each sample.
3. Note: A background measurement is made before
each ten-sample analyzing using air. ATR crystal is
carefully clean before and between each analysis with
MilliQ water and isopropanol and the ATR crystal
should be dried with soft tissue paper. All experi-
ments are carried out at room temperature.
4. In a first screening all abnormal spectra are remove
and if it was necessary a new sample spectra acquisi-
tion is performed.
5. PLS regression is done by using OPUS 7.5.18 BRUKER
software. Firstly, the spectral data from the calibration
samples are regress against the measured parameters
and by means of full cross-validation with one sample
omitted in order to obtain a significant number of PLS
components (rank).
Different preprocess spectra could be tested for PLS
analysis with the following methods:
a. multiplicative scatter correction (MSC);
b. minimum maximum normalization (MinMax);
c. vector normalization (VecNor);
d. straight line subtraction (SLS);
e. constant offset elimination (ConOff);
f. first derivative (1stDer);
g. second derivative (2ndDer);
h. combination of the previously methods namely
1stDerþMSC, 1stDerþVecNor, 1stDerþ SLS.
For PLS analysis models with no spectral data processing
(None) are tested. All models can be calculated to a maximum
rank of 10 and the results of the cross-validation test for the
better values of coefficient of determination (r2), root mean
square error of cross-validation (RMSECV) and ratios of
performance to deviation (RPD) (Anjos et al., 2018).
15.3. Concluding remarks
These developed models have acceptable accuracy for
the screening analysis for total phenolic compounds,
total flavonoids and antioxidative capacity of bee pollen.
They are fast and easy to handle and could give a good
evaluation of the properties of bee pollen moisture
available in the market. In the future is important pro-
duce more robust models with a higher number of sam-
ples, produced in different countries to have a better
chemical characterization of the bee pollen to use in
the labeling.
A total of 134 pollen samples mixture was analyzed
and the frequency of each plant genus as predominant
pollen, secondary pollen, important minor pollen and
minor pollen is represented on Table 52. The variability
of genus observed in all samples is given by the richness
of floral resources of the region where the bee pollen
samples were collected. In a higher amount of samples,
the Cistus spp and Echium spp pollen were the predom-
inant. As secondary pollen the Castanea spp pollen was
the highest representative genus found in 34 samples,
followed by the Cytisus spp, Eucalyptus spp, Cistus spp
and Echium spp.
Castanea spp pollen was found in 60% of the samples
from as Predominant pollen to minor pollen. The same
distribution was observed for Cistus spp pollen that was
presented in 61% of the analyzed samples.
The spectra for pollen are very similar, like it can be
observed in Figure 54, which shows the average spec-
trum of different types of analyzed samples. All the
spectra analyses are characterized by bands between
1600 and 900 cm1, which correspond to the signal
Standard methods for pollen research 85
characteristic of the phenols compounds and is due to
the C-C and C-O vibration of phenolic and flavonoids
compounds. An absorbance band appeared at
1632 cm1, which could be assigned to the carboxyl
group in asymmetric stretching, presented also in the
honey samples (Anjos et al., 2015). The band from 3400
to 3000 cm1 is usually associated to the water content
in the samples because is due to OH groups but this
groups are also important in the identification of phen-
olic compounds, alcohols and carboxylic groups and
also the hydrogen vibration of amide N-H groups from
olive wastewater samples (Droussi et al., 2009). The
water content in this kind of samples usually varies
between 6 and 8%. The other important bands that
were observed in bee pollen samples are described in
Anjos et al. (2017).
Table 53 summarizes the statistical data for each par-
ameter evaluated. The higher values of coefficient of
variation denote the variability of data important for
good model calibration. The samples data must be div-
ided into two parts, one for calibration (30% to 50% of
data) and the other for validation set (70% to 50% of
data). PLS multivariate analysis was done with spectral
data in order to obtain calibration models for each ana-
lyzed parameter in pollen samples. The calibration mod-
els determination of total phenolic compounds, total
flavonoids and antioxidant capacity in pollen mixtures
were developed based on the highest r2, lowest stand-
ard error of calibration and cross-validation, lower
number of factors used in the calculation and higher
residual prediction deviation. Table 54 represents the
results for optimal calibration established. Values of r2
for selected models were greater than 0.859 for calibra-
tion and greater than 0.72 for the Cross validation. The
worst model founded was for the ABTS with a lower
RPD for cross-validation. In calibration all models pre-
sent values of RPD higher than 2.7. However, more
accurate model could be obtained with more samples,
from different regions and countries.
16. Standard methods for determination of
antioxidant bioactivity
16.1. Introduction
Bee pollen is considered as an important source of anti-
oxidant flavonoids and phenolic acids for human con-
sumption. Given the complex composition (view the
first sections of this chapter), the natural variation of
that composition, the storage conditions (Campos et al.,
2003; Melo & Almeida-Muradian, 2010), the different
ways to prepare the pollen extracts, and the different
assay conditions to evaluate the antioxidant capacity,
the determination and comparison of antioxidant cap-
acity is difficult to assess. A great deal of methods for
assessing the antioxidant capability of foods and bever-
ages have been developed, and reviews on this topic
have been published (Prior et al., 2005). In this section,
a proposal for standardizing some methods to assess
the antioxidant levels of bee pollen is done. As the anti-
oxidant capacity varies with the type of oxidants, more
than one assay is needed (Niki & Noguchi, 2000).
- DPPH
 free radical scavenging activity and the oxygen
radical absorbance capacity (ORAC) methods are pro-
posed for standardization of the routine assessment of
bee pollen antioxidant capacity; the former because of
its wide use in bee pollen, easiness of doing it, and
good correlation with other assays like b-carotene
bleaching (Morais et al., 2011) among others; and the
Figure 54. Average ATR spectrum of pollen samples.
86 M. G. Campos et al.
second one because of it has been proposed to be con-
sidered for standardization for foods, botanicals, nutra-
ceuticals, and other dietary supplements (Prior et al.,
2005), and this would allow to compare this property
between bee pollen and any other food.
16.2. Methods
16.2.1. 2,2-diphenyl-1-picrylhydrazyl (- DPPH
) free radical
scavenging activity principle
- DPPH
 is a stable organic nitrogen free radical. It has
a spare electron, which being delocalized in the mol-
ecule, confers a deep violet color and makes the com-
pound no capable to form dimers (Molyneux, 2004)
neither reacts with oxygen (Ionita, 2005). In ethanolic
or methanolic solutions, - DPPH
 absorbs between 515
and 528 nm (Almaraz-Abarca et al., 2004; Molyneux,
2004; S
anchez-Moreno, 2002), but when it is mixture
with a reducing substance, the reduced form of
-DPPH
 (- DPPH
-H, 2,2-diphenyl-1-picrylhydrazine) is
formed (Ionita, 2005), then the deep purple color
becomes to a pale purple and eventually, this becomes
to a pale yellow. The wide range of maximum absorb-
ance reported for the alcoholic - DPPH
 solution indi-
cates that the kmax must to be determined under a
given set of particular conditions. Stoichiometric deter-
minations have indicated that two moles of - DPPH
 are
reduced by one mole of water soluble vitamin E ana-
logue (Trolox) (Leong & Shui, 2002), but the interaction
of a potential antioxidant with -DPPH
 depends on its
structural conformation (Kaur & Kapoor, 2001). In a
-DPPH
 assay, this molecule represents the radical
formed in a system suffering oxidation and the antioxi-
dant to evaluate represents the reductant compound.
Although the -DPPH
 scavenging activity can be eval-
uated by electron spin resonance (Wasek et al., 2001;
Yu, 2001), the evaluation of the absorbance reduction
at 515 to 528 nm when in contact with a hydrogen
donator compound by using visible spectroscopy, is the
most common manner of evaluation. A quantitative esti-
mation of the activity using visible spectroscopy is based
on the registration of the differences between the initial
concentration of -DPPH
 in the reaction medium and
its diminution with increased amounts of antioxidant.
The -DPPH
 scavenging activity of one sample can
be expressed as the percentage of -DPPH
 free radical
scavenging remaining in the reaction (Equation (1)).
%DPPH
remaining ¼ ½½DPPH
remaining=½- DPPH
initial
	100 (1)
That manner has been used to express the -DPPH

scavenging activity in around 65.6% of reports concern-
ing bee pollen. However, this activity can also be
expressed in terms of the efficient concentration at 50%
(EC50), defined as the amount of antioxidant needed for
reducing by 50% the initial concentration of -DPPH
,
the larger the EC50 the less efficient the antioxidant.
S
anchez-Moreno et al. (1998) have proposed that
besides the concentration of antioxidant, the reaction
time to reach the plateau of the scavenging reaction
should be taken in account for estimating this activity,
according to Equation (2). Results from quercetin evalu-
ations suggest that the increasing in the reductor com-
pound do not affect the time to reach the plateau
(Campos, 1997).
AE ¼ 1=ðEC50ÞðTEC50Þ (2)
where AE means the antiradical efficiency, EC50 the effi-
cient concentration at 50% and TEC50 the time needed
to reach the steady state with EC50.
As both -DPPH
 and phenols are highly soluble in
methanol and ethanol, the antiradical activity is eval-
uated in either of those solvents without apparent
interference with the reaction; other solvents, like
water and acetone seem to under estimate the extent
of reduction (Guo et al., 2020). The use of buffered
methanolic solutions has been reported for the evalu-
ation of antiradical activity of bee pollen (Wang et al.,
2010); that solvent can affect the estimation of IC50 in
methanol and buffered methanol for antioxidants like
BHT, but the effect is lower for phenolic antioxidants
like propyl gallato, which bears several hydroxyl groups,
being the IC50 values comparable in both kinds of sol-
vents (Sharma & Bhat, 2009).
A great variation in the initial concentration of
-DPPH
 solution can be found in many reports of the
antiradical activity. According to Molyneux (2004), that
concentration should be in the range of 50 to 100 mM in
order to get solutions giving absorbance values less than
1, thus, observing the normal practice in spectrophotom-
etry to reach accurate measurements. The range given by
Sharma and Bhat (2009) was 25 to 70 mM.
Concerning the stability of the -DPPH
 solution, its
spectrum (200 to 700 nm) did not reveal neither
molecular nor quantitative alterations after eight hours
in light (Campos, 1997). For longer periods, the free
radical activity of the -DPPH
 solution is reduced about
2 to 4% per week (Molyneux, 2004).
Among the disadvantage of -DPPH
 assay are the
reversibility of the initial step of the reaction (Bondet
et al., 1997) and the fact of some substance, like pro-
teins, most likely present in total extracts of plants, can
precipitate in the alcoholic reaction medium (S
anchez-
Moreno, 2002). Prior et al. (2005) informed about the
following three disadvantages of this assay, the not easy
interpretation when the test compounds have spectra
overlapping -DPPH
 kmax; the overestimation of the
activity for small molecules, having better access to the
radical site (steric accessibility); and the inaccurate
interpretation of the antioxidant activity for a given
sample because -DPPH
 solutions are decolorized by
both reducing and hydrogen atom transfers. Contrary,
among the advantage of the -DPPH
 free radical scav-
enging activity are that it is a rapid method, does not
Standard methods for pollen research 87
have to be generated previously to the proper radical
scavenging as in other assays, the compound is commer-
cially available, the method is accurate and reproducible,
and needs only a UV-Vis spectrophotometer or a color-
imeter to perform. Also, - DPPH
 assay is independent
of sample polarity (Koleva et al., 2002). Besides, good
correlations between this assay and TEAC (Marghitas¸
et al., 2009) and with b-carotene bleaching assay
(Morais et al., 2011) have been reported, meaning that
-DPPH
 test may represent a good approaching to
evaluate the antioxidant activity for complex samples
like bee pollen extracts.
There is a high correlation between the decreasing
of DPPH
 concentration and the increasing of flavonoid
or phenolics concentration in the reaction medium
(Almaraz-Abarca et al., 2004, 2006, 2008; Barriada-
Bernal et al., 2014). However, a clear correlation
between the DPPH
 scavenging activity and the flavon-
oid or phenol concentration in different samples has
not been found in all systems (Morais et al., 2011; Negri
et al., 2011). This suggests that the types of phenolics
present in a particular combination in a given sample
are involved in the manifestation of the antioxi-
dant properties.
S
anchez-Moreno (2002) informed about a long list of
compounds from different sources analyzed with the
-DPPH
 assay, which could be update with the reports
concerning the bee pollen activity.
16.2.1.1. Variations in the - DPPH
 assay for bee pol-
len. The main variations of the -DPPH
 assay found in
the several reports concerning bee pollen refer to the
-DPPH
 concentration of the solution used to carry
out the analysis, reaction time, and the proportion of
volume of - DPPH
 solution related to volume of
extract or sample to analyze (Table 55). Differences are
also found in the form of expression of the activity.
Under this situation important difficulties are to be
faced when comparisons of scavenging activity of differ-
ent bee pollen samples are done. Thus, the expression
of the antiradical activity in terms of IC50 would repre-
sent a more accurate activity than the expression in
terms of % activity, which is estimated at one extract
concentration selected in an arbitrary manner and men-
tioned in the reports only in few cases.
16.2.1.2. Material and methods for - DPPH
 assay.
1. Prepare a solution of - DPPH
 in ethanol or methanol
(depending on the solubility of the compounds to be
tested), and calculate the UV-Vis absorption that
should be close to 1 (Absorption Unities – AU). The
concentration of the free radical depends according to
the method choose (see Table 55).
2. Prepare a test solution with the sample. For bee pol-
len a ratio of 10mg/ml of solvent can be used as a
start point. After a preliminary assay, if needed, differ-
ent concentration of the simple solution can be done.
3. Measure 1 or up to 4ml of - DPPH
 according to the
method choose (see Table 55) for at least 5
test tubes.
4. Add to each different concentration of the sample sol-
utions and adjust with solvent for an equal final vol-
ume in each test tube. Example: 1ml of - DPPH

solution þ (50 ml extract þ 950 ml of solvent), to
complete a final volume of 1ml.
5. Wait the time necessary for the complete reaction,
according to the method choose (see Table 55)
6. Measure the absorption at 517 nm in the visible range
with a spectrophotometric system.
7. Do a chart with the concentration of the sample and
the respective consume of -DPPH
. Example: O mg of
sample þ 1ml of - DPPH
 correspond to 0.96AU.
These values correspond to zero activity. When the
concentration of the sample increase (if active) the
-DPPH
 solution absorption decreases and then is
possible to determine the concentration of the sample
necessary to decrease this AU to 50% corresponding
to the EC50.
8. Proceed as above utilizing a standard (ascorbic acid,
trolox, various flavonoids or other polyphenols,
depending of the simple under evaluation) for com-
parison of the results and to validate the accuracy of
the assay.
16.2.2. Antioxidant capacity using ORAC (oxygen radical
absorbance capacity) assay principle
The ORAC method measures the sequestrating capabil-
ity of an antioxidant against the induced radical 2,20-azo-
bis(2-amidinopropane)dihydrochloride (AAPH) in
reactions which are temperature sensitive; 37 C is the
most frequent temperature of assay. The radical reacts
with a fluorescent compound forming a nonfluorescent
product. The protector effect of antioxidant is esti-
mated by calculating the area under a fluorescence
decreasing curve over time, comparing with a reference
substance, which is not an antioxidant. ORAC method
was developed by Glazer (1990), using phycoerythrin,
but currently fluorescein and dichlorofluorescein are
preferred because of their stability and less reactiveness
(Prior et al., 2005). At present, ORAC assay is highly
automated. Special plates of fluorometers are used.
Arruda (2013) used plate wells for 300 ll, and it was used
for the external ones 300 ll of distilled water. Other
holes were filled with 25 ll of phosphate buffer (blank)
and the others with 25ll of the extract dilutions. Then it
was injected 150 ll of fluoresceine, waits for 30minutes,
and then injected 25ll of the AAPH solution to initiate
the reaction of peroxyl radical with the fluorescein (oxy-
gen radical absorbance capacity). ORAC values are usu-
ally expressed as Trolox equivalents/g sample, then a
standard curve using three to five concentrations of
TroloxVR (25ll) vs. the fluorescence intensity is con-
structed. ORAC unit is equivalent of a protection against
oxidation offered by a 1 mmol of Trolox. The
88 M. G. Campos et al.
fluorescence must be monitored every minute, using as
work temperature 37 C. It is used extinction wavelength
at 493 nm (485/20 filter) and at 515 nm (filer 528/20) of
emission. For calculation, it is used the sample area under
the curve of the testing compounds minus the area of the
blank. The same calculation is made for the Trolox curve.
The area under the curve (AUC) is calculated after
30minutes, when it is injected 25ll of the AAPH solution
to start the reaction, this delay in the starting of reaction
avoids underestimation of antioxidant activity due to
effects of secondary potential antioxidant products (Prior
et al., 2005).
The area under the curve (AUC) is calculated using
Equation (3).
AUC ¼ 1þ
XN
i¼1
fi
f0
where fo ¼ initial fluoresence at time zero; fi ¼ fluores-
cence registration after i
At the beginning of the reaction kinetics there is a
time in which the fluorescence is maintained, this is the
lag phase, in it there is no consume of peroxide by
fluorescent probe due to the presence of the antioxi-
dants and consequently enhance the area under the
curve. Prior et al. (2005) consider that methods like
ORAC, which measures the antioxidant capacity by
AUC, take in account both inhibition time and degree;
this gives a more accurate estimation of the action of
an antioxidant than the methods which are based on
the evaluation at a fixed time or inhibition degree.
16.2.2.1. Material and methods for antioxidant activ-
ity (ORAC) assay. The ORAC procedure recommended
use an automated plate reader (Synergi HT, Bio TeK,
USA) with 96-well plates (Arruda et al., 2013; Ou et al.,
2001), as follows:
1. Conduct all the analyses in a pH 7.4 (75mM) phos-
phate buffer at 37 C under condition with a blank
sample in parallel.
2. Dilute the extracts with sodium phosphate buffer
(75mM, pH 7.4). Peroxyl radical will generated using
2, 20-azobis (2-amino-propane) dihydrochoride aque-
ous solution – that will be prepared for each run
(153mM). Fluorescein will be used as sub-
strate (40 nM).
3. Program the ORAC analyzer to record the fluores-
cence of FL every minute after addition of AAPH. All
fluorescent measurements express relatively to the
initial reading. Fluorescence analysis conditions are
performed as follows: excitation at 493 nm (filter 485/
20) and 515 nm (filter 528/20). The standard curve
must be linear between 6.25 and 100 mM TroloxVR .
4. Calculate the final results using the differences of
areas under the FL decay curves between the blank
and a sample and express as mM TE/g bee pollen.
16.3. Concluding remarks
The great diversity of bee pollen that can be found
worldwide, varying in the antioxidant capabilities,
encourages to propose standard evaluations for this
property, allowing a comparison among bee pollens
from different geographical and botanical origins.
Standards evaluations also would be useful for under-
standing the effects of several environmental factors on
the antioxidant properties, and for monitoring bee pol-
len quality.
17. Standard methods for anti-inflammatory
activities in bee pollen
17.1. Introduction
The inflammation process has been defined by several
pathologists with different approach. Most pathologists
would probably agree that inflammation represents a
response of living tissue to local injury; that it leads to
the local accumulation of blood cells and fluid (Ryan &
Majno, 1977). Briefly, inflammation is a physiological
response to the damage of tissues or cells that is
caused by physical or biological agents and also free
radicals involving different reactions intended to remove
the cause and repair the damage (Bogdanov, 2015).
Paulino et al. (2008) defined the inflammation as the
process that involves production and/or release of
mediators from neurons or damaged tissues, which are
responsible for different responses including pain.
For determination of standard methods on anti-
inflammatory activities much methodology in vitro and
in vivo has been developed by several researchers
worldwide. Recently Pascoal et al. (2014) following the
methodologies in vitro described by da Silva et al.
(2012), present a method based on the measurement of
the inhibitory effect of bee pollen (BP) on the reactions
catalyzed by hyaluronidase was evaluated.
17.2. Determination of standard methods
17.2.1. Anti-inflammatory activities of BP in vitro
17.2.1.1. BP extract preparation. For BP extract prep-
aration the procedures described by Pascoal et al. (2014)
are described below regarding to the procedures.
17.2.1.1.1. Materials and reagents.
1. BP sample(s).
2. Methanol (CH3OH).
3. Whatman filter paper No. 4.
4. Flasks.
5. Rotary vacuum evaporator.
17.2.1.1.2. Procedure. Note: The sample concentration
and diluents used varies from the author to other:
1. BP and Methanol (MeOH) were mixed (1:2) (w/v).
2. Macerate for 72 h at room temperature.
Standard methods for pollen research 89
3. Then, the solution was filtered by Whatman filter
paper No. 4.
4. The solid residue was re-extracted.
5. Mix the two filtered into a flask, thus obtaining the
methanol extracts of bee pollen.
6. Evaporate the extracts on a rotary evaporator at
45 C for approximately 1h.
7. Finally, the dried extract of pollen was stored in the
dark at room temperature, for further analysis.
17.2.1.2. BP sample preparation. 17.2.1.2.1. Materials
and reagents.
1. BP extract.
2. Hyaluronidase enzyme (Type IV-S: bovine testes).
3. Calcium chloride (CaCl2).
4. Hyaluronic acid sodium salt.
5. Potassium tetraborate (K2B4O74H2O).
6. Water-bath.
7. p-dimethylaminebenzaldehyde (DMSO).
8. Spectrophotometer.
17.2.1.2.2. Procedure.
1. Prepare BP extract concentrations [5mg/ml – 25mg/
ml] using dimethyl sulfoxide (DMSO) as solvent.
2. Prepare a mixture of 50 ll of BP’ extract and 50ll
(350 units) of hyaluronidase enzyme (Type IV-S:
bovine testes).
3. Incubate at 37 C for 20min.
4. Add 1.2ll of calcium chloride (2.5	 103 M/l) to
activate the enzyme.
5. Incubate at 37 C for 20min.
6. Add 0.5ml of hyaluronic acid sodium salt (0.1 M/l) to
start the reaction.
7. Incubate the mixture at 37 C for 40minutes.
8. Then add 0.1ml of potassium tetraborate 0.8 M.
9. Incubate in boiling water bath for 3minutes.
10. Place the mixture at 10 C.
11. Add 3ml of (DMSO).
12. Afterwards incubate at 37 C for 20minutes.
13. Measured the absorbance at 585 nm using water
as control.
17.2.2. Anti-inflammatory activities of BP in vivo
17.2.2.1. BP extract preparation. For BP extract prep-
aration the procedures described by Maruyama et al.
(2010) are suggest.
17.2.2.1.1. Materials and reagents.
1. BP sample(s)
2. Ethanol (EtOH) and others.
17.2.2.1.2. Procedure.
1. BP (200 g) extracted with water (Water BP) or 95%
ethanol (EtOH BP).
2. Stir for 4 h at 60 C (Water BP) or stir for 16 h
(EtOH BP) at room temperature.
3. Take the filtrate to dryness and powdered.
4. The yields are stored for further use.
17.2.3. Animals
For the animal experimental protocol procedures, the
Guide for the Care and Use of Laboratory Animals
Eighth Edition (2011) is suggested.
17.3. Concluding remarks
The high anti-inflammatory effects of BP compared to
such nonsteroidal anti-inflammatory drugs as naproxen,
analgin, phenylbutazone, or indomethacin also has been
reported (Pascoal et al., 2014). The mechanism of the
inflammatory effect is about inhibiting the activity of
enzymes (cyclooxygenase and lipoxygenase) which are
responsible for turning arachidonic acid into such toxic
compounds as prostaglandin and leukotrienes, inducing
acute and chronic inflammatory conditions in tissues.
Flavonoids and phenolic acids are mainly responsible for
such actions, but fatty acids and phytosterols also take
part in this process (Akkol et al., 2010; Choi, 2007;
Komosinska-Vassev et al., 2015; Pascoal et al., 2014).
Kaempferol it is a compound present in BP and has
been attributed ability to inhibit the activity of two
enzymes: hyaluronidase, which is the enzyme catalyzing
depolymerization of hyaluronic acid, and elastase, which
hydrolyzes elastin, strengthens the connective tissue
and seals blood vessels. This results in decreased tran-
sudates, inflammatory reactions and swellings. Blood cir-
culations in the vessels improve and, therefore, skin
becomes moistened and tight. Quercetin another fla-
vonoid compound with several bioactivities as anti-
inflammatory, antiedematous and analgesic action, may
result by inhibiting the activity of histidine decarboxyl-
ase, lowers the histamine level in the organism
(Kanashiro et al., 2007; Lee et al., 2009; Sahin et al.,
2011). Studies done by Lee et al., 2009 in order to elu-
cidated the anti-inflammatory activities of several sub-
class flavonoids such as the antioxidant property,
inhibition of NO production, and inhibition of inflamma-
tory enzymes (hyaluronidase, collagenase, LOX, and
COXs), found that Kaempferol, quercetin, myricetin,
and rutin inhibited hyaluronidase reaction, while apige-
nin, baicalin, and baicalein showed specific inhibition to
collagenase reaction and catechin, which is flavon-3-ols-
subclass, had suppressed both hyaluronidase and colla-
genase reaction. The degradation of the hyaluronic acid
by the hyaluronidase enzyme may cause bone loss,
inflammation and pain. As consequence, the determin-
ation of the hyaluronidase enzyme is an indirect way to
assess the anti-inflammatory activity (da Silva et al.,
2012). Choi (2007) evaluated the antinociceptive and
anti-inflammatory activity of pine (Pinus densiflora) pollen
in mice. The antinociceptive activity was determined
using acetic acid-induced abdominal constriction and
formalin-induced licking, and the hot plate test and anti-
inflammatory effects were evaluated using carrageenan
and formalin-induced paw edema, and arachidonic acid-
induced ear edema. The obtained results showed that
90 M. G. Campos et al.
ethanol extract of pine pollen (100 and 200mg/kg) pro-
duced a significant inhibition of both phases of the for-
malin pain test in mice. Also, the pine pollen extract
produced a significant inhibition of carrageenan and for-
malin-induced paw edema as well as arachidonic acid-
induced ear edema in mice. These activities could be
due to different polyphenols found in pine pollen
related with antinociceptive and anti-inflammatory
actions. Among other flavonoids, Chrysin has been
reported as to possess anti-inflammatory activity, anti-
oxidant properties and promotes cell death by perturb-
ing cell cycle progression (Pichichero et al., 2011).
Another studies developed by Maruyama et al. (2010) in
order to investigate the anti-inflammatory effect of bee
pollen from Cistus sp. of Spanish origin by a method of
carrageenan-induced paw edema in rats, and to investi-
gate the mechanism of anti-inflammatory, found that the
ethanol extract of bee pollen show a potent anti-inflam-
matory activity and its effect acts via the inhibition of
nitric oxide (NO) production, besides the inhibitory
activity of COX-2. Maruyama et al. (2010) and,
reported that some flavonoids found in plant foods
included also in BP may partly participate in some of
the anti-inflammatory action. Therefore, the anti-inflam-
matory effects of BP may be due to their flavonoids
compounds as Kaempferol, quercetin, myricetin, and
rutin that inhibit the inflammatory enzymes such as hya-
luronidase, collagenase, LOX, and COXs, helping the
retardation of inflammatory responses.
18. Standard methods for antimicrobial activities
in bee pollen
18.1. Introduction
In order to investigate new antimicrobial agents from
extracts of plants and other natural products to be
used in pharmaceutical and cosmetic products, several
methods have been developed by several researchers
for evaluate the antibacterial and antifungal activity of
the plant extracts. Nowadays, the most known assays
have been based on diffusion in agar, macro and micro-
dilution methods (Ostrosky et al., 2008). Table 56
shows the advantages and disadvantages of these meth-
ods with references. Microdilution methods have been
reported as the most widely used testing methods that
use commercially marketed materials and devices, and
the manual methods that provide flexibility and possible
cost savings include the disk diffusion and gradient diffu-
sion methods (Jorgensen & Ferraro, 2009). Some meth-
ods provide quantitative results and all provide
qualitative assessments using the categories susceptible,
intermediate or resistant. Also, it’s important to high-
light that all of these methods have strengths and weak-
nesses, including organisms that may be accurately
tested by the method; but in general way, these current
testing methods provide accurate detection of common
antimicrobial resistance mechanisms (Jorgensen &
Ferraro, 2009).
According to the information available the diffusion
in agar technique is widely used but problems associ-
ated with this have been reported. However, the Micro-
dilution technique in microtiter plate (96 wells) has been
proved to be quick provides reproducible results and
was useful in screening plants for antimicrobial activity.
All of these methods have been reported that each of
them has strengths and weaknesses. Also, it is import-
ant to highlight that those methods are widely used in
the worldwide by several researchers for antimicrobial
activities determinations but sometimes suffer minor
modifications (Estevinho et al., 2008).
18.2. Materials and methods
18.2.1. BP sample collection
Accordingly, with Morais et al. (2011), for getting BP
samples from beekeepers, the selected beehives were
equipped with bottom-fitted pollen traps. After dried
the BP, the samples were delivered to the Microbiology
Lab, where they were stored in a dark place at room
temperature (20 C) for further analysis.
18.2.2. Extract preparation of BP for antimicrobial analysis
Nowadays, many experiments in vitro and in vivo are
performed with Pollen Methanolic Extracts (PME) and
Pollen Ethanolic Extracts (PEE) to confirm its antimicro-
bial activity. According to the revised bibliography, once
the pollen is received in the laboratories from beekeep-
ers the methods for samples preparation commonly
used are based on Methanol (MeOH) (Graikou et al.,
2011; Kac
aniov
a et al., 2012; Pascoal et al., 2014) and
Ethanol (EtOH) (Eswaran & Bhargava, 2014; Hleba
et al., 2013; Kac
aniov
a et al., 2014) pollen extracts. In
dependence of the final concentration desired, amounts
of BP are weighted and dissolved in methanol or etha-
nol aqueous solutions. These methodologies have been
used by several researchers with some modifications.
The BP extracts could be prepared accordingly with
Pascoal et al. (2014) as described below regarding to
the procedures:
1. The sample concentration and diluents used varies
from the author to other.
2. Place the mixture under magnetic stirring at 150 rpm
for 1h.
3. Filtration using Whatman No. 4 paper.
4. Repeat this procedure once more.
5. Mix the two filtered into a flask, thus obtaining the
methanol extracts of BP.
6. Evaporate the extracts on a rotary evaporator at
45 C for approximately 1h.
7. Keep extracts in dark and in refrigeration until their
further use.
Standard methods for pollen research 91
8. Preparing different sample concentrations using
dimethyl sulfoxide (DMSO) as solvent.
18.2.3. Antimicrobial analysis by agar disk diffusion
18.2.3.1. Preparation of inoculum suspension. The
preparation of inoculum could be accordingly with
Matuschek et al. (2014) with bit modifications as
described below regarding to the procedures:
1. Inoculate the bacteria in broth of Muller Hinton Agar
or Brain Heart Infusion Broth (BHIA); and Sabouraud
dextrose Agar or potato dextrose agar for yeast.
2. Incubate for 18 to 24 hours for bacteria and 48 h
for yeast.
3. Inoculate 3 to 4 colonies of bacteria in 5ml Tryptic
Soy Broth (TSB) and use Yeast Extract Peptone
Dextrose (YPD) for yeasts.
4. Incubate at 37 C for bacteria and 25 C for yeast
until to reach an optical density 0.5 of Marc Farland
standard, which is equivalent approximately from
1–2	 108 CFU/ml for bacteria and 1–5	 106 UFC/ml
for yeasts.
Note: Alternatively, may prepare directly from the culture
medium with agar 18 to 24 hours for bacteria and 48 hours for
yeast, suspending the colony in saline or sterile nutrient broth
until an optical density of 0.5 Marc Farland standard, to
perform this step properly, either a photometric device should
be used or, if done visually, adequate light is needed to visually
compare the inoculum tube and the 0.5 Marc Farland standard
against a card with a white background and contrasting black
lines. For 0.5 Marc Farland standard preparations, see Cavalieri
et al. (2005) on page 220.
5. With swab, remove a small portion of inoculum is
evenly spread on plates containing Muller-Hinton broth
by passing the swab over the entire surface in
three directions.
6. Let dry for 5minutes.
7. Apply 10 ml of the antimicrobial solution on filter
paper disks of 6mm in diameter, at different concen-
trations to be tested.
8. Place the antibiotic disks on the agar surface using
sterile forceps, by pressing the agar.
9. Let the plates 15minutes at room temperature, in
order to initiate diffusion of antibiotics.
10. Incubate the plate upside down at 37 C for 18
to 24 hours.
11. Measure the diameter of the inhibition zones with a
ruler. To facilitate the measurement, the plates should
be placed in a dark background surface. Figure 55
shows the inhibition zone around the disk.
18.2.4. Antimicrobial analysis by agar well diffusion
18.2.4.1. Preparation of inoculum suspension. The
preparation of inoculum could be accordingly with
(Clinical and Laboratory Standards Institute (CLSI),
2012) with bit modifications as described below regard-
ing to the procedures:
1. Please, see proceedings described in points 1 to 6
regarding to agar Disk diffusion, then, follow the
next step.
2. Forming wells in the culture medium previously inocu-
lated from 6 to 8mm, with the aid of stainless-
steel cylinders.
3. Fill the wells with 100 ml of the substances to
be analyzed.
4. Incubate the plate upside down at 37 C for 18 to
24 hours for bacteria and yeast to 25 C for 48 h.
5. Measure the diameter of the inhibition zones with a
ruler. To facilitate the measurement, the plates should
be placed in a dark background surface.
Note: Use a positive control (may be an antibiotic)
and a negative control (solvent used for dissolving the
dry extract in different concentrations); The disks are
positioned at 15mm plate edge and widely separated
from each other, so there is no overlap of zones of
inhibition (approximately 20mm). The pH of the culture
medium should be between 7.2 and 7.4, and the recom-
mended depth is approximately 4mm (Barry &
Thornsberry, 1991). Figure 56 shows the inhibition
zone around the well diffusion.
18.2.5. Antimicrobial analysis by micro-dilution in microtiter
plate (96 wells)
18.2.5.1. Preparation of inoculum suspension. The
preparation of inoculum could be accordingly with
Cavalieri et al. (2005) in Manual of Antimicrobial
Susceptibility Testing, with bit modifications as
described below.
Procedure
1. In sterile 96 well microplate, placing 200 ml of the
sample (pollen extract) in the wells of the
first column.
Figure 55. Inhibition zones around the disks.
92 M. G. Campos et al.
2. In wells 2 to 8 and 11 to 12 put 100 ml medium
(nutrient broth), the wells 11 and 12 being the posi-
tive control.
3. In wells 9, 10 (negative control), putting 200 ml
of medium.
4. Perform serial dilutions until the well 8.
5. Please, see proceedings described in points 1 to 4
regarding to agar Disk diffusion, then, follow the
next step.
6. Then, within 15minutes of adjusting the inoculum to
the 0.5 Marc Farland turbidity standard, mix the sus-
pension and dilute it so that the final concentration
in each well is 5 x 105 CFU/ml. Deliver 2.0ml of the
original suspension into 38ml of water (1:20 dilu-
tion). From 1:20 dilution, prepare another 1:10 dilu-
tion that will be the inoculum.
7. Add 100 ml of inoculum in each well except for the
columns 1, 9 and 10. Inoculate MIC panel carefully to
avoid splashing from one well to another. Failure to
adjust the inoculum and to dilute it within 15minutes
may adversely affect the concentration of organisms
and test results.
8. Cover the microplate with a breathable film.
9. Incubate the yeast to 48 hours at 25 C and the bac-
teria at 37 C for 24 hours.
10. After this period, all wells except the first column,
add 20 mL of TTC (2, 3, 5-triphenyl tetrazolium
chloride) at 1%.
11. Incubate plates again for another 3 hours at 37 C
and 25 C for bacteria and yeast, respectively.
12. Reading: See the color change (red/pink), indicating
the reduction of TTC by microorganisms. The min-
imum inhibitory concentration (MIC) is defined as
the lowest concentration of pollen extract that is
capable of inhibiting visible growth, as indicated by
TTC staining. Figure 57 shows the color change indi-
cating the reduction of TTC by microorganisms in
the wells of microplates.
18.2.5.2. Calculating the lethal concentrations.
1. After reading the microplates at 48 hours.
2. Seeding in plates (Mueller-Hinton agar for bacteria and
Sabouraud for yeasts) in duplicate 20 ml of each of the
wells in which there has been no microbial growth
and the last in which growth was obtained.
3. Then mix thoroughly the contents of each well.
4. Once homogenized, inoculate 20 ml of each well in
duplicate plates in their respective medium.
5. Incubate during 24 and 48 hours for bacteria and yeast
respectively.
6. Then make counts.
7. If they are <10 or >10 CFU means that the sub-
stance has bactericidal and bacteriostatic activity,
respectively.
18.3. Concluding remarks
The Agar well diffusion techniques (Table 57) are used
widely, problems associated with this technique are
reported in Table 56). Therefore, Morais et al. (2011),
Graikou et al. (2011) and Pascoal et al. (2014) used a
Micro-dilution technique because is quick, worked well
with Staphylococcus aureus, Enterococcus faecalis,
Pseudomonas aeruginosa (P. aeruginosa), and Escherichia
coli (E. coli) also this method gave reproducible results;
provided a permanent record of the results and was
useful in screening plants for antimicrobial activity (Eloff,
1998). Although BP has antimicrobial properties, it is
important to highlight that these properties are depend-
ing on the type of pollen, concentrations applied and
the inhibitions can occur against different species of bac-
teria and inhibition ratios can also vary (Carpes et al.,
2007; Erkmen & Ozcan, 2008). For elucidate these affir-
mations, Campos et al. (1998) and Erkmen and Ozcan
(2008) in their studies when investigated the antibacter-
ial activities of BP on some bacteria, yeasts and molds,
the obtained results showed that, the used concentra-
tions had no antimicrobial effect on the bacteria and
fungi. Carpes et al. (2007) in their studies found similar
results as previously mentioned researchers. In general
way the results obtained by all of cited researchers
Figure 56. Inhibition zone around the well diffusion.
Figure 57. Microtiter plate (96 wells), showing the color
change indicating the reduction of TTC by microorganisms.
Standard methods for pollen research 93
even if the methodologies applied were different
showed that the Gram-negative bacteria were more
resistant to the BP action than the Gram-positive bac-
teria, what may be related with the additional outer
layer membrane, impermeable to most molecules, that
consists of phospholipids, proteins and lipopolysacchar-
ides (Silici & Kutluca, 2005).
19. Determination of standard methods for anti-
mutagenic activities in bee pollen
19.1. Introduction
Some published studies, reported the beneficial effects of
BP to human health such as the prevention of prostate
problems (Shoskes, 2002) and tumors (Zhang et al., 1995).
The increase of the of the large number of mutagens that
has widespread in the world, led to researchers to investi-
gate and discovering anti-mutagenic materials that limit and
stop the harmful effects of mutations on the physiology of
the organism. The most important of these materials has
been reported as a plant extracts (Jaiswal et al., 2014).
Over the past decade, some studies have lauded the anti-
mutagenic properties of diverse beekeeping products, such
as BP, propolis, honey mixtures and royal jelly, against the
influence of some chemical and physical mutagens (Bariliak
et al., 1996). Recently, studies done by Abdella et al. (2009)
and Tohamy et al. (2014) when studied the protective effi-
ciency of Egyptian BP extracts and propolis, treated mice
with cisplatin [cis-diammine-dichloro-platinum(II)]
(CDDP), observed that BP extracts presented potential
chemoprotective against CDDP induced genotoxicity in
mice by another side found that the chemoprotective fre-
quency of BP extracts was much greater than propolis.
Cisplatin (CDDP) is known as an inorganic platinum com-
pound with a broad spectrum anti-neoplastic activity
against different types of human tumors as testicular, ovar-
ian, breast, lung, bladder, head and neck cancer (Khynriam
& Prasad, 2003; Pabla et al., 2008; Siddik, 2003). The geno-
toxicity action of CDDP is due to its ability to bind with
DNA, block and prolong the cell division in the G2 phase
that is related to the inhibition of chromatin condensation
(Aly et al., 2003). Bariliak et al. (1996) when studied apicul-
ture products (propolis, beebread, honey and queen bee
milk) observed a chemoprotective effect of these beehive
products against the influence of some chemical and phys-
ical mutagens. BP was able to reduce the chromosome
damage induced by the three cancer drugs used, thus sup-
porting their use as a chemoprotective agent (Shi
et al., 2010).
19.2. Determination of standard methods for anti-
mutagenic activities in BP
19.2.1. Preparation of BP extracts
Different methods for preparations of BP extracts have
been reported (Carpes et al., 2007; Kaur et al., 2015;
Pascoal et al., 2014; Pinto et al., 2010).
Procedure (according with Kaur et al., 2015).
1. The sample concentration and diluents used varies
from the author to other.
2. Milling 1 g of pollen grains.
3. Homogenize with the help of needle in Petri dish.
4. Add 7.5ml of 70% ethanol to the pollen mixture.
5. Keep at 70 C for 1 h with agitation of 1min after
every 10min interval.
6. Separate the supernatant.
7. Re-extract solid residue with 70% ethanol.
8. Repeat the process 3–4 times.
9. Group supernatants in a 25ml conical flask.
10. The collected supernatant is considered as
100% extract.
11. Store these extracts at 5 C till further analysis.
12. Prepare different concentrations viz., 25%, 50%, 75%
and 100% of pollen extracts using double dis-
tilled water.
19.2.2. Anti-mutagenic activities in vitro
19.2.2.1. Materials and reagents.
a. BP.
b. Ethyl methanesulfonate (EMS) and other chemicals.
c. Methanol (CH3OH) and ethanol (CH3CH2OH).
d. Folin–Ciocalteu reagent chloroform (CHCl3).
e. Sodium carbonate (Na2CO3).
f. Gentaminice and fluconazole.
g. Gallic acid and (þ)-catechin.
h. The bovine testicular hyaluronidase (350 units).
i. Potassium salt of human umbilical cord hyalur-
onic acid.
j. The culture mediums.
k. The TTC solution (235-Triphenyl-2H-tetrazo-
lium chloride).
l. High purity water (18MX cm) and other chemicals.
19.2.2.2. Procedure.
1. Yeast cells (D7 diploid strain of Saccharomyces cerevi-
siae ATCC 201137), according Pascoal et al. (2014).
2. Prior to each experiment the S. cerevisiae D7 strain
must be tested for the frequency of spontaneous con-
vertants at the tryptophan (trp) locus and revertants
at the isoleucine (ilv) locus.
3. Cells from a culture with low spontaneous gene con-
version and reverse point mutation frequencies were
grown in a liquid medium at 28 C until they reached
the stationary growth phase.
4. Yeast cells pelleted and re-suspended in a volume of
0.1 M sterile potassium phosphate buffer, pH 7.4, to
obtain a final mixture of 2	 108 cell/ml.
5. The mixture (4ml) contain 1ml of cell suspension,
potassium phosphate buffer and methanolic extracts
of BP, in order to reach final concentrations of 0.00,
0.25, 0.50 and 0.75mg/L.
6. Incubate the mixture under shaking for 2 h at 37 C.
94 M. G. Campos et al.
7. Cells are plated in complete and selective media to
ascertain survival, trp convertants and ilv revertants.
8. Ethyl methanesulfonate (EMS), a mutagenic compound,
is used as control.
19.2.3. Anti-mutagenic activities of BP in vivo
19.2.3.1. Materials and reagents.
a. BP.
b. Cis-platin (CDDP) (cis-diamminedicholorplatinum).
c. Ethyl methanesulfonate (EMS) and other chemicals.
19.2.3.2. Experimental animals. For animal experimen-
tation, we recommend to proceed according to the
Guide for the Care and Use of Laboratory Animals
Eighth Edition (2011) in order to perform as per inter-
national ethical standards. The used animals are usually
mice (Abdella et al., 2009; Tohamy et al., 2014).
19.2.3.3. Preparation of BP extracts. The preparation
of BP extract can be according with Orsoli
c et al.
(2005), Yamaguchi et al. (2007), Abdella et al. (2009)
and Tohamy et al. (2014).
Procedure
1. BP (280mg) was suspended in 10ml of distilled water.
2. Mix vigorously.
3. Keep the suspension overnight in dark.
4. Centrifuge at 10,000 rpm in a cooling centrifuge for
45minutes at 10 C.
5. Collect and filtrate the supernatant fraction.
6. Keep the filtrated in freeze condition at 20 C
until use.
19.3. Concluding remarks on anti-mutagenic effects
of BP
Studies published by Pinto et al. (2010) suggest that some
BP preparations could be employed as future chemopro-
tective and/or chemopreventive agents, especially for mam-
mary or other estrogen-dependent cancers because in
their studies found that at least one preparation of both
species showed a marked anti-genotoxic effect against the
action of tested anticancer drug. These anti-mutagenic
properties of BP have been attributed to their composition
rich in phenolic compounds that justify the higher protec-
tion (Pinto et al., 2010). This author also reported that the
most effective species of BP from Salix alba in inducing
inhibitory activity of E2 (17b-estradiol) and protected
against the three mutagenic drugs and these results could
be due its higher flavonoid content. Some of two studied
samples were rich in luteolin, quercetin, chrysin, and
kaempferol compounds that have been previously demon-
strated to be antiestrogenic (Oh & Chung, 2006; Pinto
et al., 2008; van Meeuwen et al., 2007), also exhibit anti-
mutagenic and genotoxic activity (Lautraite et al., 2002;
Nagy et al., 2009) whereas Galangin, pinocembrin and
chrysin, as well as isorhamnetin 3-O-neohesperidoside, a
derivative of isorhamnetin were proven to effectively con-
trast the DNA-damaging activity of several mutagens/carci-
nogens without exhibiting any genotoxic activity (Heo
et al., 2001; Snyder & Gillies, 2002). Similar studies revealed
that phenolic and flavonoids compounds as luteolin,
chrysin, quercetin and taxifolin showed anti-mutagenic and
genotoxic activity (Benkovic et al., 2008; Dutta et al., 2007;
Makena et al., 2009; Uhl et al., 2003). Therefore, the anti-
mutagenic effects, can be affected by total phenolic com-
pound content as well as variation in their composition,
and the higher phenolic compound content found in S. alba
would justify the higher protection elicited toward the
mutagenic drugs (Pinto et al., 2010).
The specific mechanism of protection of the BP extract
is reported to involve the scavenging of potentially toxic
and mutagenic electrophiles and free radicals and the modi-
fication of antioxidant pathways (Ohta et al., 2007). By
another side, Nagy et al. (2009) referred that the anti-
mutagenic effect of most phenolic compounds could be
clearly ascribed to their -DPPH
 (1,1-diphenyl-2-picrylhy-
drazyl) scavenging activity, substitution patterns and lipo-
philicity. Barzin et al. (2011) in their studies evaluated the
anti-mutagenic effects of pollen grams of Phoenix dactylifera
using Ames assay and found that had 46% anti-mutagenic
response. According to Ames et al. (1973, 1976) when the
percent (46% of anti-mutagenic response), is more than 40,
mutagenesis effect of the test sample is strong. Also,
Bariliak et al. (1996) observed the ability of apiculture prod-
ucts to decrease the mutagenic effects of some chemical
and physical mutagens. In general way, the bioactivity of the
pollens has been reported to be due to presence of com-
pounds such as proteins, carbohydrates, amino acids, lipids,
vitamins carotenoids, steroids, terpenes, flavonoids and
polyphenols (Leja et al., 2007; Kao et al., 2020; Kroyer &
Hegedus, 2001).
Acknowledgements
The COLOSS (Prevention of honey bee COlony LOSSes)
Association aims to explain and prevent massive honey bee
colony losses. It was funded through the COST Action
FA0803. COST (European Cooperation in Science and
Technology) is a unique means for European researchers to
jointly develop their own ideas and new initiatives across all
scientific disciplines through trans-European networking of
nationally funded research activities. Based on a pan-European
intergovernmental framework for cooperation in science and
technology, COST has contributed since its creation more
than 40 years ago to closing the gap between science, policy
makers and society throughout Europe and beyond. COST is
supported by the EU Seventh Framework Program for
research, technological development and demonstration activ-
ities (Official Journal L 412, 30 December 2006). The European
Science Foundation as implementing agent of COST provides
the COST Office through an EC Grant Agreement. The
Council of the European Union provides the COST
Secretariat. The COLOSS network is now supported by the
Ricola Foundation – Nature & Culture. Figures 26–28 are
reproduced from Sawyer (1981) with the permission of the
publishers University College Cardiff Press and Northern Bee
Standard methods for pollen research 95
Books. Lidia Barreto and J Nordi wish to thank the Apiculture
Research Center of Taubate University (UNITAU-SP/Brazil)
and Agriculture Secretary of Bahia State (SEAGRI-BA/
BRAZIL). Maria Campos wishes to thank (UI0204): UIDB/
00313/2020, Center of Chemistry from Faculty of Sciences
and Technology of University of Coimbra, Portugal. Of
elia
Anjos wishes to thank to Forest Research Centre, a research
unit funded by Fundac¸~ao para a Cie^ncia e a Tecnologia I.P.
(FCT), Portugal (UIDB/00239/2020), and to Centro de
Biotecnologia de Plantas da Beira Interior for the OPUS soft-
ware availability. Norma Almaraz Abarca thanks to the
Instituto Politecnico Nacional for financial and logistic support.
Manuel Chica and Pascual Campoy wish to thank the
APIFRESH project. APIFRESH has been co-funded by the
European Commission under the R4SMEs 7th Framework
Program. Olena Lokutova thanks the Austrian Institute of bee-
keeping, Dr H Pehhacker and the same members of the
Institute H Hagel and E H€uttinger for conducting photo-
microscopic studies and pollen analysis Ukrainian samples of
pollen loads, which were the basis of the atlas of pollen
“honey plants” of Ukraine. Olena is grateful also to Polish col-
leagues Z Warakomska (Department of Botany University of
Lublin) and D Teper (Polish Institute of beekeeping) for their
professionalism and consultations to determine the botanical
origin of some Ukrainian honey. Also thanks to their scientific
advisers’ academician G Bogdanov (National Academy of
Agrarian Sciences of Ukraine), Prof. V Polishuk (Department
of beekeeping National University of Life and Environmental
Sciences of Ukraine) and O Martynyuk (M.G. Kholodny
Institute of Botany, Kiev, Ukraine) for his helpful co-operation
in the field of beekeeping and palinology. Janka Nozkova
wishes to thank the Operational Program Research and
Development of the European Regional Development Fund in
the frame of the project “Support of technologies innovation
for special bio-food products for human healthy nutrition”
(ITMS 26220220115) and also by the Excellence Center for
Agrobiodiversity Conservation and Benefit – project imple-
mented under the Operational Program Research and
Development financed by European Fund for Regional
Development ITMS 26220120015 (Slovak Republic) and all
colleagues from Institute of Biodiversity and Biosafety, Slovak
University of Agriculture in Nitra for their help with image
analysis. Ananias Pascoal, Georgina Tolentino and Let
ıcia
Estevinho would like to thank Fundac¸~ao para a Cie^ncia e
Tecnologia (FCT), Programa Operacional Pontencial Humano
(POPH) and European Union (EU) for his Postdoctoral grant
SFRH/BPD/91380/2012. Wiebke Sickel, Markus Ankenbrand,
Gudrun Grimmer, Frank F€orster, Ingolf Steffan-Dewenter and
Alexander Keller thank the financial support by the DFG
Collaborative Research Center 1047, Insect Timing. MJA was
further supported by a grant of the German Excellence
Initiative to the Graduate School of Life Sciences of the
University of W€urzburg. They are grateful to the members of
the Departments of Animal Ecology and Tropical Biology;
Bioinformatics; and Human Genetics, University of W€urzburg,
for constructive input on the design of the workflow.
Additionally thank to the Department of Human Genetics,
especially S. Rost, for granting access to the Illumina MiSeq
device. Zivoslav Tesi
c, Mirjana Mosi
c, Aleksandar Kosti
c,
Mirjana Pesi
c, Dusanka Milojkovi
c-Opsenica thank the Ministry
of Education, Science and Technological Development of
Serbia, Grants 172017 and TR 31069. Gina Tolentino would
like to thank the Mountain Research Center (CIMO),
Agricultural College of Braganc¸a, Polytechnic Institute of
Braganc¸a for his research grant in the project titled "
Development of new bee products in biological produc-
tion way.”
Disclosure statement
No potential conflict of interest was reported by the authors.
ORCID
Maria G. Campos http://orcid.org/0000-0003-1012-6240
Of
elia Anjos http://orcid.org/0000-0003-0267-3252
Manuel Chica http://orcid.org/0000-0002-4717-1056
Pascual Campoy http://orcid.org/0000-0002-9894-2009
Norma Almaraz-Abarca http://orcid.org/0000-0003-
1603-4865
Leticia M. Estevinho http://orcid.org/0000-0002-9249-1948
Ananias Pascoal http://orcid.org/0000-0003-3823-0402
Luis G. Dias http://orcid.org/0000-0002-1210-4259
Zivoslav L. j. Tesi
c http://orcid.org/0000-0002-5162-3123
Aleksandar Z. Kosti
c http://orcid.org/0000-0002-1012-4029
Mirjana B. Pesi
c http://orcid.org/0000-0001-9779-1686
Dusanka M. Milojkovi
c-Opsenica http://orcid.org/0000-0001-
6274-4222
Wiebke Sickel http://orcid.org/0000-0002-0038-1478
Markus J. Ankenbrand http://orcid.org/0000-0002-
6620-807X
Ingolf Steffan-Dewenter http://orcid.org/0000-0003-
1359-3944
Alexander Keller http://orcid.org/0000-0001-5716-3634
Frank F€orster http://orcid.org/0000-0003-4166-5423
Chrysoula H. Tananaki http://orcid.org/0000-0002-
0056-483X
Vasilios Liolios http://orcid.org/0000-0001-7353-2364
Dimitrios Kanelis http://orcid.org/0000-0001-7736-4783
Maria-Anna Rodopoulou http://orcid.org/0000-0003-
1643-4615
Andreas Thrasyvoulou http://orcid.org/0000-0001-6731-4184
Christina Kast http://orcid.org/0000-0002-5673-6848
Ligia Bicudo de Almeida-Muradian http://orcid.org/0000-
0001-5322-1358
Teresa SzczeR sna http://orcid.org/0000-0002-0161-0287
Norman L. Carreck http://orcid.org/0000-0001-7779-9736
References
Abdella, E. M., Tohamy, A., & Ahmad, R.R. (2009).
Antimutagenic activity of Egyptian propolis and bee pollen
water extracts against cisplatin-induced chromosomal
abnormalities in bone marrow cells of mice. Iranian Journal
of Cancer Prevention, 2(4), 175–181.
Abouda, Z., Zerdani, I., Kalalou, I., Faid, M., & Ahami, M. T.
(2011). The antibacterial activity of Moroccan bee bread
and bee-pollen (fresh and dried) against pathogenic bacteria.
Research Journal of Microbiology, 6(4), 376. http://.dx.doi.org/
10.3923/jm.2011.376.384
Akkol, E. K., Orhan, D. D., Gurbuz, I., & Yesilada, E. (2010).
In vivo activity assessment of a "honey-bee pollen mix" for-
mulation. Pharmaceutical Biology, 48(3), 253–259. https://doi.
org/10.3109/13880200903085482
Allen, G. (2006). An automated pollen recognition system
[Master’s thesis]. Institute of information Sciences and
Technology, Massey University.
Almaraz-Abarca, N., Campos, M. d. G., 
Avila-Reyes, J. A.,
Naranjo-Jimenez, N., Herrera-Corral, J., & Gonzalez,
V. L. S. (2004). Variability of antioxidant activity among
honey bee-collected pollen of different botanical origin.
Interciencia, 29(10), 574–578. doi: 0378-1844/04/10/574-05.
Almaraz-Abarca, N., Campos, M. D. G., Delgado-Alvarado, A.,

Avila-Reyes, J. A., Naranjo-Jim
enez, J. A., & Herrera Corral,
H. (2006). Variaci
on intrapoblacional en el Perfil Fenolico
96 M. G. Campos et al.
del polen de Stenocactus multicostatus subsp. zacatecasensis
(Cactaceae). Bolet
ın Nakari, 17(3), 59–64.
Almaraz-Abarca, N., Campos, M. G., 
Avila-Reyes, J. A.,
Naranjo-Jim
enez, N., Herrera Corral, J., & Gonz
alez-Valdez,
L. S. (2007). Antioxidant activity of polyphenolic extract of
monofloral honey bee – collected pollen from mesquite
(Prosopis juliflora, Leguminosae). Journal of Food Composition
and Analysis, 20(2), 119–124. https://doi.org/10.1016/j.jfca.
2006.08.001
Almaraz-Abarca, N., Campos, M. G., Delgado-Alvarado, A.,

Avila-Reyes, J. A., Naranjo-Jim
enez, N., Herrera Corral, J.,
Tomatas, A. F., Almeida, A. J., & Vieira, A. (2007). Fenoles
del polen de Stenocactus, Echinocereus y Mammillaria
(Cactaceae). Polibotanica, 23, 37–55. http://www.herbario.
encb.ipn.mx/pb/pdf/pb23/3Fenoles.pdf
Almaraz-Abarca, N., Campos, M. G., Delgado-Alvarado, E. A.,

Avila-Reyes, J. A., Herrera-Corral, J., Gonz
alez-Valdez, L. S.,
Naranjo-Jim
enez, N., Frigerio, C., Tomatas, A. F., Almeida,
A. J., Vieira, A., & Uribe-Soto, J. N. (2008). Pollen flavonoid/
phenolic acid composition of four species of Cactaceae and
its taxonomic significance. American Journal of Agricultural
and Biological Science, 3(3), 534–543. https://doi.org/10.3844/
ajabssp.2008.534.543
Almaraz-Abarca, N., Gonz
alez-Elizondo, M. S., Campos, M. G.,

Avila-Sevilla, Z. E., Delgado-Alvarado, E. A., & 
Avila, R. J. A.
(2013). Variability of the foliar phenol profiles of Agave vic-
toriae-reginae complex (Agavaceae). Botanical Sciences, 91(3),
295–306. https://doi.org/10.17129/botsci.9
Almaraz-Abarca, N., Rivera-Rodr
ıguez, D. M., Arr
aez-Rom
an,
D., Segura-Carretero, A., S
anchez-Gonz
alez, J. J., Delgado-
Alvarado, A., 
Avila-Reyes, J. A. (2013). Los fenoles del
polen del g
enero Zea. Acta Botanica Mexicana, 105, 59–85.
http://www.ibiologia.unam.mx/sociedad/www/pdf/BSBM91_
3/Almaraz-Abarca_91_3_295-306.pdf
Almeida-Muradian, L. B., Arruda, V. A. S., & Barreto, L. M. R.
C. (2012). Manual de Controle de Qualidade do P
olen Ap
ıcola
(p. 28). APACAME.
Almeida-Muradian, L. B., Pamplona, L. C., Coimbra, S., &
Barth, O. M. (2005). Chemical composition and botanical
evaluation of dried bee pollen pellets. Journal of Food
Composition and Analysis, 18(1), 105–111. https://doi.org/10.
1016/j.jfca.2003.10.008
Aly, M. S., Ashour, M. B., EL-Nahas, S. M., & Abo Zeid,
M. A. F. (2003). Genotoxicity and cytotoxicity of the anti-
cancer drugs gemcitabine and cisplatin, separately and in
comination: in vivo studies. Journal of Biological Sciences,
3(11), 961–972. https://doi.org/10.3923/jbs.2003.961.
972[Mismatch] .
Ambruster, L., & Oenike, G. (1929). Die pollenformen als mittel
zur honig herkunftsbestimmung (p. 116). Wachholtz.
Ames, B. N., Durston, W. E., Yamasaki, E., & Lee, F. D.
(1973). Carcinogens are mutagens: A simple test system
combining liver homogenates for activation and bacteria for
detection. Proceedings of the National Academy of Sciences of
the United States of America, 70(8), 2281–2285. https://doi.
org/10.1128/AAC.04918-14.
Andrada, A. C., & Telleria, M. C. (2005). Pollen collected by
honey bees (Apis mellifera L.) from south of Cald
en district
(Argentina): Botanical origin and protein content. Grana, 44(2),
115–122. https://doi.org/10.1080/00173130510010459
Anjos, O., Campos, M. G., Ruiz, P. C., & Antunes, P. (2015).
Application of FTIR-ATR spectroscopy to the quantification
of sugar in honey. Food Chemistry, 169, 218–223. https://doi.
org/10.1016/j.foodchem.2014.07.138
Anjos, O., Santos, A. J. A., Dias, T., Estevinho, L. M. (2017).
Application of FTIR-ATR Spectroscopy on the bee pollen
characterization. Journal of Apicultural Research, 56(3),
210–218. https://doi.org/10.1080/00218839.2017.1289657
Anjos, O., Santos, A. F. A., Paix~ao, V., Estevinho, M. L. (2018).
Physicochemical Parameters of Lavandula spp. honey
accessed by FT-RAMAN. Talanta, 178, 43–48. https://doi.
org/10.1016/j.talanta.2017.08.099
Ankenbrand, M. J. (2015). The ITS2 Database V – Twice as
much. Molecular Biology and Evolution, 32, 3030–3032.
AOAC. (1995). Official methods of analysis (16th Ed.).
Association of Official Analytical Chemists.
Arroyo, J., Devesa, J. A., Herrera, P., Ortiz, P. L., & Lozano, S. T.
(1986). Resumen del proyecto de investigaci
on: La flora
melit
ofila en Andaluc
ıa occidental. Vida Ap
ıcola, 18, 33–39.
Arruda, V. A. S. (2013). Dried bee pollen: Physicochemical, micro-
biological quality, phenolic and flavonoid compounds, antioxidant
activity and botanical origin [PhD thesis]. Pharmaceutical
Science School, University of S~ao Paulo.
Arruda, V. A. S., Freitas, A. S., B., O. M., Estevinho, L. M., &
Almeida-Muradian, L. B. (2013). Propriedades biol
ogicas do
p
olen ap
ıcola de coqueiro, coletado na Regi~ao Nordeste do
Brasil. Magistra, 25, 1–6.
Arruda, V. A. S., Santos Pereira, A. A., de Freitas, A. S., Barth,
O. M., & de Almeida-Muradian, L. B. (2013b). Dried bee
pollen: B complex vitamins, physicochemical and botanical
composition. Journal of Food Composition and Analysis, 29(2),
100–105. https://doi.org/10.1016/j.jfca.2012.11.004
Arruda, V. A. S., Santos Pereira, A. A., Estevinho, L. M., & de
Almeida-Muradian, L. B. (2013a). Presence and stability of B
complex vitamins in bee pollen using different storage con-
ditions. Food and Chemical Toxicology, 51, 143–148. https://
doi.org/10.1016/j.fct.2012.09.019
Avula, B., Satyanarayanaraju, S., Wang, Y., Zweigenbaum, J.,
Wang, M., & Khan, I. A. (2015). Characterization and
screening of pyrrolizidine alkaloids and N-oxides from bota-
nicals and dietary supplements using UHPLC-high resolution
mass spectrometry. Food Chemistry, 178, 136–148. https://
doi.org/10.1016/j.foodchem.2015.01.053
Balkanska, R. G., Ignatova, M. M. (2012). Chemical composition of
multifloral bee pollen from Bulgaria. In Proceedings of 6th Central
European Congress on Food, CE Food 2012 (pp. 375–378).
Ball, G. (1998). Bioavailability and analysis of vitamins in foods.
Chapman & Hall.
Baltrusaityt _e, V., Venskutonis, P. R., & Ceksteryt _e, V. (2007b).
Radical scavenging activity of different floral origin honey
and beebread phenolic extracts. Food Chemistry, 101(2),
502–514. https://doi.org/10.1016/j.foodchem.2006.02.007
Barbosa, S. I. T. R., Silvestre, A. J. D., Sim~oes, M. M. Q., &
Estevinho, M. L. (2006). Composition and antibacterial
activity of the lipophilic fraction of honey bee pollen from
native species of Montesinho Natural Park. International
Journal of Agricultural Research, 1(5), 471–749.
Bariliak, I. R., Berdyshev, G. D., & Dugan, A. M. (1996). The
antimutagenic action of apiculture products. The Journal of
Cytology and Genetics, 30, 48–55.
Barreto, L. M. R. C. (2004). P
olen Ap
ıcola: beneficiamento,
armazenamento e legislac¸~ao. Botucatu, 2004, 150 (Tese
Doutorado) – Faculdade de Medicina Veterin
aria e
Zootecnia – Universidade Estadual Paulista.
Barreto, L. M. R. C., Dib, A. P. S., & Pe~ao, G. F. R. (2006).
Higienizac¸~ao e Sanitizac¸~ao da Produc¸~ao Ap
ıcola (p. 150).
Cabral Editora e Livraria Universit
aria.
Barreto, L. M. R. C., Funari, S. R. C., Orsi, R. O., & Dib,
A. P. S. (2006). Produc¸~ao de P
olen no Brasil (p. 100). Cabral
Editora e Livraria Universit
aria.
Barriada-Bernal, L. G., Almaraz-Abarca, N., Delgado-Alvarado,
E. A., Gallardo-Vel
azquez, T., 
Avila-Reyes, J. A., Torres-Mor
an,
M. I., Gonz
alez-Elizondo, M. d. S., & Herrera-Arrieta, Y. (2014).
Flavonoid composition and antioxidant capacity of the edible
flowers of Agave durangensis (Agavaceae). Cyta – Journal of Food,
12(2), 105–114. https://doi.org/10.1080/19476337.2013.801037
Standard methods for pollen research 97
Barry, A. L., & Thornsberry, C. (1991). Susceptibility tests:
Diffusion test procedures. In A. Balows, W. J. Hauser, K. L.
Hermann, H. D. Isenberg, & H. J. Shamody (Eds.),Manual of clin-
ical microbiology (Vol. 5, pp. 1117–1125). American Society for
Microbiology.
Barth, O. M. (2004). Melissopalynology in Brazil: A review of
pollen analysis of honeys, propolis and pollen loads of bees.
Scientia Agricola, 61(3), 342–350. https://doi.org/10.1590/
S0103-90162004000300018
Barth, O. M., Freitas, A. S., Oliveira, E. S., Silva, R. A.,
Maester, F. M., Andrella, R. R., & Cardozo, G. M. (2010).
Evaluation of the botanical origin of commercial dry bee
pollen load batches using pollen analysis: A proposal for
technical standardization. Anais da Academia Brasileira de
Ciencias, 82(4), 893–902. https://doi.org/10.1590/s0001-
37652010000400011
Barzin, G., Entezari, M., Hashemi, M., Hajiali, S., Ghafoori., &
Gholami, M. (2011). Survey of Antimutagenicity and
Anticancer effect of Phoenix dactylifera pollen grains.
Advances in Environmental Biology, 5(12), 3716–3718.
Basim, E., Basim, H., & Ozcan, M. (2006). Antibacterial activ-
ities of Turkish pollen and propolis extracts against plant
bacterial pathogens. Journal of Food Engineering, 77(4),
992–996. https://doi.org/10.1016/j.jfoodeng.2005.08.027
Bassi, E. A. (2000). Programa de qualidade total na produc¸~ao
de mel. In: Congresso brasileiro de apicultura, 13, 2000.
Sonopress. 1CD.
Basuny, A. M., Arafat, S. M., & Soliman, H. M. (2013).
Chemical analysis of olive and palm pollen: Antioxidant and
antimicrobial activation properties. Herald Journal of
Agriculture and Food Science Research, 2(3), 091–097.
Bell, K. L., Burgess, K. S., Botsch, J. C., Dobbs, E. K., Read, T.
D., & Brosi, B. J. (2019). Quantitative and qualitative assess-
ment of pollen DNA metabarcoding using constructed spe-
cies mixtures. Molecular Ecology, 28, 431–455. https://doi.
org/10.1111/mec.14840.
Benkovic, V., Knezevic, A. H., Dikic, D., Lisicic, D., Orsolic, N.,
Basic, I., Kosalec, I., & Kopjar, N. (2008). Radioprotective
effects of propolis and quercetin in gamma-irradiated mice
evaluated by the alkaline comet assay. Phytomedicine, 15(10),
851–858. https://doi.org/10.1016/j.phymed.2008.02.010
Bertoncelj, J., Polak, T., Kropf, U., Korosec, M., & Golob, T.
(2011). LC-DAD-ESI/MS analysis of flavonoids and abscisic
acid with chemometric approach for the classification of
Slovenian honey. Food Chemistry, 127(1), 296–302. https://
doi.org/10.1016/j.foodchem.2011.01.003
Bessonov, V. N. (1992). Status of pollen as an indicator of pol-
lution by heavy metals. Ecology. – Ekaterinburg, 1992(3),
45–50. (Article in Russian).
Betteridge, K., Cao, Y., & Colegate, S. M. (2005). Improved
method for extraction and LC-MS analysis of pyrrolizidine alka-
loids and their N-oxides in honey: Application to Echium vulgare
honeys. Journal of Agricultural and Food Chemistry, 53(6),
1894–1902. https://doi.org/10.1021/jf0480952
Biesaga, M., & Pyrzynska, K. (2009). Liquid chromatography/
tandem mass spectrometry studies of the phenolic com-
pounds in honey. Journal of Chromatography. A, 1216(38),
6620–6626. https://doi.org/10.1016/j.chroma.2009.07.066
Biodiversity International. (2007). Guidelines for the development
of crop descriptor lists. Biodiversity Technical Bulletin Series (xii
þ 72 pp.). Biodiversity International.
Bobis, O., Marghitas, L., Al, Dezmirean, D., Morar, O., Bonta,
V., & Chirila, F. (2010). Quality parameters and nutritional
value of different commercial bee products. Bulletin UASVM
Animal Science and Biotechnologies, 67(1-2), 91–96.
Bogdanov, S. (2015). Pollen: Production, nutrition and health: A
review. Bee Product Science. www.bee-hexagon.net
Bogdanov, S., Martin, P., & L€ullmann, C. (1997). Harmonized
methods of the European Honey Commission.
Determination of sugars by HPLC. Apidologie, 28, 42–44.
Bondet, V., Brand-Williams, W., & Berset, C. (1997). Kinetics
and mechanisms of antioxidant activity using the DPPH

free radical method. LWT – Food Science and Technology,
30(6), 609–615. https://doi.org/10.1006/fstl.1997.0240
Boppr
e, M. (2011). The ecological context of pyrrolizidine
alkaloids in food, feed and forage: An overview. Food
Additives & Contaminants. Part A, Chemistry, Analysis, Control,
Exposure & Risk Assessment, 28(3), 260–281. https://doi.org/
10.1080/19440049.2011.555085
Boppr
e, M., Colegate, S. M., & Edgar, J. A. (2005).
Pyrrolizidine alkaloids of Echium vulgare honey found in
pure pollen. Journal of Agricultural and Food Chemistry, 53(3),
594–600. https://doi.org/10.1021/jf0484531
Boucher, A., Hidalgo, P. J., Thonnat, M., Belmonte, J., Galan,
C., Bonton, P., & Tomczak, R. (2002). Development of a
semi-automatic system for pollen recognition. Aerobiologia,
18(3/4), 195–201. https://doi.org/10.1023/A:1021322813565
Boudet, A. M. (2007). Evolution and current status of research
in phenolic compounds. Phytochemistry, 68(22-24),
2722–2735. https://doi.org/10.1016/j.phytochem.2007.06.012
Brasil. (1997). Minist
erio da Agricultura e do Abastecimento.
Portaria no 368, de 4 de setembro de 1997. Regulamento
T
ecnico sobre as Condic¸~oes Higie^nico-Sanit
arias e de Boas
Pr
aticas de Elaborac¸~ao para Estabelecimentos Elaboradores/
Industrializadores de Alimentos. Di
ario Oficial da Uni~ao da
Republica Federativa do Brasil, Bras
ılia.
Brasil. (2001). Instruc¸~ao Normativa do Minist
erio da
Agricultura n 3, de 19 de janeiro de 2001. (DOU – 23/01/
01, Secc¸~ao I, pg. 18-23) Regulamentos t
ecnicos de identi-
dade e qualidade de apitoxina, cera de abelhas, gel
eia real,
gel
eia real Liofilizada, p
olen ap
ıcola, pr
opolis e extrato de
pr
opolis. Di
ario Of
ıcial (DO), Bras
ılia, DF.
Brindza, J., & Brovarskyi, V. (2014). Pollen and bee pollen of
some plant species (p. 138). Kyiv: Korsunskiy vidavnichiy dim
Vsesvit.
Bruni, I., Galimberti, A., Caridi, L., Scaccabarozzi, D., De
Mattia, F., Casiraghi, M., & Labra, M. (2015). A DNA bar-
coding approach to identify plant species in multiflower
honey. Food Chemistry, 170, 308–315. https://doi.org/10.
1016/j.foodchem.2014.08.060
Bucher, E., Rjfler, V., Vorwoht, G., & Zieger, E. (2004). Das
pollebild der sudtiroler honige (p. 676). Biologischer labor.
Burmistrov, N., & Nikitina, V. A. (1990). Honey plants and their
pollen (p. 192). Rosagropromizdat. (Book on Russian).
Cabrera, C., & Montenegro, G. (2013). Pathogen control using
a natural Chilean bee pollen extract of known botanical ori-
gin. Ciencia e Investigaci
on Agraria, 40(1), 223–230. https://
doi.org/10.4067/S0718-16202013000100020
Campos, M. G. (1997). Caracterizac¸~ao do p
olen ap
ıcola pelo seu
perfil em compostos fen
olicos e pesquisa de algumas actividades
biologicas [PhD thesis]. Faculty of Pharmacy, University of
Coimbra.
Campos, M. G., Bogdanov, S., Almeida-Muradian, L. B.,
Szczesna, T., Mancebo, Y., Frigerio, C., & Ferreira, F.
(2008). Pollen composition and standardization of analytical
methods. Journal of Apicultural Research, 47(2), 154–161.
https://doi.org/10.3896/IBRA.1.47.2.12[10.1080/00218839.
2008.11101443]
Campos, M. G., Frigerio, C., Lopes, J., & Bogdanov, S. (2010).
What is the future of bee pollen? Journal of ApiProduct and
ApiMedical Science, 2(4), 131–144. https://doi.org/10.3896/
IBRA.4.02.4.01
Campos, M. G., & Markham, K. R. (2007). Structure information
from HPLC and on-line measured absorption spectra: Flavones,
flavonols and phenolic acids. Imprensa da Universidade de
Coimbra. https://doi.org/10.14195/978-989-26-0480-0
98 M. G. Campos et al.
Campos, M. G., Markham, K. R., Mitchell, K. A, & da Cunha,
A. P. d. (1997). An approach to the characterization of bee
pollens via their flavonoid/phenolic profiles. Phytochemical
Analysis, 8(4), 181–185. https://doi.org/10.1002/(SICI)1099-
1565(199707)8:4< 181::AID-PCA359> 3.0.CO;2-A
Campos, M. G., Markham, K., & Proenc¸a da Cunha, A. (1996).
Quality assessment of bee-pollens using flavonoid/phenolic
profiles. Bull. Groupe Polyphenols, 18, 54–55.
Campos, M. G., Mitchel, K., Cunha, A., & Markham, K. R.
(1997). A systematic approach to the characterization of
bee pollens via their flavonoid/phenolic profiles.
Phytochemical Analysis, 8(4), 181–185. 2-A https://doi.org/10.
1002/(SICI)1099-1565(199707)8:4< 181::AID-PCA359> 3.
0.CO,[10.1002/(SICI)1099-1565(199707)8:4< 181::AID-
PCA359> 3.0.CO;2-A]
Campos, M. G., Webby, R. F., Markham, K. R., Mitchell, K. A.,
& da Cunha, A. P. (2003). Age-induced diminution of free
radical scavenging capacity in bee pollens and the contribu-
tion of constituent flavonoids. Journal of Agricultural and Food
Chemistry, 51(3), 742–745. https://doi.org/10.1021/jf0206466
Caroli, S., Forte, G., Iamiceli, A. L., & Galoppi, B. (1999).
Determination of essential and potentially toxic trace ele-
ments in honey by inductively coupled plasma-based techni-
ques. Talanta, 50(2), 327–336. https://doi.org/10.1016/
S0039-9140(99)00025-9
Carpes, S. T., Begnini, R., DE Alencar, S. M., & Masson, M. L.
(2007). Study of preparations of bee pollen extracts, anti-
oxidant and antibacterial activity. Ci^encia e Agrotecnologia,
31(6), 1818–1825. https://doi.org/10.1590/S1413-
70542007000600032
Carpes, S. T., Prado, A. P., Moreno, I. A. M., Mour~ao, G. B.,
DE Alencar, S. M., & Masson, M. L. (2008). Avaliac¸~ao do
potencial antioxidante do p
olen ap
ıcola producido na regi~ao
sul do Brasil. Qu
ımica Nova, 31(7), 1660–1664. https://doi.
org/10.1590/S0100-40422008000700011
Carri
on, P., Cernadas, E., G
alvez, J. F., & D
ıaz-Losada, E.
(2003). Determine the composition of honey bee pollen by
texture classification. In F. J. Perales (Ed.), IbPRIA 2003,
LNCS 2652 (pp. 158–167). https://doi.org/10.1007/978-3-
540-44871-6_19
Carvell, C., Westrich, P., Meek, W. R., Pywell, R. F., &
Nowakowski, M. (2006). Assessing the value of annual and
perennial forage mixtures for bumble bees by direct obser-
vation and pollen analysis. Apidologie, 37(3), 326–340.
https://doi.org/10.1051/apido:2006002
Cavalieri, S. J., Harbeck, R. J., Mccarter, Y. S., Ortez, J. H.,
Rankin, I. D., Sautter, R. L., Sharp, S. E., & Spiegel, C. A.
(2005). Manual of antimicrobial susceptibility testing. In
M. B. Coyle (Ed.), American Society for Microbiology # 2005
(pp. 1–236). American Society for Microbiology.
Chandola, V., Banerjee, A., & Kumar, V. (2009). Anomaly
detection: A survey. ACM Computing Surveys, 41(3), 1–15.
58. https://doi.org/10.1145/1541880.1541882
Chantarudee, A., Phuwapraisirian, P., Kimura, K., Okuyama,
M., Mori, H., Kimura, A., & Chanchao, C. (2012). Chemical
constituents and free radical scavenging activity of corn pol-
len collected from Apis mellifera hives compared to floral
corn pollen at Nam, Thailand. BMC Complementary &
Alternative Medicine, 12, 45.
Chaturvedi, M. (1973). An analysis of honey bee pollen loads
from Banthra, Lucknow, India. Grana, 13(3), 139–144.
https://doi.org/10.1080/00173137309429890
Chen, S., Yao, H., Han, J., Liu, C., Song, J., Shi, L., Zhu, Y., Ma, X.,
Gao, T., Pang, X., Luo, K., Li, Y., Li, X., Jia, X., Lin, Y., & Leon,
C. (2010). Validation of the ITS2 region as a novel DNA bar-
code for identifying medicinal plant species. PLoS One, 5(1),
e8613. https://doi.org/10.1371/journal.pone.0008613
Chica, M., & Campoy, P. (2012). Discernment of bee pollen
loads using computer vision and one-class classification
techniques. Journal of Food Engineering, 112(1-2), 50–59.
https://doi.org/10.1016/j.jfoodeng.2012.03.028
Choi, E. (2007). Antinociceptive and antiinflammatory activities
of pine (Pinus densiflora) pollen extract. Phytotherapy
Research, 21(5), 471–475. https://doi.org/10.1002/ptr.2103
Chudzinska, M., & Baralkiewicz, D. (2011). Application of ICP-
MS method of determination of 15 elements in honey with
chemometric approach for the verification of their authenti-
city. Food and Chemical Toxicology, 49(11), 2741–2749.
https://doi.org/10.1016/j.fct.2011.08.014
Cirnu, I., Slusanschi, N., Maronescu, R., Filipescu, H., & Grosu,
E. (1969). Compozitia chimica a polenului de porumb (Zea
mays L.) si floarea-soarelui (Helianthus annuus L.) recoltat la
diferite epoci, in cardul amestecului furajer-melifer.
Apicultura, 22(8), 22–25.
Clinical and Laboratory Standards Institute (CLSI). (2009).
Performance standards for antimicrobial disk susceptibility
tests; approved standard-tenth edition M02-A10. National
Committee for Clinical Laboratory Standards, 29. 76 pp
Clinical and Laboratory Standards Institute (CLSI). (2012).
Methods for dilution antimicrobial susceptibility tests for
bacteria that grow aerobically; approved standard – Ninth
Edition M07 – A9. National Committee for Clinical Laboratory
Standards, 32(2), 1–64.
Comaniciu, D., & Meer, P. (2002). Mean shift: A robust
approach toward feature space analysis. IEEE Transactions on
Pattern Analysis and Machine Intelligence, 24(5), 603–619.
https://doi.org/10.1109/34.1000236
Costa, C. T., Dalluge, J. J., Welch, M. J., Coxon, B., Margolis,
S. A., & Horton, D. (2000). Characterization of prenylated
xanthones and flavanones by liquid chromatography/atmos-
pheric pressure chemical ionization mass spectrometry.
Journal of Mass Spectrometry, 35(4), 540–549. https://doi.org/
10.1002/(SICI)1096-9888(200004)35:4< 540::
AID-JMS966> 3.0.CO;2-Y
Crane, E. E. (1990). Bees and beekeeping: Science, practice and
world resources (p. 593). Cornstock Publ.
Crews, C., Berthiller, F., & Krska, R. (2010). Update on analyt-
ical methods for toxic pyrrolizidine alkaloids. Analytical and
Bioanalytical Chemistry, 396(1), 327–338. https://doi.org/10.
1007/s00216-009-3092-2
Crews, C., Driffield, M., Berthiller, F., & Krska, R. (2009). Loss
of pyrrolizidine alkaloids on decomposition of ragwort
(Senecio jacobaea) as measured by LC-TOF-MS. Journal of
Agricultural and Food Chemistry, 57(9), 3669–3673. https://
doi.org/10.1021/jf900226c
Cuyckens, F., & Claeys, M. (2002). Optimization of a liquid
chromatography method based on simultaneous electro-
spray ionization mass spectrometric and ultraviolet photo-
diode array detection for analysis of flavonoid glycosides.
Rapid Communications in Mass Spectrometry, 16(24),
2341–2348. https://doi.org/10.1002/rcm.861
Cuyckens, F., & Claeys, M. (2004). Mass spectrometry in the
structural analysis of flavonoids. Journal of Mass
Spectrometry, 39(1), 1–15. https://doi.org/10.1002/jms.585
Danner, N., H€artel, S., & Steffan-Dewenter, I. (2014). Maize
pollen foraging by honey bees in relation to crop area and
landscape context. Basic and Applied Ecology, 15(8),
677–684. https://doi.org/10.1016/j.baae.2014.08.010
Dasarathy, B. (1991). Nearest Neighbor (NN) norms: Nearest neigh-
bor pattern classification techniques. IEEE Computational Society.
de Rijke, E., Out, P., Niessen, W. M. A., Ariese, F., Gooijer,
C., & Brinkman, U. A. (2006). Analytical separation and
detection methods for flavonoids. Journal of Chromatography.
A, 1112(1-2), 31–63. https://doi.org/10.1016/j.chroma.2006.
01.019
de Rijke, E., Zappey, H., Ariese, F., Gooijer, C., & Brinkman,
U. A. (2003). Liquid chromatography with atmospheric
pressure chemical ionization and electrospray ionization
Standard methods for pollen research 99
mass spectrometry of flavonoids with triple quadrupole and
ion-trap instruments. Journal of Chromatography. A, 984(1),
45–58. https://doi.org/10.1016/S0021-9673(02)01868-X
Delaplane, K. S., Dag, A., Danka, R. G., Freitas, B. M.,
Garibaldi, L. A., Goodwin, R. M., & Hormaza, J. I. (2013).
Standard methods for pollination research with Apis melli-
fera. In V. Dietemann, J. D. Ellis, & P. Neumann (Eds.), The
COLOSS BEEBOOK, Volume I: standard methods for Apis
mellifera research. Journal of Apicultural Research, 52(4).
https://doi.org/10.3896/IBRA.1.52.4.12.
Di Paola-Naranjo, R. D., S
anchez, S. J., Param
as, A. M. G., &
Rivas Gonzalo, J. C. (2004). Liquid chromatographic-mass
spectrometric analysis of anthocyanin composition of dark
blue bee pollen from Echium plantagineum. Journal of
Chromatography. A, 1054(1-2), 205–210. https://doi.org/10.
1016/j.chroma.2004.05.023
Droussi, Z., D'orazio, V., Provenzano, M. R., Hafidi, M., &
Ouatmane, A. (2009). Study of the biodegradation and
transformation of olive-mill residues during composting
using FTIR spectroscopy and differential scanning calorim-
etry. Journal of Hazardous Materials, 164(2-3), 1281–1285.
https://doi.org/10.1016/j.jhazmat.2008.09.081
D€ubecke, A., Beckh, G., & L€ullmann, C. (2011). Pyrrolizidine
alkaloids in honey and bee pollen. Food Additives &
Contaminants. Part A, Chemistry, Analysis, Control, Exposure &
Risk Assessment, 28(3), 348–358. https://doi.org/10.1080/
19440049.2010.541594
Dustman, J. H. (1992). Honey, quality and its control.
Abstracts of the 8th International Palynological Congress.
France, p. 40.
Dutkiewicz, S. (1996). Usefulness of Cernilton in the treatment of
benign prostatic hyperplasia. International Urology and
Nephrology, 28(1), 49–53. https://doi.org/10.1007/BF02550137
Dutta, D., Devi, S. S., Krishnamurthi, K., Kumar, K., Vyas, P.,
Muthal, P. L., Naoghare, P., & Chakrabarti, T. (2007).
Modulatory effect of distillate of Ocimum sanctum leaf
extract (Tulsi) on human lymphocytes against genotoxi-
cants. Biomedical and Environmental Sciences, 20(3), 226–234.
Dzyuba, O. F. (2006). Palinoindikatsiya environmental quality.
SPb.: Publishing House "Nedra", 198 p. (Article in Russian).
Dzyuba, O. F. (2007). Teratomorfnye pollen grains in a modern
and paleopalinologicheskih pollen spectra and some problems.
Palynostratigraphy/OF Dziuba//Oil and Gas Technologies: Theory
and Practice, 2007(2), 1–22. (Article in Russian).
Dzyuba, O. F., & Tarasovich, V. F. (2001). The morphological
features of pollen grains of Tilia Cordata Mill. In the mod-
ern metropolis. International Workshop: Pollen as an indi-
cator of environmental and paleoecological reconstruction
– St. Petersburg, pp. 79–90. (Article in Russian).
Echigo, T., Takenaka, T., & Yatsunami, K. (1986). Comparative
studies on chemical composition of honey, royal jelly and
pollen loads. Bulletin of the Faculty of Agriculture Tamagawa
University, 26, 1–8.
Edgar, R. C. (2010). Search and clustering orders of magnitude
faster than BLAST. Bioinformatics, 26(19), 2460–2461.
https://doi.org/10.1093/bioinformatics/btq461
Eloff, J. N. (1998). A sensitive and quick microplate method to
determine the minimal inhibitory concentration of plant
extracts for bacteria. Planta Medica, 64(8), 711–713. https://
doi.org/10.1055/s-2006-957563
Eraslan, G., Kanbur, M., & Silici, S. (2009). Effect of carbaryl on
some biochemical changes in rats: The ameliorative effect
of bee pollen. Food and Chemical Toxicology, 47(1), 86–91.
https://doi.org/10.1016/j.fct.2008.10.013
Eraslan, G., Kanbur, M., Silici, S., Liman, B., Altinordulu, S., &
Sarica, Z. S. (2009). Evaluation of protective effect of bee
pollen against propoxur toxicity in rat. Ecotoxicology and
Environmental Safety, 72(3), 931–937. https://doi.org/10.
1016/j.ecoenv.2008.06.008
Erkmen, O., & Ozcan, M. M. (2008). Antimicrobial effects of
Turkish propolis, pollen, and laurel on spoilage and patho-
genic food-related microorganisms. Journal of Medicinal Food,
11(3), 587–592. https://doi.org/10.1089/jmf.2007.0038
Estevinho, L. M., Pereira, A. P., Moreira, L., Dias, L. G., &
Pereira, E. (2008). Antioxidant and antimicrobial effects of
phenolic compounds extracts of Northeast Portugal honey.
Food and Chemical Toxicology, 46(12), 3774–3779. https://doi.
org/10.1016/j.fct.2008.09.062
Eswaran, V. U., & Bhargava, H. R. (2014). Chemical analysis
and anti-microbial activity of Karnataka bee bread of Apis
species. World Applied Sciences Journal, 32(3), 379–385.
https://doi.org/10.5829/idosi.wasj.2014.32.03.1006
Evans, J. D., Schwarz, R. S., Chen, Y.-P., Budge, G., Cornman, R.
S., De La Rua, P., De Miranda, J. R., Foret, S., Foster, L.,
Gauthier, L., Genersch, E., Gisder, S., Jarosch, A., Kucharski,
R., Lopez, D., Lun, C. M., Moritz, R. F. A., Maleszka, R.,
Muñoz, I., & Pinto, M. A. (2013). Standard methodologies for
molecular research in Apis mellifera. In V. Dietemann, J. D. Ellis
& P. Neumann (Eds), The COLOSS BEEBOOK, Volume I: stand-
ard methods for Apis mellifera research. Journal of Apicultural
Research, 52(4), https://doi.org/10.3896/IBRA.1.52.4.11
Fabre, N., Rustan, I., Hoffmann, E., & Quetin-Leclercq, J. (2001).
Determination of flavone, flavonol and flavanone aglycones by
negative ion LC–ES ion trap mass spectrometry. Journal of the
American Society for Mass Spectrometry, 12(6), 707–715. https://
doi.org/10.1016/S1044-0305(01)00226-4
Faegri, K., Kaland P, E., & Krzywinski, K. (1993).
Bestimmungsschlussel fur die nordwesteuropaische pollenflora
(p. 85). G. Fischer.
Fanali, C., Dugo, L., & Rocco, A. (2013). Nano-liquid chroma-
tography in nutraceutical analysis: determination of polyphe-
nols in bee pollen. Journal of Chromatography. A, 1313,
270–274. https://doi.org/10.1016/j.chroma.2013.06.055
FAO/IPGRI. (2001). Multi-crop passport descriptors. http://www.
bioversityinternational.org/fileadmin/bioversity/publications/
pdfs/124.pdf
Fatrcov
a-Sramkov
a, K., Nȏzkov
a, J., Kac
aniov
a, M., M
ariassyov
a,
M., & Kropkov
a, Z. (2010). Microbial properties, nutritional
composition and antioxidant activity of Brassica napus subsp.
napus L. bee pollen used in human nutrition. Ecological
Chemistry and Engineering A, 17(1), 45–54.
Fatrcov
a-Sramkov
a, K., No^zkov
a, J., Kac
aniov
a, M., &
Mari
assyov
a, M. (2011). Antioxidant and antimicrobial prop-
erties of monofloral bee pollen usable in human nutrition.
In Chemick
e Listy, 105(Suppl), 1005–1006. https://doi.org/10.
1080/03601234.2013.727664
Fatrcov
a-Sramkov
a, K., No^zkov
a, J., Kac
aniov
a, M.,
M
ari
assyov
a, M., Rovn
a, K., & Stric
ık, M. (2013).
Antioxidant and antimicrobial properties of monofloral bee
pollen. Journal of Environmental Science and Health. Part. B,
Pesticides, Food Contaminants, and Agricultural Wastes, 48(2),
133–138. https://doi.org/10.1080/03601234.2013.727664
Fe
as, X., V
azquez-Tato, M. P., Estevinho, L., Seijas, J. A., &
Iglesias, A. (2012). Organic bee pollen: Botanical origin,
nutritional value, bioactive compounds, antioxidant activity
and microbiological quality. Molecules, 17(7), 8359–8377.
http://dx.doi.org/10.3390/molecules17078359
Forcone, A., Aloisi, P. V., Ruppel, S., & Munoz, M. (2011).
Botanical composition and protein content of pollen col-
lected by Apis mellifera L. in the north-west of Santa Cruz
(Argentinean Patagonia). Grana, 50(1), 30–39. https://doi.
org/10.1080/00173134.2011.552191
Ferreres, F., Pereira, D. M., Valent~ao, P., & Andrade, P. B. (2010).
First report of non-coloured flavonoids in Echium plantagineum
bee pollen: Differentiation of isomers by liquid chromatography/
ion trap mass spectrometry. Rapid Communications in Mass
Spectrometry, 24(6), 801–806. https://doi.org/10.1002/rcm.4454
100 M. G. Campos et al.
Fogle, H. W. (1977). Identification of tree fruit species by pol-
len ultrastructure. Journal – American Society for Horticultural
Science, 102(5), 548–551.
Fontanals, N., Marc
e, R. M., & Borrull, F. (2007). New materi-
als in sorptive extraction techniques for polar compounds.
Journal of Chromatography. A, 1152(1-2), 14–31. https://doi.
org/10.1016/j.chroma.2006.11.077
Forcone, A., Aloisi, P. V., Ruppel, S., & Munoz, M. (2011).
Botanical composition and protein content of pollen col-
lected by Apis mellifera L. in the north-west of Santa Cruz
(Argentinean Patagonia). Grana, 50(1), 30–39. https://doi.
org/10.1080/00173134.2011.552191
Fossel, A.-M., & Pechhacker, H. (2001). Bienen and blumen (p.
676). Institut fur Bienenkunde.
Franco, B. D. G. M. (2002). Fatores intr
ınsecos e extr
ınsecos
que controlam o desenvolvimento microbiano nos alimen-
tos. In B. D. G. M. Franco & M. Landgraf (Eds.),
Microbiologia dos Alimentos (pp. 13–26). Atheneu.
Fukunaga, K., & Hostetler, L. (1975). The estimation of the
gradient of a density function, with applications in pattern
recognition. IEEE Transactions on Information Theory, 21(1),
32–40. https://doi.org/10.1109/TIT.1975.1055330
Galimberti, A., De Mattia, F., Bruni, I., Scaccabarozzi, D.,
Sandionigi, A., Barbuto, M., Casiraghi, M., & Labra, M.
(2014). A DNA barcoding approach to characterize pollen
collected by honeybees. PLoS One, 9(10), e109363. https://
doi.org/10.1371/journal.pone.0109363
Gasi
c, U., Keckes, S., Dabi
c, D., Trifkovi
c, J., Milojkovi
c-
Opsenica, D., Nati
c, M., & Tesi
c, Z. (2014b). Phenolic pro-
file and antioxidant activity of Serbian polyfloral honeys.
Food Chemistry, 145, 599–607. https://doi.org/10.1016/j.food-
chem.2013.08.088
Gasi
c, U., Sikoparija, B., Tosti, T., Trifkovi
c, J., Milojkovi
c-
Opsenica, D., Nati
c, M., & Tesi
c, Z. Z (2014a).
Phytochemical fingerprints of lime honey collected in
Serbia. Journal of AOAC International, 97(5), 1259–1267.
https://doi.org/10.5740/jaoacint.SGEGasic
Germano, M. I. S., Germano, P. M. L., Kamei, C. A. K., Abreu,
E. S. d., Ribeiro, E. R., Silva, K. C. d., Lamardo, L. C. A., Rocha,
M. F. G., Vieira, V. K. I., & Kawasaki, V. M. (2000). Manipuladores
de alimentos: capacita? 
E preciso! Regulamentar? Ser
a preciso?
HigieneAlimentar, 14(78/79), 18–22.
Gilber, M. D., & Lisk, D. J. (1978). Honey as an environmental
indicator of radionuclide contamination. Bulletin of
Environmental Contamination and Toxicology, 19(1), 32–34.
https://doi.org/10.1007/BF01685763
Gilliam, M., Prest, D. B., & Lorenz, B. J. (1989). Microbiology of pol-
len and bee bread: taxonomy and enzimology of molds.
Apidologie, 20(1), 53–68. https://doi.org/10.1051/apido:19890106
Glazer, A. N. (1990). Phycoerythrin fluorescence-based assay
for reactive oxygen species . Methods in Enzymology, 186,
161–168. https://doi.org/10.1016/0076-6879(90)86106-6
Glazunov, K. P. (2001 Pollen as an indicator of negative environmental
factors: embryological aspect [Paper presentation]. International
Seminar: Pollen as an Indicator of Environmental and
Paleoecological Reconstruction, pp. 61–63. (Article in Russian).
Goh, K., Chang, E., & Cheng, K. (2001). SVM binary classifier
ensembles for image classification [Paper presentation].
Proceedings of the ACM Conference Information and
Knowledge Management, New York, pp. 395–402. https://
doi.org/10.1145/502585.502652
Goh, K., Chang, E., & Li, B. (2005). Using one-class and two-
class SVMs for multiclass image annotation. IEEE
Transactions on Knowledge and Data Engineering, 17(10),
1333–1346. https://doi.org/10.1109/TKDE.2005.170
Gok, S., Severcan, M., Goormaghtigh, E., Kandemir, I., &
Severcan, F. (2015). Differentiation of Anatolian honey sam-
ples from different botanical origins by ATR-FTIR
spectroscopy using multivariate analysis. Food Chemistry,
170, 234–240. https://doi.org/10.1016/j.foodchem.2014.08.
040
Gonzalez, R. C., & Woods, R. E. (2008). Digital image process-
ing (3rd ed.). Prentice Hall.
Graikou, K., Kapeta, S., Aligiannis, N., Sotiroudis, G.,
Chondrogianni, N., Gonos, E., & Chinou, I. (2011).
Chemical analysis of Greek pollen – Antioxidant, antimicro-
bial and proteasome activation properties. Chemistry Central
Journal, 5(1), 33–39. https://doi.org/10.1186/1752-153X-5-33
Greenham, J., Williams, C., & Harborne, J. B. (1995). Identification
of lipophilic flavonols by a combination of chromatographic and
spectral techniques. Phytochemical Analysis, 6(4), 211–217.
https://doi.org/10.1002/pca.2800060407
Grembecka, M., & Szefer, P. (2013). Evaluation of honeys and
bee products quality based on their mineral composition
using multivariate techniques. Environmental Monitoring and
Assessment, 185(5), 4033–4047. https://doi.org/10.1007/
s10661-012-2847-y
Gu, D., Yang, Y., Abdulla, R., & Aisa, H. A. (2012).
Characterization and identification of chemical compositions
in the extract of Artemisia rupestris L. by liquid chromatog-
raphy coupled to quadrupole time-of-flight tandem mass
spectrometry. Rapid Communications in Mass Spectrometry,
26(1), 83–100. https://doi.org/10.1002/rcm.5289
Guide for the Care and Use of Laboratory Animals Eighth Edition.
(2011). The National Academies Press 500 Fifth Street, NW
Washington, DC 20001. www.national-academies.org
Guo, J. T., Lee, H. L., Chiang, S. H., Lin, F. I., & Chang, C. Y.
(2020). Antioxidant properties of the extracts from different
parts of broccoli in Taiwan. Journal of Food and Drug Analysis,
9(2), 96–101. https://doi.org/10.38212/2224-6614.2795
Harborne, J. B., & Boardley, M. (1984). Use of high performance
liquid chromatography in the separation of flavonol glycosides
and flavonol sulphates. Journal of Chromatography A, 299(2),
377–385. https://doi.org/10.1016/S0021-9673(01)97853-7
Hartmann, T., & Toppel, G. (1987). Senecionine n-oxide. The
primary product of pyrrolizidine alkaloid biosynthesis in
root cultures of Senecio vulgaris. Phytochemistry, 26(6),
1639–1643. https://doi.org/10.1016/S0031-9422(00)82261-X
Hartmann, T., & Witte, L. (1995). Chemistry, biology and che-
moecology of the pyrrolizidine alkaloids. In S. W. Pelletier
(Ed.), Alkaloids: Chemical and biological perspectives (Vol. 9,
pp. 155–233). Pergamon Press. https://doi.org/10.1016/
b978-0-08-042089-9.50011-5
Haussmann, B. I. G., & Parzies, H. K. (2009). Methodologies
for generating variability: Part 1: Use of genetic resources
in plant breeding. In S. Ceccarelli (Ed.), Plant breeding and
farmer participation (pp. 107–128). FAO of the UN.
Heo, M. Y., Sohn, S. J., & Au, W. W. (2001). Anti-genotoxicity
of galangin as a cancer chemopreventive agent candidate.
Mutation Research, 488(2), 135–150. https://doi.org/10.1016/
S1383-5742(01)00054-0 http://www.mercola.com/article/
diet/bee_pollen.htm.
Herbert, E. W. J. (1992). Honey bee nutrition. In J. E. Graham
(Ed.), The hive and the honey bee (pp. 197–233). Dadant &
Sons Inc.
Herbert, E. W. J., & Shimanuki, H. (1978). Chemical composition
and nutritive value of bee-collected and bee-stored pollen.
Apidologie, 9(1), 33–40. https://doi.org/10.1051/apido:19780103
Hewitt, W., & Vincent, S. (1989). Theory and application of
microbiological assay. Academic Press.
Hleba, L., Pochop, J., Fels€ociov
a, S., Petrov
a, J., Cubon, J.,
Pavelkov
a, A., & Kac
aniov
a, M. (2013). Antimicrobial effect
of bee collected pollen extract to Enterobacteriaceae
Genera after application of bee collected pollen in their
feeding. Animal Science and Biotechnologies, 46(2), 108–113.
Hleba, L., Pochop, J., Fels€ociov
a, S., Petrov
a, J., Cubon, J. &
Pavelkov
a, A. (2013). Characterization of chemical
Standard methods for pollen research 101
composition of bee pollen in China. Food Chemistry, 61(3),
708–718. https://doi.org/10.1021/jf304056b
Hodges, D. (1952). The pollen loads of the honey bee (p. 154).
Bee Research Association.
Human, H., Brodschneider, R., Dietemann, V., Dively, G., Ellis,
J. D., Forsgren, E., Fries, I., Hatjina, F., Hu, F.-L., Jaffé, R.,
Jensen, A. B., Köhler, A., Magyar, J., Özikrim, A., Pirk, C.
W. W., Rose, R., Strauss, U., Tanner, G., Tarpy, D. R., van
der Steen, J. J. M., Vaudo, A., Vejsnaes, F., Wilde, J.,
Williams, G. R., & Zheng, H.-Q. (2013). Miscellaneous
standard methods for Apis mellifera research. In V.
Dietemann, J. D. Ellis & P. Neumann (Eds), The COLOSS
BEEBOOK, Volume I: standard methods for Apis mellifera
research. Journal of Apicultural Research, 52(4), https://doi.org/
10.3896/IBRA.1.52.4.10
Human, H., & Nicolson, S., W. (2006). Nutritional content of
fresh, bee-collected and stored pollen of Aloe greatheadii
var. davyana (Assphodelaceae). Phytochemistry, 67(14),
1486–1492. https://doi.org/10.1016/j.phytochem.2006.05.023
Ibrahim, A., & Spivak, M. (2006). The relationship between
hygienic behavior and suppression of mite reproduction as
honey bee (Apis mellifera) mechanisms of resistance to
Varroa destructor. Apidologie, 37(1), 31–40. https://doi.org/10.
1051/apido:2005052
Ionita, P. (2005). Is DPPH stable free radical a good scavenger
for oxygen active species? Chemical Papers, 50(1), 11–16.
Jaiswal, S., Mansa, N., Prasad, M. S., Jena, B. S., & Negi, P. S.
(2014). Antibacterial and antimutagenic activities of (Dillenia
indica) extracts. Food Bioscience, 5, 47–53. https://doi.org/10.
1016/j.fbio.2013.11.005
Jakobek, L. (2015). Interactions of polyphenols with carbohy-
drates, lipids and proteins. Food Chemistry, 175, 556–567.
https://doi.org/10.1016/j.foodchem.2014.12.013
Jiang, L. (2011). Comparison of disk diffusion, agar dilution, and
broth microdilution for antimicrobial susceptibility testing of five
chitosans [Thesis in Master of Science]. Faculty of the
Louisiana State University and Agricultural and Mechanical
College, Department of Food Science.
Jorgensen, J. H., & Ferraro, M. J. (2009). Antimicrobial suscep-
tibility testing: A review of general principles and contem-
porary practices. Clinical Infectious Diseases, 49(11),
1749–1755. https://doi.org/10.1086/647952
Kac
aniov
a, M., Vukovi
c, N., Chlebo, R., Hasc
ık, P., Rovn
a, K.,
Cubon, J., Dzugan, M., & Pasternakiewicz, A. (2012). The
antimicrobial activity of honey, bee pollen loads and bees-
wax from Slovakia. Archives of Biological Sciences, 64(3),
927–934. https://doi.org/10.2298/ABS1203927K
Kac
aniov
a, M., Terentjeva, M., Petrov
a, J., Hleba, L., Shevtsova,
T., Kluz, M., Rozek, P., & Vukovic, N. (2014). Antimicrobial
activity of non-traditional plant pollen against different spe-
cies of microorganisms. Journal of Microbiology Biotechnology
and Food Science, 4(1), 80–82. http://dx.doi.org/10.15414/
jmbfs.2014.4.1.80-82
Kao, Y. T., Lu, M. J., & Chen, C. (2020). Preliminary analysis of
phenolic compounds and antioxidant activities in tea pollen
extracts. Journal of Food and Drug Analysis, 19(4), 470–477.
https://doi.org/10.38212/2224-6614.2177
Kaur, C., & Kapoor, H. C. (2001). Antioxidants in fruits and
vegetables-the millennium’s health. International Journal of
Food Science and Technology, 36(7), 703–725. https://doi.org/
10.1046/j.1365-2621.2001.00513.x
Kaur, R., Nagpal, A., & Katnoria, J. K. (2015). Exploration of
antitumor properties of pollen grains of plant species
belonging to Fabaceae family. Journal of Pharmaceutical
Sciences and Research, 7(3), 127–129.
Keckes, S., Gasi
c, U., Velickovi
c, T. 
C., Milojkovi
c-Opsenica,
D., Nati
c, M., & Tesi
c, Z. (2013). The determination of
phenolic profiles of Serbian unifloral honeys using ultra-
high-performance liquid chromatography/high resolution
accurate mass spectrometry. Food Chemistry, 138(1), 32–40.
https://doi.org/10.1016/j.foodchem.2012.10.025
Kedves, M., Kocsicska, I., & Priskin, K. (2003). Transmission electron
microscopy of the connectives of the pollen grains of Ginkgo
biloba L. Plant Cell Biology and Development, Szeged, 15, 28–42.
Kedves, M., Pardutz, A., & Horvath, A. (2002). Transmission
electron microscopy of partially degraded pollen grains of
Ambrosia artemisiifolia (ragweed). Plant Cell Biology and
Development, Szeged, 14, 49–58.
Kelina, I., Saavedra, C., Consuelo Rojas, I., Guillermo, E., &
Delgado, P. (2013). Pollinic characteristics and chemical
composition of bee pollen collected in Cayalti (Lambayeque
– Peru). Revista Chilena de Nutricion, 40(1), 71–78.
Keller, A., Danner, N., Grimmer, G., Ankenbrand, M., von der
Ohe, K., von der Ohe, W., Rost, S., H€artel, S., & Steffan-
Dewenter, I. (2015). Evaluating multiplexed next-generation
sequencing as a method in palynology for mixed pollen samples.
Plant Biology, 17(2), 558–566. https://doi.org/10.1111/plb.12251
Keller, A., Hohlfeld, S., Kolter, A., Schultz, J., Gemeinholzer, B., &
Ankenbrand, M. J. (2020). BCdatabaser: On-the-fly reference
database creation for (meta-)barcoding. Bioinformatics, 36(8),
2630–2631. https://doi.org/10.1093/bioinformatics/btz960
Keller, I., Fluri, P., & Imdorf, A. (2005). Pollen nutrition and
colony development in honey bees: Part I. Bee World, 86(1),
3–10. https://doi.org/10.1080/0005772X.2005.11099641
Kempf, M., Beuerle, T., B€uhringer, M., Denner, M., Trost, D., von
der Ohe, K., Bhavanam, V. B. R., & Schreier, P. (2008).
Pyrrolizidine alkaloids in honey: Risk analysis by gas chromatog-
raphy-mass spectrometry. Molecular Nutrition & Food Research,
52(10), 1193–1200. https://doi.org/10.1002/mnfr.200800051
Kempf, M., Heil, S., Hasslauer, I., Schmidt, L., von der Ohe, K.,
Theuring, C., Reinhard, A., Schreier, P., & Beuerle, T.
(2010). Pyrrolizidine alkaloids in pollen and pollen products.
Molecular Nutrition & Food Research, 54(2), 292–300. https://
doi.org/10.1002/mnfr.200900289
Kempf, M., Wittig, M., Reinhard, A., von der Ohe, K.,
Blacquiere, T., Raezke, K.-P., Michel, R., Schreier, P., &
Beuerle, T. (2011). Pyrrolizidine alkaloids in honey:
Comparison of analytical methods. Food Additives &
Contaminants. Part A, Chemistry, Analysis, Control, Exposure &
Risk Assessment, 28(3), 332–347. https://doi.org/10.1080/
19440049.2010.521772
Khynriam, D., & Prasad, S. B. (2003). Changes in glutathione-
related enzymes in tumor-bearing mice after cisplatin treat-
ment. Cell Biology and Toxicology, 18(6), 1573–6822.
Kirk, W. D. J. (1994). A color guide to pollen loads of the honey
bee (p. 54). International Bee Research Association.
Klancnik, A., Piskernik, S., Jersek, B., & Mozina, S. S. (2010).
Evaluation of diffusion and dilution methods to determine
the antibacterial activity of plant extracts. Journal of
Microbiological Methods, 81(2), 121–126. https://doi.org/10.
1016/j.mimet.2010.02.004
Klungness, L. M., & Peng, Y.-S. (1983). A scanning electron
microscopic study of pollen loads collected and stored by
honey bees. Journal of Apicultural Research, 22(4), 264–271.
https://doi.org/10.1080/00218839.1983.11100598
Koleva, I. I., VAN Beek, T. A., Linssen, J. P. H., De Groot, A.,
& Evstatieva, L. N. (2002). Screening of plant extracts for
antioxidant activity: A comparative study on three testing
methods. Phytochemical Analysis, 13(1), 8–17. https://doi.org/
10.1002/pca.611
Komosinska-Vassev, K., Olczyk, P., Kafmierczak, J., Mencner,
L., & Olczyk, K. (2015) Bee pollen: Chemical composition
and therapeutic application. Evidence-Based Complementary
and Alternative Medicine, 2015, 1–6. https://doi.org/10.1155/
2015/297425.
K€oppler, K., Vorwohl, G., & Koeniger, N. (2007). Comparison
of pollen spectra collected by four different subspecies of
102 M. G. Campos et al.
the honey bee Apis mellifera. Apidologie, 38(4), 341–353.
https://doi.org/10.1051/apido:2007020
Kosti
c, A. Z., Bara
c, M. B., Stanojevi
c, S. P., Milojkovi
c-Opsenica,
D. M., Tesi
c, Z. L., Sikoparija, B., Radisi
c, P., Prentovi
c, M., &
Pesi
c, M. B. (2015). Physicochemical composition and techno-
functional properties of bee pollen collected in Serbia. LWT –
Food Science and Technology, 62(1), 301–309. https://doi.org/10.
1016/j.lwt.2015.01.0310023-6438
Kozich, J. J., Westcott, S. L., Baxter, N. T., Highlander, S. K.,
& Schloss, P. D. (2013). Development of a dual-index
sequencing strategy and curation pipeline for analyzing
amplicon sequence data on the MiSeq Illumina sequencing
platform. Applied and Environmental Microbiology, 79(17),
5112–5120. https://doi.org/10.1128/AEM.01043-13
Kraaijeveld, K., de Weger, L. A., Ventayol Garc
ıa, M.,
Buermans, H., Frank, J., Hiemstra, P. S., & den Dunnen, J. T.
(2015). Efficient and sensitive identification and quantifica-
tion of airborne pollen using next-generation DNA
sequencing. Molecular Ecology Resources, 15(1), 8–16. https://
doi.org/10.1111/1755-0998.12288
Kroyer, G., & Hegedus, N. (2001). Evaluation of bioactive prop-
erties of pollen extracts as functional dietary food supplement.
Innovative Food Science & Emerging Technologies, 2(3), 171–174.
https://doi.org/10.1016/S1466-8564(01)00039-X
Langford, M., Taylor, G., & Flenley, J. (1990). Computerized
identification of pollen grains by texture analysis. Review of
Palaeobotany and Palynology, 64(1-4), 197–203. https://doi.
org/10.1016/0034-6667(90)90133-4
Lautraite, S., Musonda, A. C., Doehmer, J., Edwards, G. O., &
Chipman, J. K. (2002). Flavonoids inhibit genetic toxicity
produced by carcinogens in cells expressing CYP1A2 and
CYP1A1. Mutagenesis, 17(1), 45–53. https://doi.org/10.1093/
mutage/17.1.45
LeBlanc, B. W., Davis, O. K., Boue, S., DeLucca, A., & Deeby,
T. (2009). Antioxidant activity of Sonoran Desert bee pol-
len. Food Chemistry, 115(4), 1299–1305. https://doi.org/10.
1016/j.foodchem.2009.01.055
Lee, K.-H., Kim, A.-J., & Choi, E.-M. (2009). Antioxidant and anti-
inflammatory activity of pine pollen extract in vitro. Phytotherapy
Research, 23(1), 41–48. https://doi.org/10.1002/ptr.2525
Leja, M., Mareczek, A., Wy_zgolik, G., Klepacz-Baniak, J., &
Czeko
nska, K. (2007). Antioxidative properties of bee pol-
len in selected plant species. Food Chemistry, 100(1),
237–240. https://doi.org/10.1016/j.foodchem.2005.09.047
Leong, L. P., & Shui, G. (2002). An investigation of antioxidant
capacity of fruits in Singapore market. Food Chemistry, 76(1),
69–75. https://doi.org/10.1016/S0308-8146(01)00251-5
Li, A., Zhang, Q., Zhang, G., Chen, J., Fei, Z., & Liu, F. (2002).
Adsorption of phenolic compounds from aqueous solutions
by a water-compatible hypercrosslinked polymeric adsorb-
ent. Chemosphere, 47(9), 981–989. https://doi.org/10.1016/
S0045-6535(01)00222-3
Li, P., & Flenley, J. R. (1999). Pollen texture identification using
neural networks. Grana, 38(1), 59–64. In https://doi.org/10.
1080/001731300750044717
Lindtner, P. (2014). Garden plants for honey bees (p. 389).
Wicwas Press LLC.
Liolios, V., Tananaki, C., Dimou, M., Kanelis, D., Goras, G.,
Karazafiris, E., Thrasyvoulou, A. (2014). Predict the chemical
composition of pollen. In Proceedings of International Symposium on
Bee Products, 3rd edition, September 28th –October 1st, p. 28.
Lokutova, O. (2014). Quality of bee products in context of envir-
onmental safety. Zborn
ık vedeck
ych pr
ac a v
ysledkov z riesenia
v
yskumn
eho projektu Podpora inov
acie technol
ogi
ı
speci
alnych v
yrobkov biopotrav
ın pre zdrav
u v
yzivu l'ud
ı –
Slovensk
a pol'nohospod
arska univerzita v Nitre, pp. 182–188.
Loper, G. M., Standifer, L. N., Thompson, M. J., & Gilliam, M.
(1980). Biochemistry and microbiology of bee collected
almond (Prumus dulcis) pollen and bee bread. I Fatty acids,
vitamins and minerals. Apidologie, 11(1), 63–73. https://doi.
org/10.1051/apido:19800108
Louveaux, J., Maurizio, A., & Vorwohl, G. (1978). Methods of
melissopalynology. Bee World, 59(4), 139–157. https://doi.
org/10.1080/0005772X.1978.11097714
Lucchese, L., Mitra, S. (2001). Color image segmentation: A
state-of-the-art survey. In Proceedings of the Indian National
Science Academy (INSA-A), pp. 207–221.
Lv, H., Wang, X., He, Y., Wang, H., & Suo, Y. (2015).
Identification and quantification of flavonoid aglycones in
rape bee pollen from Qinghai-Tibetan Plateau by HPLC-
DAD-APCI/MS. Journal of Food Composition and Analysis, 38,
49–54. https://doi.org/10.1016/j.jfca.2014.10.011
Madejczyk, M., & Baralkiewicz, D. (2008). Characterization of
Polish rape and honeydew honey according to their mineral
contents using ICP-MS and F-AAS/AES. Analytica Chimica Acta,
617(1-2), 11–17. https://doi.org/10.1016/j.aca.2008.01.038
Makena, P. S., Pierce, S. C., Chung, K. T., & Sinclair, S. E.
(2009). Comparative mutagenic effects of structurally simi-
lar flavonoids quercetin and taxifolin on tester strains
Salmonella typhimurium TA102 and Escherichia coli WP-2
uvrA. Environmental and Molecular Mutagenesis, 50(6),
451–459. https://doi.org/10.1002/em.20487
Margaoan, R., Marghitas, L., Al, Dezmirean, D. S., Bobis, O., &
Mihai, C. M. (2012). Physical-Chemical composition of fresh
bee pollen from Transylvania. Bulletin UASVM Animal Science
and Biotechnologies, 69(1-2), 351–355.
Marghitas¸, L. A., Stanciu, O. G., Dezmirean, D. S., Bobis¸, O.,
Popescu, O., Bogdanov, S., & Campos, M. G. (2009). In vitro
antioxidant capacity of honey bee-collected pollen of
selected floral origin harvested from Romania. Food
Chemistry, 115(3), 878–883. https://doi.org/10.1016/j.food-
chem.2009.01.014
Markham, K., & Campos, M. G. (1996). 7- and 8-O-methylher-
bacetin-3-O-sophoroside from bee pollens and some struc-
ture/activity observations. Phytochemistry, 43(4), 763–767.
https://doi.org/10.1016/0031-9422(96)00286-5
Markham, K. R., & Geiger, H. (1993). 1H nuclear magnetic reson-
ance spectroscopy of flavonoids and their glycosides in hexadeu-
terodimethylsulphoxide. In J. B. Harborne (Ed.), The flavonoids:
Advances in research since 1986. Chapman and Hall, pp. 441–496.
Markowicz Bastos, D. H., Barth, M. O., Rocha, C. I., Cunha,
I. B. S., Carvalho, P. O., Torres, E. A. S., & Michelan, M.
(2004). Fatty acid composition and palynological analysis of
bee (Apis) pollen loads in the states of S~ao Paulo and Minas
Gerais, Brazil. Journal of Apicultural Research, 43(2), 35–39.
https://doi.org/10.1080/00218839.2004.11101107
Marsh, J., & Goode, J. A. (Eds.) (1994). Antimicrobial Peptides
ICiba Foundation Symposium 186]. John Wiley & Sons.
Martins, M. C. T., Morgano, M. A., Vincente, E., Baggio, S. R.,
& Rodriguez-Amaya, D. B. (2011). Physicochemical compos-
ition of bee pollen from eleven Brazilian states. Journal of
Apicultural Science, 55(2), 107–116.
Maruyama, H., Sakamoto, T., Araki, Y., & Hara, H. (2010).
Anti-inflammatory effect of bee pollen ethanol extract from
Cistus sp of Spanish on carrageenan-induced rat hind paw
edema. BMC Complementary and Alternative Medicine, 10(1),
11. https://doi.org/10.1186/1472-6882-10-30
Matuschek, E., Brown, D. F. J., & Kahlmeter, G. (2014).
Development of the EUCAST disk diffusion antimicrobial sus-
ceptibility testing method and its implementation in routine
microbiology laboratories. Clinical Microbiology and Infection,
20(4), O255–O266. https://doi.org/10.1111/1469-0691.12373
Maurizio, A., & Louveaux, J. (1965). Pollens de plantes melliferes
d’Europe (p. 148). Union des Groupements apicoles Franc¸ais.
McMurdie, P. J., & Holmes, S. (2013). phyloseq: An R package
for reproducible interactive analysis and graphics of micro-
biome census data. PLoS One, 8(4), e61217. https://doi.org/
10.1371/journal.pone.0061217
Standard methods for pollen research 103
Mej
ıas, E., & Montenegro, G. (2012). The antioxidant activity
of Chilean honey and bee pollen produced in the Llaima
volcano’s zonez. Journal of Food Quality, 35(5), 315–322.
https://doi.org/10.1111/j.1745-4557.2012.00460.x
Melo, E. A., Filho, J. M., & Guerra, N. B. (2005) Characterization
of antioxidant compounds in aqueous coriander extract
(Coriandrum sativum L.). LWT—Food Science and Technology, 38,
15–19. https://doi.org/10.1016/j.lwt.2004.03.011.
Melo, I. L. P., & Almeida-Muradian, L. B. (2010). Stability of anti-
oxidants vitamins in bee pollen samples. Qu
ımica Nova, 33(3),
514–518. https://doi.org/10.1590/S0100-40422010000300004
Melo, I. L. P., & Almeida-Muradian, L. B. (2011). Comparison
of methodologies for dried bee pollen's moisture determin-
ation. Ci^encia e Tecnologia de Alimentos, 31(1), 194–197. v.
https://doi.org/10.1590/S0101-20612011000100029
Melo, I. L. P., Freitas, A. S., Barth, O. M., & Almeida-Muradian,
L. B. (2009). Relac¸~ao entre a composic¸~ao nutricional e a
origem floral de p
olen ap
ıcola desidratado. [Correlation
between nutritional composition and floral origin of dried
bee pollen]. Revista Do Instituto Adolfo Lutz, 68, 346–353.
Melo, A. A. M., Meira, D. F. S., Chan, J., Sattler, J. A. G., &
Almeida-Muradian, L. B. (2012a). Comparac¸~ao entre
m
etodos de extrac¸~ao de compostos antioxidantes do p
olen
ap
ıcola desidratado. In: 11 Encontro de Qu
ımica de
Alimentos, Braganc¸a. 11 Encontro de Qu
ımica de Alimentos –
Qualidade dos alimentos: novos desafios/Resumos, p. 65–65.
Melo, A. A. M., Meira, D. F. S., Sattler, J. A. G., & Almeida-
Muradian, L. B. (2012b). Atividade antioxidante do p
olen
ap
ıcola desidratado avaliada por diferentes m
etodos de
extrac¸~ao. Mensagem Doce (Associac¸~ao Paulista de
Apicultores, Criadores de Abelhas Melificas Europ
eias), S~ao
Paulo, p. 59, 22 maio.
Melo, A. A. M., Meira, D. F. S., Sattler, J. A. G., & Almeida-
Muradian, L. B. (2012c). Antioxidant activity of bee pollen produced
in two Brazilian states [Paper presentation]. 5th European
Conference of Apidology, Halle an Der Saale. EurBee 5 – 5th
European Conference of Apidology (p. 223).
Melo, A. A. M., Meira, D. F. S., Sattler, J. A. G., Almeida-
Muradian, L. B. (2012d). Antioxidant activity of dehydrated
bee pollen produced in Rio Grande do Sul state, Brazil. In
II International Symposium on Bee Products/Annual Meeting of
IHC, Braganc¸a/Book of Abstracts (p. 75).
Melo, A. A. M., Nagai, M. K., Chan, J., & Almeida-Muradian,
L. B. (2013). Chemical composition, total phenolic com-
pounds and antioxidant activity of dehydrated bee pollen
produced in four Brazilian states. In XXXXIII International
Apicultural Congress. Apimondia – Scientific Program/Posters.
Menezes, J. D. S., Maciel, L. F., Miranda, M. S., & Druzian, J. I.
(2010). Compostos bioactivos e potencial antioxidante de
p
olen ap
ıcola producido por abelhas africanizadas (Apis mel-
lifera L.). Revista do Instituto Adolfo Lutz, 69(2), 233–242.
Molyneux, P. (2004). The use of the stable free radical diphenyl-
picryl-hydracyl (DPPH) for estimating antioxidant activity.
Songklanakarin Journal of Science and Technology, 26, 211–219.
Montenegro, G., Pizarro, R., Mej
ıas, E., & Rodr
ıguez, S. (2013).
Evaluaci
on biol
ogica de polen ap
ıcola de plantas de Chile.
Revista Internacional de Bot
anica Experimental, 82, 7–14.
Morais, M., Moreira, L., Fe
as, X., & Estevinho, L. M. (2011).
Honeybee-collected pollen from five Portuguese Natural Parks:
Palynological origin, phenolic content, antioxidant properties
and antimicrobial activity. Food and Chemical Toxicology, 49(5),
1096–1101. https://doi.org/10.1016/j.fct.2011.01.020
Moreira, L., Dias, L. G., Pereira, J. A., & Estevinho, L. (2008).
Antioxidant properties, total phenols and pollen analysis of
propolis samples from Portugal. Food and Chemical Toxicology,
46(11), 3482–3485. https://doi.org/10.1016/j.fct.2008.08.025
Moreti, A. C. C. C., Marchini, L. C., Souza, V. C., & Rodrigues,
R. R. (2002). Atlas do p
olen de plantas ap
ıcolas. Papel Virtual.
Moya, M., Koch, M., Hostetler, L. (1993). One-class classifier
networks for target recognition applications. In Proceedings
on World Congress on Neural Networks, International Neural
Network Society (pp. 797–801).
Mullins, J., & Emberlin, J. (1997). Sampling pollens. Journal of
Aerosol Science, 28(3), 365–370. https://doi.org/10.1016/
S0021-8502(96)00439-9
Mußmann, P., Levsen, K., & Radeck, W. (1994). Gas-chromato-
graphic determination of phenols in aqueous samples after
solid phase extraction. Fresenius' Journal of Analytical Chemistry,
348(10), 654–659. https://doi.org/10.1007/BF00325568
Muszynska, J., & Warakomska, Z. (1999). Honey bees in the
monitoring of the environment’s contamination: I. The
numerousness of the monitoring apiary giving samples of
pollen loads, which are representative for a neighborhood
rich in pollen flow. Pszczelnicze scientific notebooks –
1999. – Rok XLIII, pp. 197–208.
Nagai, T., Inoue, R., Inoue, H., & Suzuki, N. (2002). Scavenging
capacities of pollen extracts from Cistus ladanifeus on aut-
oxidation, superoxide radicals, hydroxyl radicals, and DPPH
radicals. Nutrition Research, 22(4), 519–526. https://doi.org/
10.1016/S0271-5317(01)00400-6
Nagai, T., Nagashima, T., Suzuki, N., & Inoue, R. (2005).
Antioxidant activity and angiotensin I-converting enzyme
inhibition by enzymatic hydrolysates from bee bread.
Zeitschrift fur Naturforschung. C, Journal of Biosciences, 60(1-
2), 133–138. https://doi.org/10.1515/znc-2005-1-224
Nagy, M., Krizkov
a, L., Mucaji, P., Kontsekov
a, Z., Sersen, F.,
& Krajcovic, J. (2009). Antimutagenic activity and radical
scavenging activity of water infusions and phenolics from
ligustrum plants leaves. Molecules, 14(1), 509–518. https://
doi.org/10.3390/molecules14010509
Nati
c, M. M., Dabi
c, D., C; Papetti, A., Fotiri
c Aksi
c, M. M.,
Ognjanov, V., Ljubojevi
c, M., & Tesi
c, Z L. J. (2015). Analysis and
characterisation of phytochemicals in mulberry (Morus alba L.)
fruits grown in Vojvodina, North Serbia. Food Chemistry, 171,
128–136. https://doi.org/10.1016/j.foodchem.2014.08.101
Negri, G., Teixeira, E., Alves, M. L. T. F., Moreti, A. C. C. C.,
Otsuk, I. P., Borguini, R. G., & Salatino, A. (2011).
Hydroxycinnamic acid amide derivatives, phenolic compounds
and antioxidant activities of extracts of pollen samples from
Southeast Brazil. Journal of Agricultural and Food Chemistry,
59(10), 5516–5522. https://doi.org/10.1021/jf200602k
Niki, E., & Noguchi, N. (2000). Evaluation of antioxidant cap-
acity. What capacity is being measured by which method?
IUBMB Life, 50(4-5), 323–329. https://doi.org/10.1080/
15216540051081119
Nogueira, C., Iglesias, A., Feas, X., & Estevinho, L. M. (2012).
Commercial bee pollen with different geographical origins: A com-
prehensive approach. International Journal of Molecular Sciences,
13(9), 11173–11187. https://doi.org/10.3390/ijms130911173
No^zkov
a, J., Brindza, J., Ostrovsk
y, R., Stehl
ıkov
a, B., &
Fatrcov
a-Sramkov
a, K. (2010). Hodnotenie kvantitat
ıvnych a
kvalitat
ıvnych znakov vcelieho pel'u a ich klasifik
acia podl'a
navrhnut
ych deskriptorov: Evaluation of quantitative and
qualitative traits of bee pollen and their classification
according to proposed list of descriptors. In Potravin
arstvo.
Zdruzenie HACCP Consulting 2007: Nitrianske
Hrnciarovce, Slovensko. 4, 204–216.
No^zkov
a, J., Ostrovsk
y, R., Brindza, J., & Moln
arov
a, E. (2010).
Morfologick
a charakteristika vcel
ıch pel'ov
ych obno^zok. In
B. Brovarskyi (Ed.), Vcel
ı obno^zkov
y pel'. SPU v Nitre (pp.
101–113).
No^zkov
a, J., Pavelek, M., Bjelkov
a, M., Brutch, N., Tejklov
a, E.,
Porokhovinova, E., & Brindza, J. (2011). Descriptor List for
Flax – Linum usitatissimum L., Nitra (p. 104).
Oh, S. M., & Chung, K. H. (2006). Antiestrogenic activities of
Ginkgo biloba extracts. The Journal of Steroid Biochemistry
104 M. G. Campos et al.
and Molecular Biology, 100(4-5), 167–176. https://doi.org/10.
1016/j.jsbmb.2006.04.007
Ohta, S., Fujimaki, T., Uy, M. M., Yanai, M., Yukiyoshi, A., & Hirata,
T. (2007). Antioxidant hydroxycinnamic acid derivatives iso-
lated from Brazilian bee pollen. Natural Product Research, 21(8),
726–732. https://doi.org/10.1080/14786410601000047
Oliveira, K. C. L. S., Moriya, M., Azedo, R. A. B., Almeida-
Muradian, L. B., Teixeira, E. W., Alves, M. L. T. M. F., &
Moreti, A. C. C. C. (2009). Relationship between botanical
origin and antioxidants vitamins of bee-collected pollen.
Qu
ımica Nova, 32(5), 1099–1102. https://doi.org/10.1590/
S0100-40422009000500003
Orsoli
c, N., Kosalec, I., & Basi
c, I. (2005). Synergistic antitu-
mor effect of polyphenolic components of water soluble
derivative of propolis against Ehrlich ascites tumour.
Biological & Pharmaceutical Bulletin, 28(4), 694–700. https://
doi.org/10.1248/bpb.28.694
Ostrosky, E. A., Mizumoto, M. K., Lima, M. E. L., Kaneko,
T. M., Nishikawa, S. O., & Freitas, B. R. (2008). M
etodos
para avaliac¸~ao da actividade antimicrobiana e determinac¸~ao
da concentrac¸~ao m
ınima inibit
oria (CMI) de plantas medici-
nais. Revista Brasileira de Farmacognosia, 18(2), 301–307.
https://doi.org/10.1590/S0102-695X2008000200026
Otero, M. P., & Losada, S. B. (2002). M
etodo de determi-
naci
on del origen geogr
afico del polen ap
ıcola comercial.
Lazaroa, 23, 25–34.
Otsu, N. (1979). A threshold selection method from grey-
level histograms. IEEE Transactions on Systems, Man, and
Cybernetics, 9(1), 62–66. https://doi.org/10.1109/TSMC.1979.
4310076
Ou, B., Hampsch-Woodill, M., & Prior, R. R. (2001).
Development and validation of an improved oxygen radical
absorbance capacity assay using fluorescein as the fluores-
cent probe. Journal of Agricultural and Food Chemistry, 49(10),
4619–4626. https://doi.org/10.1021/jf010586o
Ozcan, M., Ceylan, D. A., Unver, A., & Yetisir, R. (2003).
Antifungal effect of pollen and propolis extracts collected from
different regions of Turkey, Uludag. Bee Journal, 3, 27–34.
Ozdal, T., Capanoglu, E., & Altay, F. (2013) A review on pro-
tein-phenolic interactions and associated changes. Food
Research International, 51, 954–970. https://doi.org/10.1016/j.
foodres.2013.02.009.
Pabla, N., Huang, S., Mi, Q.-S., Daniel, R., & Dong, Z. (2008).
ATR-Chk2 Signaling in p53 Activation and DNA Damage
Response during Cisplatin-induced Apoptosis. The Journal of
Biological Chemistry, 283(10), 6572–6583. https://doi.org/10.
1074/jbc.M707568200
Panetta, J. C. (1998). O car
ater educativo da Vigil^ancia
Sanit
aria. Higiene Alimentar, 12(57), 8–10.
Panteli
c, M., Dabi
c, D., Matijasevi
c, S., Davidovi
c, S.,
Dojcinovi
c, B., Milojkovi
c-Opsenica, D., Tesi
c, Z., & Nati
c,
M. (2014). Chemical characterization of fruit wine made
from Oblacinska sour cherry. The Scientific World Journal,
2014, 454797–454799. https://doi.org/10.1155/2014/454797
Partap, U. (1997). Bee flora of the Hindu Kush-Himalayas:
Inventory and management (p. 297). International Centre for
Integrated Mountain Development.
Pascoal, A., Rodrigues, S., Teixeira, A., Fe
as, X., & Estevinho,
L. M. (2014). Biological activities of commercial bee pollens:
Antimicrobial, antimutagenic, antioxidant and anti-inflamma-
tory. Food and Chemical Toxicology, 63, 233–239. https://doi.
org/10.1016/j.fct.2013.11.010
Paulino, N., Abreu, S. R., Uto, Y., Koyama, D., Nagasawa, H.,
Hori, H., Dirsch, V. M., Vollmar, A. M., Scremin, A., Bretz,
W. A. (2008). Anti-inflammatory effects of a bioavailable
compound, Artepillin C, in Brazilian propolis. European
Journal of Pharmacology, 10;587(1–3), 296–301. doi: 10.1016/
j.ejphar.2008.02.067. Epub 2008 Feb 29. PMID: 18474366.
Paulo, L., Antunes, P., Campos, M. G., & Anjos, O. (2012).
Utilizac¸~ao do Teor em Metais Pesados no P
olen Como
Marcador Ambiental – Estudo Preliminar. II Congresso
Ib
erico de Apicultura, 18-20 Setembro, Guadalajara, pp. 75–76.
Paulo, L., Antunes, P., & Anjos, O. (2014). Mineral compos-
ition of pollen using inductively coupled plasma atomic
emission spectroscopy. Planta Medica, 80(16), 2–16. https://
doi.org/10.1055/s-0034-1394851
Petersen, J., Souza, E. M. P., Moreira, R. M., Pasin, L. E. V.,
Nordi, J. C., & Barreto, L. M. R. C. (2011). Comercializac¸~ao
do p
olen ap
ıcola em 11 pa
ıses da America Latina. Magistra,
Cruz das Almas-BA, 23(n
umero especial), 14–16.
Pichichero, E., Cicconi, R., Mattei, M., & Canini, A. (2011).
Chrysin-induced apoptosis is mediated through p38 and
Bax activation in B16-F1 and A375 melanoma cells.
International Journal of Oncology, 38(2), 473–483. https://doi.
org/10.3892/ijo.2010.876
Pinto, B., Bertoli, A., Noccioli, C., Garritano, S., Reali, D., &
Pistelli, L. (2008). Estradiol-antagonistic activity of phenolic
compounds from leguminous plants. Phytotherapy Research,
22(3), 362–366. https://doi.org/10.1016/j.fct.2013.11.010.
Pinto, B., Caciagli, F., Riccio, E., Reali, D., Sari
c, A., Balog, T.,
Liki
c, S., & Scarpato, R. (2010). Antiestrogenic and antige-
notoxic activity of bee pollen from Cystus incanus and Salix
alba as evaluated by the yeast estrogen screen and the
micronucleus assay in human lymphocytes. European Journal
of Medicinal Chemistry, 45(9), 4122–4128. https://doi.org/10.
1016/j.ejmech.2010.06.001
Plumb, R., Castro-Perez, J., Granger, J., Beattie, I., Joncour, K.,
& Wright, A. (2004). Ultra-performance liquid chromatog-
raphy coupled to quadrupole-orthogonal time-of-flight mass
spectrometry. Rapid Communications in Mass Spectrometry,
18(19), 2331–2337. https://doi.org/10.1002/rcm.1627
Prakash, A. S., Pereira, T. N., Reilly, P. E. B., & Seawright, A. A.
(1999). Pyrrolizidine alkaloids in human diet. Mutation Research/
Genetic Toxicology and Environmental Mutagenesis, 443(1-2),
53–76. https://doi.org/10.1016/S1383-5742(99)00010-1
Prior, R. L., Wu, X., & Schaich, K. (2005). Standardized meth-
ods for the determination of antioxidant capacity and phe-
nolics in foods and dietary supplements. Journal of
Agricultural and Food Chemistry, 53(10), 4290–4302. https://
doi.org/10.1021/jf0502698
Prokudina, E. A., Havl
ıcek, L., Al-Maharik, N., Lapc
ık, O.,
Strnad, M., & Gruz, J. (2012). Rapid UPLC–ESI–MS/MS
method for the analysis of isoflavonoids and other phenyl-
propanoids. Journal of Food Composition and Analysis, 26(1-2),
36–42. https://doi.org/10.1016/j.jfca.2011.12.001
Qian, W. L., Khan, Z., Watson, D. G., & Fearnley, J. (2008). Analysis
of sugars in bee pollen and propolis by ligand exchange chroma-
tography in combination with pulse amperometric detection and
mass spectrometry. Journal of Food Composition and Analysis,
21(1), 78–83. https://doi.org/10.1016/j.jfca.2007.07.001
Rabie, A. L., Wells, J. D., & Dent, L. K. (1983). The nitrogen con-
tent of pollen protein. Journal of Apicultural Research, 22(2),
119–123. https://doi.org/10.1080/00218839.1983.11100572
Rauha, J. P., Vuorela, H., & Kostiainen, R. (2001). Effect of eluent on
the ionization efficiency of flavonoids by ion spray, atmospheric
pressure chemical ionization, and atmospheric pressure photo-
ionization mass spectrometry. Journal of Mass Spectrometry,
36(12), 1269–1280. https://doi.org/10.1002/jms.231
R Core Team. (2014). R: A language and environment for statis-
tical computing. R Foundation for Statistical Computing.
http://www.R-project.org/
Reiter, R. (1947). The coloration of anther and corbicular pol-
len. The Ohio Journal of Science, 17(4), 137–152. http://hdl.
handle.net/1811/3589
Rex, J. H., Pfaller, M. A., Walsh, T. J., Chaturvedi, V., Espinel-
Ingroff, A., Ghannoum, M. A., Gosey, L. L., Odds, F. C.,
Rinaldi, M. G., Sheehan, D. J., & Warnock, D. W. (2001).
Standard methods for pollen research 105
Antifungal susceptibility testing: Practical aspects and cur-
rent challenges. Clinical Microbiology Reviews, 14(4), 643–658.
https://doi.org/10.1128/CMR.14.4.643-658.2001
Royal Horticultural Society (RHS). 1966, 1986, 1995, 2004.
RHS Color Chart. Royal Horticultural Society.
Ricciardelli, D. G. (1982). Analisi organolettico-microscopica
dei prodotti apistici: un problema attuale e indilazionabile.
L'Ape Nosta Amica, 4, 19–21.
Richardson, R. T., Lin, C.-H., Sponsler, D. B., Quijia, J. O.,
Goodell, K., & Johnson, R. M. (2015). Application of ITS2 meta-
barcoding to determine the provenance of pollen collected by
honey bees in an agroecosystem. Applications in Plant Sciences,
3(1), 1400066. https://doi.org/10.3732/apps.1400066
Robards, K., & Antolovich, M. (1997). Analytical chemistry of
fruit bioflavonoids. The Analyst, 122(2), 11R–34R. https://doi.
org/10.1039/a606499j
Rochat, B. (2012). Quantitative/qualitative analysis using LC-
HRMS: the fundamental step forward for clinical laborato-
ries and clinical practice. Bioanalysis, 4(14), 1709–1711.
https://doi.org/10.4155/bio.12.159
Rodlauer, R., & H€uttinger, E. (1992). A Computer-supported
method of honey pollen analysis. Abstracts of the 8th
International palynological Congress, France. p. 40.
Rodr
ıguez, I., Llompart, M. P., & Cela, R. (2000). Solid-phase
extraction of phenols. Journal of Chromatography. A, 885(1-
2), 291–304. https://doi.org/10.1016/S0021-9673(00)00116-3
Rodr
ıguez-Dami
an, M., Cernadas, E., Formella, A., Fern
andez-
Delgado, M., & S
a-Otero, P. D. (2006). Automatic detection
and classification of grains of pollen based on shape and
texture. IEEE Transactions on Systems, Man, and Cybernetics,
Part C: Applications and Reviews, 36, 531–542. https://doi.org/
10.1109/TSMCC.2005.855426
Rodr
ıguez-Dami
an, M., Cernadas, E., Formella, A., S
a-Otero,
P. (2004). Pollen classification using brightness-based and
shape-based descriptors. IEEE Transactions on System, Man,
and Cybernetics.
Rodr
ıguez-Dami
an, M., Formella, A., & S
a-Otero DE, P. (2004
Pollen classification using brightness-based and shape-based descrip-
tors [Paper presentation]. Proceedings of the 17th International
Conference on Pattern Recognition, ICPR 2004, Vol. 2, pp.
212–215. https://doi.org/10.1109/ICPR.2004.1334098
Rodr
ıguez-Delgado, M. A., Malovan
a, S., P
erez, J. P., Borges,
T., & Garc
ıa-Montelongo, F. J. (2001). Separation of phen-
olic compounds by high-performance liquid chromatography
with absorbance and fluorimetric detection. Journal of
Chromatography. A, 912(2), 249–257. https://doi.org/10.1016/
S0021-9673(01)00598-2
R€osemann, G. M., Botha, C. J., & Eloff, J. N. (2014).
Distinguishing between toxic and non-toxic pyrrolizidine
alkaloids and quantification by liquid chromatography–mass
spectrometry. Phytochemistry Letters, 8, 126–131. https://doi.
org/10.1016/j.phytol.2014.03.002
Rosenthal, C., Petrescu, A., & Caragiani, S. (1969). Cercetari
ascupra compozitiei biochimice si valorii biologice a unui
produs inlocuitor de pollen. Apicultura, 22(8), 18–21.
Rouessac, F., & Rouessac, A. (2007). Sample preparation: Solid phase
extraction (SPE). In Francis, A. Rouessac, & S. Brooks (Trans.),
Chemical analysis – Modern instrumentation methods and techniques
(2nd ed., chapter 21 pp. 487–490). JohnWiley & Sons, Ltd.
Rzepecka-Stojko, A., Pilawa, B., Ramos, P., & Stojko, J. (2012).
Antioxidative properties of bee pollen extracts examined
by EPR spectroscopy. Journal of Apicultural Science, 56(1),
23–30. https://doi.org/10.2478/v10289-012-0003-0
Rzepecka-Stojko, A., Stec, M., Kurzeja, E., Gawro
nska, E.,
Pawlowska-Goral, K. (2012). The effect of storage of bee
pollen extracts on polyphenol content. Polish Journal of
Environmental Studies, 21(4), 1007–1011. http://www.pjoes.
com/pdf/21.4/Pol.J.Environ.Stud.Vol.21.No.4.1007-1011.pdf
Sahin, H., Aliyazicioglu, R., Yildiz, O., Kolayli, S., Innocenti, A.,
& Supuran, C. T. (2011). Honey, pollen, and propolis
extracts show potent inhibitory activity against the zinc
metalloenzyme carbonic anhydrase. Journal of Enzyme
Inhibition and Medicinal Chemistry, 26(3), 440–444. https://
doi.org/10.3109/14756366.2010.503610
Salamanca, G., Moreno, C., Quiralte, J., Barber, D., Villalba,
M., & Rodriguez, R. (2008). Immunological characterisation
of Ole e 11, a major olive pollen allergen. Allergy, 63, 500.
S
anchez, E. T. (2004). Polen: Producci
on Envasado y Comercializaci
on.
S
anchez-Moreno, C. (2002). Methods used to evaluate the free
radical scavenging activity in foods and biological systems. Food
Science and Technology International, 8(3), 121–137.
S
anchez-Moreno, C., Larrauri, J. A., & Saura-Calixto, F. (1998). A
procedure to measure the antiradical efficiency of polyphenols.
Journal of the Science of Food and Agriculture, 76(2), 270–276.
https://doi.org/10.1002/(SICI)1097-0010(199802)76:2< 270::AID-
JSFA945> 3.0.CO;2-9
Santana, C. M., Ferrera, Z. S., Torres Padr
on, M. E., & Santana
Rodr
ıguez, J. J. (2009). Methodologies for the extraction of
phenolic compounds from environmental samples: New
approaches. Molecules, 14(1), 298–320. https://doi.org/10.3390/
molecules14010298
Sattler, J. A. G. (2013). Quantificatio of antioxidant vitamins E
(alpha-, beta-, gama- and delta-tocopherol), C (ascorbic acid),
provitamin A (alpha- and beta-carotene) and chemical compos-
ition of dehydrated bee pollen produced in georeferenced apia-
ries of southern Brazil [Master dissertation]. University of
S~ao Paulo. Retrieved July 7, 2014, from http://www.teses.
usp.br/teses/disponiveis/9/9131/tde-18032014-151137/
Sawyer, R. (1981). Pollen identification for beekeepers (p. 112).
University College Cardiff Press.
Sawyer, R. (1988). Honey identification (p. 115). Cardiff
Academic Press.
Schieber, A., Keller, P., Streker, P., Klaiber, I., & Carle, R.
(2002). Detection of isorhamnetin glycosides in extracts of
apples (Malus domestica cv. "Brettacher") by HPLC-PDA
and HPLC-APCI-MS/MS. Phytochemical Analysis, 13(2),
87–94. https://doi.org/10.1002/pca.630
Scorzoni, L., Benaducci, T., Almeida, A. M. F., Silva, D. H. S.,
Bolzani, V. S., & Gianinni, M. J. (2007). The use of standard
methodology for determination of antifungal activity of natural
products against medical yeasts Candida sp. and Cryptococcus
sp. Brazilian Journal of Microbiology, 38, 391–397. https://doi.org/
10.1590/S1517-83822007000300001.
Serra-Bonveh
ı, J., & Escol
a Jord
a, R. (1997). Nutrient compos-
ition and microbiological quality of honey bee-collected pol-
len in Spain. Journal of Agricultural and Food Chemistry, 45(3),
725–732. https://doi.org/10.1021/jf960265q
Serra-Bonveh
ı, J., Gonello Galindo, J., & Gomez Pajuelo, A.
(1986). Estudio de la composition y caracteristicas fisico-
quimicas del pollen de abejas. Alimentaria, 23(176), 63–67.
Serra-Bonveh
ı, J., Torrent
o, S. M., & Lorente, C. E. (2001). Evaluation
of polyphenolic and flavonoid compounds in honey bee –
Collected pollen produced in Spain. Journal of Agricultural and Food
Chemistry, 49(4), 1848–1853. https://doi.org/10.1021/jf0012300
Sharma, M. (1970). An analysis of pollen loads of honey bees
from Kangra, India. Grana, 10(1), 35–42. https://doi.org/10.
1080/00173137009429855
Sharma, O. P., & Bhat, T. K. (2009). DPPH antioxidant assay
revisited. Food Chemistry, 113(4), 1202–1205. https://doi.org/
10.1016/j.foodchem.2008.08.008
Shaw, D. E. (1990). The incidental collection of fungal spores by bees
and the collection of spores in lieu of pollen. Bee World, 71(4),
158–176. https://doi.org/10.1080/0005772X.1990.11099059
Shi, H. B., Kong, M., Chen, G., Zhao, J., Shi, H. L., Chen, Y., &
Rowan, F. G. (2010). Compound pollen protein nutrient
increases serum albumin in cirrhotic rats. Gastroenterology
Research, 3(6), 253–261. https://doi.org/10.4021/gr240e
106 M. G. Campos et al.
Shoskes, D. A. (2002). Phytotherapy in chronic prostatitis.
Urology, 60(6 Suppl), 35–37. https://doi.org/10.1016/S0090-
4295(02)02383-X
Sickel, W., Ankenbrand, M. J., Grimmer, G., Holzschuh, A.,
H€artel, S., Lanzen, J., Steffan-Dewenter, I., & Keller, A.
(2015). Increased efficiency in identifying mixed pollen sam-
ples by meta-barcoding with a dual-indexing approach. BMC
Ecology, 15, 20. https://doi.org/10.1186/s12898-015-0051-y
Siddik, Z. (2003). Cisplatin: mode of cytotoxic action and
molecular basis of resistance. Oncogene, 22(47), 7265–7279.
https://doi.org/10.1038/sj.onc.1206933
Silici, S., & Kutluca, S. (2005). Chemical composition and antibac-
terial activity of propolis collected by three different races of
honeybees in the same region. Journal of Ethnopharmacology,
99(1), 69–73. https://doi.org/10.1016/j.jep.2005.01.046
Silva, E. N. (1999). Tecnologia de alimentos. Higiene na
ind
ustria de alimentos. Unicamp- Campinas. 5.
Silva, J. A. (2000). T
opicos da tecnologia de alimentos. S~ao
Paulo: Livraria Varela, 229. 82 pp.
Silva, T. M. S., Camara, C. A., da Silva Lins, A. C., Maria Barbosa-
Filho, J., da Silva, E. M. S., Freitas, B. M., & de Assis Ribeiro dos
Santos, F. (2006). Chemical composition and free radical scav-
enging activity of pollen loads from stingless beeMelipona subni-
tuda Ducke. Journal of Food Composition and Analysis, 19(6-7),
507–511. https://doi.org/10.1016/j.jfca.2005.12.011
Silva, T. M. S., Camara, C. A., Lins, A. C. S., Agra, M. F., Silva,
E. M. S., Reis, I. T., & Freitas, B. M. (2009). Chemical com-
position, botanical evaluation and screening of radical scav-
enging activity of collected pollen by the stingless bees
Melipona rufiventris (Uruc¸u-amarela)). Anais da Academia
Brasileira de Ciencias, 81(2), 173–178. https://doi.org/10.
1590/s0001-37652009000200003
Silva, C. I., Bordon, N. G., da, R., Filho, L. C., & Gar
ofalo,
C. A. (2012). The importance of plant diversity in maintain-
ing the pollinator bee, Eulaema nigrita (Hymenoptera:
Apidae) in sweet passion fruit fields. Revista de Biologia
Tropical, 60(4), 1553–1565.
Smirnova, V. A., Timofeyev, N. K., Breygina, M. A., Matveyeva,
N. P., & Yermakov, P. I. (2012). Antioxidant properties of
the pollen exine polymer matrix. Biophysics, 57(2), 174–178.
https://doi.org/10.1134/S0006350912020224
Snyder, R. D., & Gillies, P. J. (2002). Evaluation of the clasto-
genic, DNA intercalative, and topoisomerase II-interactive
properties of bioflavonoids in Chinese hamster V79 cells.
Environmental and Molecular Mutagenesis, 40(4), 266–276.
https://doi.org/10.1002/em.10121
Solberg, Y., & Remedios, G. (1980). Chemical composition of
pure and bee-collected pollen. Scientific Reports Agricultural
University, Norway, 59(18), 1–13.
Somerville, D. C. (2001). Nutritional value of bee collected pollens, a
report for the Rural Industries Research and Development
Corporation # 01/047 (pp. 57–58). NSW Agriculture Goulburn.
Somerville, D. C., & Nicol, H. I. (2002). Mineral content of
honey bee collected pollen from southern New South
Wales. Australian Journal of Experimental Agriculture, 42(8),
1131–11136. https://doi.org/10.1071/EA01086
Soto-Vaca, A., Gutierrez, A., Losso, J. N., Xu, Z., & Finley, J. W.
(2012). Evolution of phenolic compounds from color and flavor
problems to health benefits. Journal of Agricultural and Food
Chemistry, 60(27), 6658–6677. https://doi.org/10.1021/jf300861c
Souza, B. R. (2014). Quantification of B complex vitamins (B1,
B2) and vitamers of vitamin B3 and B6 in dehydrated bee pol-
len samples from Southern Brazil [Master dissertation].
University of S~ao Paulo.
Souza, E. H., Souza, F. V. D., Rossi, M. L., Packer, R. M., Cruz-Barros,
M. A. V., & Martinelli, A. P. (2018) Pollen morphology and viability
in Bromeliaceae. Annals of the Brazilian Academy of Sciences, 89(4),
3067–3082. http://dx.doi.org/10.1590/0001-3765201720170450.
Sparkman, O. D. (2000). Mass spectrometry desk reference.
Global View Pub. http://www.amazon.com/exec/obidos/
ASIN/0966081323/102-8754827-3860108
Stalikas, C. D. (2007). Extraction, separation, and detection
methods for phenolic acids and flavonoids. Journal of
Separation Science, 30(18), 3268–3295. https://doi.org/10.
1002/jssc.200700261
Stanciu, O. G., Marghitas, L. A., Dezmirean, D., & Campos,
M. G. (2011). A comparison between the mineral content
of flower and honey bee collected pollen of selected plant
origin (Helianthus annuus L. and Salix sp.). Romanian
Biotechnological Letters, 16(4), 6291–6296.
Stanciu, O. G., Marghitas, L. A., & Dezmirean, D. (2012).
Nutritional and biological values of bee pollen and beebread
harvested from Romania. In Proceedings of International
Conference of Agricultural Engineering CIGR-AGENG. http://
journals.usamvcluj.ro/index.php/zootehnie/article/viewFile/
9391/7820
Stanciu, O. G., Marghitas, L. A., Dezmirean, D., & Campos, M. G.
(2012). Specific distribution of minerals in selected unifloral
bee pollen. Food Science and Technology Letters, 3(1), 27–31.
Stanley, R. G., & Linskens, H. F. (1974). Pollen – Biology, bio-
chemistry, management (p. 307). Springer-Verlag. https://doi.
org/10.1007/978-3-642-65905-8
Szcze^sna, T. (2006a). Long – Chain fatty acids composition of
honey bee-collected pollen. Journal of Apicultural Science, 50,
65–78.
Szcze^sna, T. (2006b). Protein content and amino acid compos-
ition of bee-collected pollen from selected botanical origins.
Journal of Apicultural Science, 50, 81–90.
Szcze^sna, T. (2006c). Protein content and amino acids compos-
ition of bee-collected pollen originating from Poland, South
Korea and China. Journal of Apicultural Science, 50, 91–99.
SzczeRsna, T. (2007). Study on the sugar composition of honey-
bee-collected pollen. Journal of Apicultural Science, 51(1), 15–21.
SzczeRsna, T., & Rybak-Chmielewska, H. (1998). Some properties of
honey bee collected pollen. In Polnisch-Deutsches Symposium
Salus Apis Mellifera, new demands for honey bee breeding in the
21st century. Pszczelnicze Zeszyty Naukowe, 42(2), 79–80.
SzczeRsna, T., Rybak-Chmielewska, H., & Chmielewski, W.
(2002). Sugar composition of pollen loads harvested at dif-
ferent periods of the beekeeping season. Journal of
Apicultural Science, 46(2), 107–115.
SzczeRsna, T., Rybak-Chmielewska, H., & Skowronek, W.
(1995a). Wpływ utrwalania na warto
s
c biologiczna˛ obn
o_zy
pyłkowych. Pszczelnicze Zeszyty Naukowe, 39(1), 177–187.
SzczeRsna, T., Rybak-Chmielewska, H., & Skowronek, W. (1995b).
Zmiany w składzie chemicznym obn
o_zy pyłkowych zachodza˛ce
podczas ich przechowywania w r
o_znych warunkach. I. Cukry,
tłuszcze, popi
oł. Pszczelnicze Zeszyty Naukowe, 39(2), 145–156.
Taha, E. A. (2015). Chemical composition and amounts of
mineral elements in honey bee-collected pollen in relation
to botanical origin. Journal of Apicultural Science, 59(1),
75–81. https://doi.org/10.1515/JAS-2015-0008
Tax, D. (2011). DDtools, the data description toolbox for
Matlab. Version 1.9.0.
Thakur, M., & Nanda, V. (2020). Composition and functionality
of bee pollen: A review. Trends in Food Science & Technology,
98, 82–106. https://doi.org/10.1016/j.tifs.2020.02.001
Tian, S., Nakamura, K., Cui, T., & Kayahara, H. (2005). High-per-
formance liquid chromatographic determination of phenolic
compounds in rice. Journal of Chromatography. A, 1063(1-2),
121–128. https://doi.org/10.1016/j.chroma.2004.11.075
Tichy, J., & Novak, J. (2000). Detection of antimicrobials in
bee products with activity against viridans streptococci.
Journal of Alternative and Complementary Medicine, 6(5),
383–389. https://doi.org/10.1089/acm.2000.6.383
Tohamy, A. A., Abdella, E. M., Ahmed, R. R., & Ahmed, Y. K.
(2014). Assessment of anti-mutagenic, anti-histopathologic
Standard methods for pollen research 107
and antioxidant capacities of Egyptian bee pollen and prop-
olis extracts. Cytotechnology, 66(2), 283–297. https://doi.org/
10.1007/s10616-013-9568-0
Tripathi, S. K. M., Kumar, M., Kedves, M., & Varga, B. (2003).
LM, SEM and TEM investigations on partially degraded pol-
len grains of Cycas rumphii Miq. from India. Plant Cell
Biology and Development, Szeged, 15, 28–42.
Uhl, M., Ecker, S., Kassie, F., Lhoste, E., Chakraborty, A.,
Mohn, G., & Knasm€uller, S. (2003). Effect of chrysin, a fla-
vonoid compound, on the mutagenic activity of 2-amino-1-
methyl-6-phenylimidazo[4,5- b]pyridine (PhIP) and benzo(a)-
pyrene (B(a)P) in bacterial and human hepatoma (HepG2)
cells. Archives of Toxicology, 77(8), 477–484. https://doi.org/
10.1007/s00204-003-0469-4
Ulusoy, E., & Kolayli, S. (2014). Phenolic composition and anti-
oxidant properties of anzer bee pollen. Journal of Food
Biochemistry, 38(1), 73–82. http://.dx.doi.org/10.1111/jfbc.
12027 https://doi.org/10.1111/jfbc.12027
Valentini, A., Miquel, C., & Taberlet, P. (2010). DNA barcoding
for honey biodiversity. Diversity, 2(4), 610–617. https://doi.
org/10.3390/d2040610
van Laere, O., Lagasse, A., & de Mets, M. (1969). Use of the scan-
ning electron microscope for investigating pollen grains iso-
lated from honey samples. Journal of Apicultural Research, 8(3),
139–145. https://doi.org/10.1080/00218839.1969.11100230
van Meeuwen, J. A., Korthagen, N., De Jong, P. C., Piersma, A. H.,
& van den Berg, M. (2007). (Anti)estrogenic effects of phyto-
chemicals on human primary mammary fibroblasts, MCF-7 cells
and their co-culture. Toxicology and Applied Pharmacology,
221(3), 372–383. https://doi.org/10.1016/j.taap.2007.03.016
van Rijsbergen, C. (1979). Information retrieval. Butterworths.
Villanueva, G. (1999). Pollen resources used by European and
Africanized honey bees in the Yucatan peninsula. Mexico.
Journal of Apicultural Research, 38(1-2), 105–111. https://doi.
org/10.1080/00218839.1999.11101001
Villanueva, M. T. O., D
ıaz Marquina, A., Bravo Serrano, R., &
Blazquez Abell
an, G. (2002). The importance of bee-col-
lected pollen in the diet: A study of its composition.
International Journal of Food Sciences and Nutrition, 53(3),
217–224. https://doi.org/10.1080/09637480220132832
Vorwohl, G. (1990). Progress, problems and future tasks of
Melissopalynologie. Apidologie, 21(5), 383–389. (Article in
German). https://doi.org/10.1051/apido:19900501
Wang, L., & Weller, C. L. (2006). Recent advances in extraction of
nutraceuticals from plants. Trends in Food Science & Technology,
17(6), 300–312. https://doi.org/10.1016/j.tifs.2005.12.004
Wang, Q., Garrity, G. M., Tiedje, J. M., & Cole, J. R. (2007).
Naive Bayesian classifier for rapid assignment of rRNA
sequences into the new bacterial taxonomy. Applied and
Environmental Microbiology, 73(16), 5261–5267. https://doi.
org/10.1128/AEM.00062-07
Wang, S., Xie, B., Yin, L., Duan, L., Li, Z., Eneji, A. E., Tsuji,
W., & Tsunekawa, A. (2010). Increased UV-B radiation
affects the viability, reactive oxygen species accumulation
and antioxidant enzyme activities in maize (Zea mays L.)
pollen. Photochemistry and Photobiology, 86(1), 110–116.
https://doi.org/10.1111/j.1751-1097.2009.00635.x
Wang, W., Hu, J., Cheng, J. (1987). Biological effect of pollen
from beehives radioprotective effect on hematopoietic tis-
sues of irradiated mice. In Proceedings of the 31st
International Apicultural Congress Apimondia, p. 176.
Wasek, M., Nartowska, J., Wawer, I., & Tudruj, T. (2001).
Electron spin resonance assessment of the antioxidative poten-
tial of medicinal plants. Part 1. Contribution of anthocyanosides
and flavonoids to the radical scavenging ability of fruits and
herbal teas. Acta Poloniae Pharmaceutica, 58(4), 283–288.
Wiegand, I., Hilpert, K., & Hancock, R. E. W. (2008). Agar and
broth dilution methods to determine the minimal inhibitory
concentration (MIC) of antimicrobial substances. Nature
Protocols, 3(2), 163–175. https://doi.org/10.1038/nprot.2007.521
Wilkins, T. D., & Thiel, T. (1973). Modified broth-disk method
for testing the antibiotic susceptibility of anaerobic bacteria.
Antimicrobial Agents and Chemotherapy, 3(3), 350–356.
https://doi.org/10.1128/AAC.3.3.350
Williams, N. M., & Kremen, C. (2007). Resource distributions
among habitats determine solitary bee offspring production
in a mosaic landscape. Ecological Applications, 17(3),
910–921. https://doi.org/10.1890/06-0269
Witten, I. H., & Frank, E. (2005). Data mining: Practical machine
learning tools and techniques. (2nd ed.). Morgan Kaufmann.
https://doi.org/10.1016/b978-0-12-374856-0.00023-7
Xiong, A., Yang, L., He, Y., Zhang, F., Wang, J., Han, H., Wang, C.,
Bligh, S. W. A., & Wang, Z. (2009). Identification of metabolites
of adonifoline, a hepatotoxic pyrrolizidine alkaloid, by liquid
chromatography/tandem and high-resolution mass spectrom-
etry. Rapid Communications in Mass Spectrometry, 23(24),
3907–3916. https://doi.org/10.1002/rcm.4329
Yamaguchi, M., Uchiyama, S., & Nakagawa, T. (2007).
Preventive effects of bee pollen Cistus iadaniferus extract
on bone loss in ovariectomized rats in vivo. Journal of
Health Science, 53(5), 571–575. https://doi.org/10.1248/jhs.
52.268[10.1248/jhs.53.571]
Y
a~nez, Y. A., Remsberg, C. M., Takemoto, J. K., Vega-Villa, K. R.,
Andrews, P. K., Sayre, C. L., Yang, K., Wu, D., Liu, D., Chen, J.,
& Sun, P. (2013). Characterization of chemical composition of
bee pollen in China. Journal of Agricultural and Food Chemistry,
61(3), 708–718. https://doi.org/10.1021/jf304056b
Yang, W. Z., Qiao, X., Bo, T., Wang, Q., Guo, D. A., & Ye, M.
(2014). Low energy induced homolytic fragmentation of flavo-
nol 3-O-glycosides by negative electrospray ionization tandem
mass spectrometry. Rapid Communications in Mass Spectrometry,
28(4), 385–395. https://doi.org/10.1002/rcm.6794
Yerlikaya, O. (2014). Effect of bee pollen supplement on anti-
microbial, chemical, rheological, sensorial properties and pro-
biotic viability of fermented milk beverages. Mljekarstvo, 64(4),
268–279. https://doi.org/10.15567/mljekarstvo.2014.0406
Youssef, A. M., Farag, R. S., Ewies, M. A., & El-Shakaa, S. M. A.
(1978). Chemical studies on pollen collected by honey bees
in Giza region, Egypt. Journal of Apicultural Research, 17(3),
110–113. https://doi.org/10.1080/00218839.1978.11099914
Yu, L. (2001). Free radical scavenging properties of conjugated
linoleic acids. Journal of Agricultural and Food Chemistry,
49(7), 3452–3456. https://doi.org/10.1021/jf010172v
Zhang, X., Habib, F. K., Ross, M., Burger, U., Lewenstein, A.,
Rose, K., & Jaton, J. (1995). Isolation and characterization of a
cyclic hydroxamic acid from a pollen extract, which inhibits
cancerous cell growth in vitro. Journal of Medicinal Chemistry,
38(4), 735–738. https://doi.org/10.1021/jm00004a019
Appendix: authors
1. Introduction
Maria Grac¸a R Campos, Of
elia Anjos and Norman L Carreck
2. Producing bee pollen from harvest to storage
Lidia M R C Barreto, Maria G R Campos and Jo~ao C Nordi
3. Inexpensive pollen load authentication by means of
computer vision and machine learning
Manuel Chica and Pascual Campoy
4. Evaluation and classification of corbicular pol-
len morphology
Janka Nozkova
108 M. G. Campos et al.
5. Identification of the floral origin of bee pollen
by microscopy
Olena Lokutova, Maria G R Campos, Ananias Pascoal, Let
ıcia
M Estevinho
6. Phenolic and polyphenolic profiles as fingerprints of
floral origin of bee pollen
Maria G R Campos, Norma Almaraz-Abarca, L
ıgia Muradian-
Bicudo and Of
elia Anjos
7. Determination of phenolic profile of bee pollen
extracts by LC and LC/MS
Zivoslav Lj Tesi
c, Mirjana D Mosi
c, Aleksandar Z Kosti
c,
Mirjana B Pesi
c, Dusanka M Milojkovi
c-Opsenica
8. Standard method for identification of bee pollen
mixtures through meta-barcoding
Wiebke Sickel, Markus J Ankenbrand, Gudrun Grimmer,
Frank F€orster, Ingolf Steffan-Dewenter and Alexander Keller
9. Standard method for the determination of protein
content of pollen using the Kjeldahl method
Chrysoula H Tananaki, Vasilios Liolios, Dimitrios Kanelis,
Maria-Anna Rodopoulou and Andreas Thrasyvoulou
10. Standard method for determination of lipid con-
tent of pollen using Soxhlet extraction
Chrysoula H Tananaki, Vasilios Liolios, Dimitrios Kanelis,
Maria-Anna Rodopoulou and Andreas Thrasyvoulou
11. Determination of the content of sugars in pollen,
using HPLC-RID
Teresa SzczeRsna
12. Standard method for vitamin analysis in bee pollen
Ligia Bicudo de Almeida-Muradian
13. Standard methods for mineral and trace elements
in bee pollen
Of
elia Anjos, Lu
ısa Paulo and Maria Grac¸a Campos
14. Standard methods for pyrrolizidine alkaloid profil-
ing in pollen by UHPLC-HRMS
Matteo A Lucchetti, Christina Kast and Gaetan Glauser
15. Standard methods for total phenolic compounds,
total flavonoids and antioxidative capacity bee pollen
evaluation by FTIR-ATR Spectroscopy
Of
elia Anjos and Let
ıcia M Estevinho
16. Standard methods for determination of antioxi-
dant bioactivity of bee pollen
Norma Almaraz-Abarca, Ligia Bicudo de Almeida-Muradian
and Maria G Campos
17. Standard methods for anti-inflammatory activities
in bee pollen
Ananias Pascoal, Vanessa Branco Paula and Let
ıcia M Estevinho
18. Standard methods for antimicrobial activities in
bee pollen
Ananias Pascoal, Vanessa Branco Paula and Let
ıcia
M Estevinho
19. Determination of standard methods for anti-muta-
genic activities in bee pollen
Ananias Pascoal, Vanessa Branco Paula and Let
ıcia
M Estevinho
Standard methods for pollen research 109