Using palynological evidence from royal jelly to mediate the spread of Paenibacillus larvae in Brazil

1. Instituto de Botânica, Núcleo de Pesquisa em Palinologia, Avenida Miguel Stéfano, 3.687, 04301-902 São Paulo, SP, Brasil 2. Universidade Federal de Viçosa, Pós-gradução em Entomologia, Departamento de Entomologia, 36570-900 Viçosa, MG, Brasil 3. Universidade Federal de Viçosa, Departamento de Biologia Geral, 36570-900 Viçosa, MG, Brasil 4. Universidade Federal Rural do Semi-Árido, Departamento de Ciências Animais, 59625-900 Mossoró, RN, Brasil 5. Empresa Brasileira de Pesquisa Agropecuária, Embrapa Gado de Leite, 36038-330 Juiz de Fora, MG, Brasil 6. Agência Paulista de Tecnologia dos Agronegócios, Polo Regional Vale do Paraíba, 12400-970 Pindamonhangaba, SP, Brasil 7. Universidade de Vigo, Departamento de Biologia Vegetal e Ciências do Solo, Ourense, Espanha 8. Corresponding author: cyluz@yahoo.com ABSTRACT (Using palynological evidence from royal jelly to mediate the spread of Paenibacillus larvae in Brazil). Royal jelly may contain pollen grains and their presence can be used to determine the phytogeographical origin of the product. This study analyzed the phytogeographical origin of commercial royal jelly samples from São Paulo State, tested as part of the Brazilian Federal Inspection System (SIF), found to be contaminated with spores of the bacterium Paenibacillus larvae, that causes the American Foulbrood Disease. The pollen grains of Castanea had the highest total percentage, with lower percentages of Cirsium/Carduus, Cistus, Parthenocissus, Prunus, Quercus, Robinia, Scrophulariaceae, Taraxacum, Tilia, among others. This pollen spectrum is incompatible with royal jelly samples produced in Brazil. The pollen spectrum resembled that of an imported product, compatible with the Northern Hemisphere origin. Brazilian legislation does not require the phytogeographic origin of imported bee products to be analyzed by palynological procedures, but it is mandatory to have a certifi cate issued by the country of origin attesting the absence of pathogens, monitored with the objective of preventing the exotic diseases from entering Brazil. Palynology, therefore, proved to be fundamental in detecting imports of this contaminated batches.


Introduction
A decline in the honey bee population has been reported in various countries over the past decade by the FAO (United Nations Food and Agriculture Organization) and the NAS (U.S. National Academy of Sciences), including populations in Brazil.This is a matter of great concern due to the decline in pollination of important food crops (Klein et al. 2007).The honey bee mortality observed is not seasonal and there seems to be a complex interaction between various factors, including insecticide application by farmers as well as fragmentation and degradation of habitats (Johnson 2010).However, the data are still imprecise and the causes are undefined due to the absence of standards and/or weak monitoring (Pires et al. 2016).
Although honey bees are attacked by many parasites and pathogens (Chen et al. 2004), Africanized bees are more tolerant and resistant to several pathogens.However, in recent years there has been a sudden weakening of swarms in various places in the southeast of Brazil, with mortality of adult bees and the appearance of abnormal symptoms in broods (Message et al. 2012).
American foulbrood disease, caused by the bacterium Paenibacillus larvae, one of the main cause of losses of colonies in the world (Funfhaus et al. 2013).This pathogen contaminates colonies through spore ingestion by larvae (Garrido-Bailón et al. 2013), causing their death, consequently a decrease in the colony population.The bacterium P. larvae is exotic in Brazil and American foulbrood disease was only detected twice in the country.The disease was detected for the first time in 2001 in the municipality of Candelária, Rio Grande do Sul State, where samples of adult bees, honey, pollen and honeycombs were collected and analyzed (Schuch et al. 2003).Spores of the bacterium were detected in three groups of adult bees and in honeycombs, as well as in imported honey and pollen samples.In 2006, clinical signs of the disease were reported in samples from Quatro Barras, Paraná State.This bacterium was quickly controlled and extinguished in the country (MAPA 2006).
Exotic diseases can enter new territories by natural migration of bees, or by importing queen bees or bee products contaminated with pathogens (Hitchcock & Revell 1963, Gochnauer & Corner 1974, Hale & Menapace 1980, Flores et al. 2005, Higes et al. 2008, Pettis et al. 2013, Fries et al. 2013, OIE 2014, Guimarães-Cestaro et al. 2016).Diseases caused by microorganisms are of particular concern due to the easy transmission of resistant spores, which can remain viable for decades (Hornitzky 2010).For example, spores of P. larvae can remain viable for about 70 years in the environment (Shimanuki & Knox 1994, 2000).
Inside the bee colony, products such as honey, pollen, and royal jelly are susceptible to contamination by these pathogens, mainly by being stored in comb cells by the bees and transferred through trophallaxis (OIE 2014).Due to the high risk of introducing exotic pathogens into the country by imported bees' products and queens, sanitary control is regulated by a specific resolution of the Brazilian Ministry of Agriculture.Among the requirements are documents from the country of origin that include the results of diagnostic tests.
Royal jelly is a product secreted by the hypopharyngeal glands of worker honeybees (Apis mellifera).It is derived mostly from the nutrients contained in the pollen grains ingested by the bees.Pollen is the main source of proteins, peptides and secondary metabolites in royal jelly.Therefore, pollen is the main precursor product that results in royal jelly and substantially influences its biological properties (Stocker 2005).Although it does not make sense to refer to the botanical origin of royal jelly, because it is a bee secretion, observing the pollen types and their percentages (pollen spectra) can provide some evidence regarding the phytogeographical origin of a royal jelly sample.Pollen flow into the product is an accidental process, introduced from the internal environment of the colony, during manipulation in the mouthparts of the adult bees when feeding queen larvae or adult queens.As the royal jelly is not stored in the hive, it can be found pollen grains that happened to be in the beesʼ bodies or in their nectariferous vesicles (Simpson 1955, Renner et al. 2003, Barth 2005).
Pollen grains analysis in bee products is an effective tool to support the certification of their phytogeographical origin because it shows the presence of pollen types that are characteristic of specific plants.However, according to Brazilian legislation, this analysis is not mandatory for imported or local bee products.
This study evaluated the phytogeographical origin of royal jelly samples sold in São Paulo State, Brazil, tested by the Federal Inspection System (Sistema de Inspeção Federal -SIF) in which positive signs of Paenibacillus larvae were previously detected based on PCR results.The results of these analyses are useful for elucidating the possible dispersion of this pathogen in Brazil through the sale and distribution of bee products.

Material and methods
Eight royal jelly samples, which were inspected by the S.I.F., were analyzed using palynological techniques.They were purchased in São Paulo State, Brazil and were bottled from 2012 to 2014.Positive signs of Paenibacillus larvae were detected according to molecular analyses (Guimarães-Cestaro et al. 2016, Teixeira et al. 2018).
The palynological treatment of the royal jelly followed Barth (2005) using 1 g of the product for the analysis and preparation of the microscope slides which include sample dilution, centrifugation, resuspension in acetic acid for overnight dehydration, and subsequent acetolysis (Erdtman 1952).Three slides of each sample were prepared using standard uncolored and fuchsine stain colored (one slide) glycerine jelly (Barth 2005).Fuchsin stained pollen grains were not photographed.
The identification of the pollen types was based on the European and Asian pollen flora catalogs (Punt & Clarke 1974, Lieux 1975-1977-1978, Göschl 2008, Haiqing 2015), among many publications that were consulted and that are part of the collection from the Palynological Research Centre (NPP) at the Institute of Botany, as well as others obtained from the Internet.The "Northwest European Pollen Flora" (Punt & Clarke 1974) is the world´s most detailed pollen morphological study and covers around 130 families and ca.600 species.
In the present study was adopted as counting methodology the observation of all pollen grains contained in the three slides.Pollen grains were identified as "pollen types" meaning that a single plant species, group of species or higher taxa could present with similar pollen morphologies.
In order to evaluate the phytogeographical origin of each sample, the percentage calculation representing all the pollen types per sample (Total Pollen Sum) was used.The pollen types were grouped according to their relative frequencies, following (Louveaux et al. 1978), corresponding to: Predominant pollen (P) = >45% of the total pollen grains counted; Secondary pollen (S) = 16% to 45%; Important Minor Pollen (I) = 3% to 15% and, Minor pollen (M) = <3%.
Although quantification by Lycopodium tracer spores provides a context for pollen spectra that is both useful for supporting phytogeographic origins of bee products from botanical source (bee pollen and honey) and, for later comparisons with similar studies, they were not utilized in this study to collect pollen data, due to the fact that royal jelly is a question of bee secretion.The pollen concentration calculation in royal jelly has no use whatsoever, as it does not come from botanical excretions.In addition, pollen concentration calculations using Lycopodium tracer spores were only carried out in royal jelly in Barth (2005) and Morgado & Barth (2011) studies, and even so, without an important discussion about it.Lycopodium tracer spores were not used in royal jelly in the studies conducted by Chauvin & Louveaux (1956), Ricciardelli D'Albore et al. (1977), Ricciardelli-D'Albore & Battaglini (1978), Biondi et al. (2003), Piana et al. (2006), Dimou et al. (2007), Dimou et al. (2013).Another reason for not having done concentration calculation and saturation curve is that this is a product manipulated for commercialization, without possibility to control the field production, as observed in other studies, like e.g., in Dimou et al. (2007), and there is no way of knowing if these samples suffered some type of filtering before being commercialized.Filtering could have artificially reduced the pollen quantity.Finally, in forensic terms, you never add something to a pollen sample and especially not other spores or pollen.
The similarity between the samples was analyzed using a principal component analysis (PCA) to examine the influence of quantitative and qualitative data on the ordination of the pollen types.The matrix included the absolute value of all pollen types counted in each sample.The FITOPAC software Shepherd (1996) was used for the transformation of the absolute count using the natural logarithm [log (x + 1)] and the ordination was subsequently performed using the covariance matrix in the PC-ORD 5.0 software (Mccune & Mefford 1999).The variability between samples was expressed using the first two axes of the PCA.
The MINITAB 15 software (Minitab Inc., State College, PA, 2003) was used to construct a dendrogram of the percent of similarity between the samples.
The samples were evaluated in two ways to identify its origin: 1) among samples, to verify its similarity, including all the pollen types and their absolute value and, 2) with those from other countries, to compare the pollen types found in each sample together with the presence or absence in royal jelly samples from Brazil (Barth 2005, Morgado & Barth 2011), China (Piana et al. 2006), France (Stocker 2005), Greece (Dimou et al. 2013) and Italy (Piana et al. 2006) as well as Louisiana honey, United States (Lieux 1975(Lieux -1977(Lieux -1978)).This allowed investigating if a possible illegal event occurred that could pose serious environmental risks due to the possibility of dispersing this exotic pathogen in Brazil.
The pollen types were grouped in a table showing their probable geographical origins, habitus (form) of the plant, vernacular names in Portuguese and in English, the presence or absence in the samples and a possible indication of the Brazilian extra origin.
The image captures of pollen grains were obtained using an OLYMPUS BX 50 light microscope with a video camera Olympus U-CMAD-2 and the program CellSens Standard® 1.5.for Windows.

Results and Discussion
We identified 78 pollen types representing 48 genera and 38 families of plants, in eight royal jelly samples (table 1).In addition, there were 16 unidentified pollen types.Some samples contained only very few pollen grains as observed also by Dimou et al. (2013) in Greece.
No samples showed pollen spectra compatible with Brazilian biomes or degraded areas by anthropic action of the country (table 2).
Several pollen types found in the samples were from cultivated plants for food, medicinal or ornamental purposes in many parts of the world (Allium, Apium, Brassicaceae, Daucus, Lilium, Taraxacum, Zea mays, among others).Pollen grains from these plants are not good indicators of phytogeographical origin as they can occur in various continents.
The pollen spectra observed differed from those reported by Barth (2005) and Morgado & Barth (2011) for royal jelly samples produced in Brazil, which include a predominance of Cecropia, Asteraceae and Eucalyptus in the city of Viçosa (Minas Gerais State) and Cecropia, Eucalyptus, Melastomataceae and Mimosa pudica/M.scabrella in Pindamonhangaba (São Paulo State), among other pollen grains with lower frequencies.These groups are very distinct from those of the Northern Hemisphere.The pollen spectra of the samples from Pindamonhangaba are similar to those from Vietnam (Piana et al. 2006) due to the presence of Mimosa pudica and Eucalyptus, but with different percentages.Another differentiating factor was the presence of other pollen types that only occur in either one of the countries.
The pollen variability among the samples comprised 69.68% of the two first axes in the PCA (figure 1).The first axis summed for 53.85% of the variance due to the high values of Castanea, Brassicaceae, Daucus and Taraxacum pollen types which were found in seven samples, as well as Parthenocissus and Veronica, in six and five samples, respectively (table 3).
Among the main pollen grains, those from Castanea were found in all analyzed samples and Predominant (P) in samples GR1, GR5, GR8 and GR10 whereas in samples GR2 and GR9 it was Secondary (S).However, in samples GR3 and GR4 it occurred as an Important minor pollen (I) (table 1).
The chestnut is a deciduous tree native to temperate regions of the Northern Hemisphere (North America, Asia and Europe), but it is cultivated in Brazil in some places in São Paulo State, where there are commercial orchards established as long as 20 years ago in Arapeí, Campinas, Campos de Jordão, Cunha, Itupeva, Itapecerica da Serra, Mogi Mirim, Piedade, Pindamonhangaba, Santo Antônio do Pinhal, São Bento do Sapucaí, São José dos Campos and Taubaté (Bueno & Pio 2014).In Minas Gerais State, the commercial orchards are located in the southern region, in Brazópolis, Conceição dos Ouros, Espírito Santo do Dourado, Pouso Alegre, Jacutinga and Gonçalves (Bueno & Pio 2014).However, as several other pollen types observed in all samples did not belong to Brazilian plants, it can be claimed that all samples analyzed were imported.
Other important pollen types for PCA ordination that did not occur frequently in the samples were

GR9
In Brazil there is the native T. argentinense Speg.
whose pollen is similar.In tropical and subtropical regions.His greatest diversity is found in the Americas, followed by Africa, Australia and Asia.

GR1
Several native species in Brazil.

GR1, GR3, GR4
Several native species in Brazil.Several species are cultivated worldwide to food purposes, also in Brazil.

Ulmus (Ulmaceae) tree elms/olmo
Constitute natural forests in temperate regions of Eurasia and North America, but with tropical distribution in Indonesia.
It is cropped for various purposes: its wood is used in carpentry and furniture in the shipbuilding industry; in Australasia is used with ornamental purposes, both in parks and in bonsai; its foliage are widely used in animal feed in Asia; as a source of biomass in Europe and for medicinal purposes.

Parthenocissus
(Vitaceae) climber virginia creeper, five-finger, thicket creeper, grape ivy/ hera Native to the Himalayas, East Asia and North America.Often used as ornamental plants in Europe and elsewhere, including in the tropics.GR1, GR3, GR5, GR8, GR9, GR10 Exotic.Few species are cultivated worldwide to ornamental purposes, also in Brazil.2).Some pollen types that had less importance in the PCA ordination of the samples, but that can be good indicators of a Northern Hemisphere origin were Larix, Rumex, Sambucus, Tilia and Ulmus (figure 1).
The pollen types Nyssa and Mimosa pudica are good indicators of phytogeographical origins recovered in the present study.The observations of these pollen types demonstrate an Asian or North American origin as opposed to a European origin (Piana et al. 2006, Dimou et al. 2013).
The pollen detected from Nyssa ("tupelo") may be explained because trees of this genus are highly nectariferous and are valued by beekeepers due to the high quality of honey produced by bees that explore them.Nevertheless, pollen from species of this genus are generally underrepresented in honey samples (Lieux 1978).Some of the Nyssa species are native to eastern North America to Central America.Other species also occur naturally in Southeast Asia from China to the southeastern Himalaya Mountains, but they are lacking in Europe.The pollen grains of these two groups can be differentiated through scanning electron microscopy by the presence of tiny elongated sexine elements (rugulae) in the margins of the colpori of the Asian species, while the American ones do not show them (Göschl 2008).In light microscopy, it is not easy to distinguish them.This pollen type was found in samples GR1, GR5, GR9 and GR10 (table 2).Therefore, according to the total pollen spectrum of each sample, this pollen type can be an excellent indicator of the Asian or Northern American phytogeographical origin of royal jelly.
The presence of the Mimosa pudica pollen grains indicates tropical origin.This species is native to Central and South America, but because it is an invasive weed, it now also occurs in Africa, North America, Asia and Australia.
According to Piana et al. (2006), Mimosa pudica is one of the pollen types that occur in Asian royal jelly (specifically from Vietnam) but is absent in 28 samples from Italy, as well as in 82 royal jelly samples from Greece analyzed by Dimou et al. (2013).In the samples sold in São Paulo, this pollen type was only found in GR8, as an Important minor pollen (I) (table 1).This sample had 97.33% similarity with the group formed by samples GR1, GR5, GR9, and GR10, which are characterized by the presence of Nyssa pollen type and probably originated from Asia or North America (figure 2).
Among the frequent pollen types in sample GR4, Quercus and Fagopyrum esculentum occur in Brazil.Quercus is restricted to Paraná State, used for the cultivation of Shiitake mushrooms on oak logs (Bett & Perondi 2011).Buckwheat (Fagopyrum esculentum) was introduced in the south of the country during the 1920s by the Slavic immigrants (Gorgen 2013).However, the pollen spectrum of sample GR4 has no similarity with the pollen spectra of Brazilian royal jelly, honey, or bee pollen (Barth 2004-2005, Morgado & Barth 2011).Therefore, the probable phytogeographical origin of this sample is Asia.
According to the PCA analysis, sample GR4, typically Asian, was similar to sample GR3.They were grouped based on the presence of Castanea, Brassicaceae, Daucus, Lotus corniculatus, Trifolium pratense, Poaceae, Prunus, Solanum and Ulmus.These pollen types are common in honey from Europe, Asia and North America (Lieux 1975-1977-1978, Louveaux 1970, Song et al. 2012).
The pollen spectra showing the presence of cultivated plants Brassicaceae, Castanea, Daucus and Trifolium pratense in samples GR3 and GR4 were also reported by Diaz-Losada et al. (1998) in Italian honey.These authors stated that the location of the apiary near a temperate forest is refl ected in Italian honey by the presence of Quercus and Ulmus pollen grains or by the abundance of pollen from prairie herbs such as Trifolium pratense and Lamiaceae or also near areas of moist soil due to the presence of pollen of plants adapted to that environment, e.g., Salix.
The royal jelly sample GR2 had a high percentage of pollen of cultivated plants (Brassicaceae and Castanea), similar to many others, but it was the only one with pollen of several species of cultivated plants as celery (Apium, Apiaceae) and Pleurospermum (Apiaceae), two pollen types of Fabaceae (Trifolium and Fabaceae 2), as well as Primulaceae, a family with several ornamental herbaceous plant species widely cultivated around the world.This sample had 88.42% similarity with the group formed by samples with pollen of Nyssa (GR1, GR5, GR9 and GR10) and Mimosa pudica (GR8) (fi gure 2), supposedly Asian or North American.As pointed out previously, these pollen types do not occur in European royal jelly.
The set with various pollen grains of exotic plants in the samples was similar to the group of European and Chinese royal jelly having 49.54% similarity with those of Brazilian origin (fi gures 3-12).This percentage was due to the presence of seven pollen types (Amaranthaceae/Chenopodiaceae, Borreria, Brassicaceae, Mimosa pudica/M.scabrella, Poaceae, Senecio and Solanum) reported in the Brazilian samples analyzed by Barth (2005) and by Morgado and Barth (2011) as well as in some of the samples analyzed in the present study.
Palynological evidence indicates that royal jelly samples contaminated with spores of P. larvae are sold in São Paulo State as Brazilian products, however, they were produced in the Northern Hemisphere.The fact that royal jelly samples from Europe analyzed by other authors have been found similar to those from China, due to the presence of some pollen types, suggests that palynology alone cannot accurately identify the continent or country of origin.No scientifi c specialty (plant anatomy, palynology, physicochemical) alone can solve problems of the origin of bee products and illegal imports of bee products.Thus, only fi eld observations of bees' visits to fl owers cannot defi ne the botanical origin of the fl oral resources used by bees.Beekeepers' empirical observations should be followed up by physicochemical, anatomical and palynological analyses of bee products, as these analyses broaden the knowledge about bee visits to fl owers.In order to be sure of the geographical origin of these contaminated royal jelly batches, public inspection agents should audit in the warehouses that packed the royal jelly about the origin of the contaminated product that is being marketed in São Paulo.The Brazilian Ministry of Agriculture had no records of imported royal jelly from Europe, North America or Asia related to the period of the corresponding lots.Thus, the contaminated lots were probably illegally imported, threatening national beekeepers.As royal jelly production in Brazil is no longer economically viable due to the illicit entry of royal jelly over the past decade from China at low prices, we suggest that the analyzed products are original without the mixture of royal jelly produced locally in Brazil.This is a matter of concern because the dispersion of P. larvae spores from imported products without sanitary inspection is associated with the following contamination routes: using of contaminated royal jelly to feed queen larvae; manipulating contaminated products and/or utensils used to handle healthy colonies; using contaminated wax as comb substrate; access of bees to remains of contaminated imported products discarded near apiaries; cross-contamination in warehouses (e.g., pollen dryers, where pollen grains of different origins are dehydrated) and supply of contaminated food supplements to healthy bees during the winter (especially bee pollen).
The possibility of P. larvae dispersion and contamination of healthy colonies is aggravated by traditional handling procedures such as transferring wax frames among colonies, using the same utensils in multiple bee hives and transferring colonies among apiaries, as well the natural trophallaxis among bees.
We stress the urgency of regulatory agencies taking appropriate measures at warehouses to verify the origin of the lots of royal jelly marketed and thus detect and avoid the illicit trade, as well as to implement risk analysis to identify critical points of the chain for traceability.

Conclusions
No royal jelly sample contaminated with Paenibacillus larvae had a pollen spectrum compatible with Brazilian vegetation, showing that the products analyzed were imported.The pollen types identified in the samples indicated probable phytogeographical origin from the Northern Hemisphere, either from Europe, Asia or North America.
By carrying out a palynological analysis, it was observed that the pollen grains in the royal jelly samples are from species occurring in temperate climates, such as boreal forests with conifers and milder climates, for example in the Mediterranean.Pollen grains from forage grasses, as well various crops were observed in the royal jelly samples showing similarity with the samples produced by Apis mellifera bees in most parts of the world.Therefore, these pollen types are not appropriate for the characterization of the phytogeographical origin of royal jelly.
The presence of P. larvae spores in bee products from Brazilian apiaries is an unexpected finding because the country is officially considered free from the disease.However, a previous study (molecular analysis) done for pathogen detection in the same samples showed that in São Paulo State royal jelly has spores of this bacterium.The lack of effective control in this commercialization, as proven in this study, can ruin Brazilian beekeeping if this pathogen is spread throughout the country and applied palynology in forensic science is helping in this battle.
Overall, our findings show that palynology is a powerful tool in identifying the phytogeographic origins of royal jelly from the Northern Hemisphere.Although the current legislation does not require bee products to be examined palynologically to verify phytogeographic origins, this analysis can help to control pathogens from spreading.In addition, we show the fragility of actions regarding the control of imported products and thus the urgent need for more effective rules and actions by sanitary authorities.

Figure 2 .
Figure 2. Dendrogram of similarity among the eight royal jelly samples with Paenibacillus larvae spores purchased in São Paulo State, Brazil.

Figure 3 .
Figure 3. Dendrogram of similarity among the eight royal jelly samples with Paenibacillus larvae spores purchased in São Paulo State (Brazil) and those ones from Minas Gerais and São Paulo States (Brazil), China, France, Greece and Italy.

Table 1 .
Frequencies of pollen types in the royal jelly samples with Paenibacillus larvae spores purchased in São Paulo State, Brazil, inspected by the S.I.F.Predominant pollen (P), >45% of the total pollen grains counted; Secondary pollen (S) = 16% to 45%.Obs: values are in percentage.The designations after pollen type name indicate the code used in the PCA.

Table 1 (
continuation) Table2.Geographical origin, life form, vernacular names, economic data of plant usage, and the main occurring environment in Brazil of the plants identifi ed at the genus level by means of the pollen types found in the royal jelly samples with Paenibacillus larvae spores purchased in São Paulo State, Brazil.Note: "Native" means a plant indigenous to a particular region, "Naturalized" means after being introduced into a region, develops without the need to be treated or cultivated under human intervention and "Exotic" means introduced from abroad, but not fully naturalized.Native and naturalized vegetation quoted by Flora do Brasil (Brazilian Flora).continue Figure 1.Principal Component Analysis performed with the pollen type counting variables in eight royal jelly samples with Paenibacillus larvae spores purchased in São Paulo State, Brazil.Codes for pollen types can be found in Table1.

Table 3 .
Pearson and Kendall correlation coefficients for pollen grains counting variables of the first and the second axis of PCA ordination in eight royal jelly samples with Paenibacillus larvae spores purchased in São Paulo State, Brazil.Acer negundo, Betula, Cistus, Ligustrum, Myosotis, Nyssa, Prunus, Quercus, Salix, Trifolium incarnatum, Zea mays, some Asteraceae and Fabaceae and some unidentified pollen types (table1).Many of these are indicators of the Northern Hemisphere origin of the samples (table