**Improvement and Selection of Honeybees Assisted by Molecular Markers**

Maria Claudia Colla Ruvolo-Takasusuki, Arielen Patricia Balista Casagrande Pozza, Ana Paula Nunes Zago Oliveira, Rejane Stubs Parpinelli, Fabiana Martins Costa-Maia, Patricia Faquinello and Vagner de Alencar Arnaut de Toledo

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/62426

**Abstract**

Royal jelly is an important apiarian product for honeybees and has been used as an important ingredient to human health and healthy life style. Because of its wide use, there is great demand in their production. As royal jelly is a secretion of the cephalic glands of bees and it is produced at a certain age of the workers, it is necessary to perform the selection of producing queens to increase the amount produced. The employment of molecular markers is a tool that can be used to identify the genotypes of the best producers. Among the molecular markers, one of them called MRJP3 (Major Royal Jelly Protein 3) has been used in the Program of Improvement of *Apis mellifera* Royal Jelly Producing (PIAMRJP), State University of Maringá, Brazil. This molecular marker has been efficient in genotyping queens' royal jelly producers. Combined with classical breeding studies, the selection of queens assisted by MRJP3 marker has allowed to keep the selected genotypes for royal jelly production since 2006 (10 years). In this chapter, we present the main aspects of royal jelly, the hypopharyngeal glands, the major proteins of royal jelly and how it can be used as molecular markers.

**Keywords:** *Apis mellifera*, MRJP3, microsatellite, royal jelly, honeybee queen

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

## **1. Introduction**

Genetic improvement in any organism has the objective of increasing the gene frequencies of the economic importance of loci to be selected in the population. In relation to bees, this means increasing the frequency of the number of colonies that produce above the average genera‐ tion from which the selection was made.

The production of royal jelly and honey production are the result of the combined work of the workers [1], and therefore the entire colony becomes a unit of selection, where the assessment of improved queens is carried out by production workers' progeny [1,2]. Royal jelly production studies allowed to observe considerable variation in its production by Africanized honeybees [3,4]. These results show the need for selection of queens [5].

Selection of bees with genotypes involves the use of improved queens' replacement techniques and instrumental insemination and for molecular marker-assisted selection. There are few data in the literature linking molecular markers for the production of royal jelly. The identification and characterization of several loci of the major royal jelly proteins (MRJPs) allowed using one of the loci *Mrjp3* as a molecular marker for selection of *Apis mellifera* queens Africanized. Early studies by selecting queens and genotyping the best producers began in 2006, in the apiary of the State University of Maringá, Brazil. The first study associating the MRJP3 marker with royal jelly production was realized by [6].

High variability in major royal jelly proteins (MRJPs), especially MRJP3 to contain microsa‐ tellite regions, indicated a great potential of using these proteins, particularly microsatellite regions occurring in the *Mrjp3* gene as a marker for selecting studies for the improvement of the production of royal jelly. Subsequently, other researches were conducted using classic improvement parameters such as MRJP3 marker. The results to date have shown that this marker is important to genotype producing arrays of royal jelly.

Thus, this chapter shows the importance of royal jelly to honey bees and to human health, the importance of improving assisted by molecular markers and the results obtained with the selection of royal jelly producing queens and genotyped for MRJP3.

## **2. Royal jelly**

Royal jelly is secreted by the mandibular and hypopharyngeal glands located at the head of honeybees [7]. Hypopharyngeal gland secretion has a clear, water-like consistency and is rich in protein, while the mandibular gland produces a white secretion with milky consistency [7, 8]. Royal jelly can be described as a viscous substance, white-yellowish or grayish white, slightly opalescent with a characteristic pungent odor, although not unpleasant or rancid (**Figure 1**) [9–10]. These glands have the highest growth rate and activity of worker nurses between days 10 and 14 [11–15]. The development of the glands can be influenced by internal factors of the colony such as offspring and population density and external factors such as foraging and enabling bees to adapt quickly to the colony [13,16–18].

**Figure 1** Queen cell with queen larvae of *A. mellifera* and royal jelly.

Royal jelly is a glandular secretion recognized for complex composition, containing minerals, proteins, amino acids, steroids, phenols, carbohydrates, vitamins, lipids, acetylcholine and other unknown substances [9,19]; it is important too in reproduction and development. Royal jelly is the larval food until the third day of development when it becomes the exclusive food of the queen throughout her life, guaranteeing fertility and increased longevity. From third day, worker larvae are fed a mixture of honey, pollen and water, known as brood food; drones receive food brood and royal jelly [20].

The average lifespan of queens of *A. mellifera* live is 1 to 2 years [21]; they become sexually mature 6 days after emergence, mate about 17 drones and store all of the sperm needed to fertilize eggs for the duration of their lifespan [22]. Few drones rear in the summer, but a slight rise in drone rearing occurs during swarming [20]. Queens can lay 1500–2000 eggs per day throughout their lives [23,24], depending on the needs of the hive and environmental factors, while a large number of workers (sterile females) are responsible for maintaining the hive.

Due to the fertility and longevity of queens, related to the exclusive feeding with royal jelly, studies have been conducted considering similar effects in humans. Some beneficial effects have been attributed to consumption of royal jelly, as elimination of physical and mental fatigue, appetite normalization activation of brain function, improved vision, increased resistance against viral infections and skin rejuvenation [10].

Owing to considerable amount of proteins, free amino acids, lipids, vitamins, sugars and bioactive substances such as 10-hydroxy-trans-2-decenoic acid and antibacterial protein 350 KDa proteins, royal jelly becomes an ingredient for various healthy foods [25]. Review carried out by [25] shows several studies have reported that the royal jelly exhibits beneficial physio‐ logical and pharmacological effects in mammals, including vasodilative and hypotensive activities, antihypercholesterolemic activity and antitumor activity.

## **3. Molecular marker Major Royal Jelly Protein 3**

Royal jelly contains from 12 to 15% crude protein consisting of soluble proteins in water and water-insoluble proteins. The fraction of soluble proteins of royal jelly produced by the hypopharyngeal and mandibular glands contains several major proteins with molecular weight between 47 and 80 kDa [26] besides a small amount of minor proteins such as antibiotics and peptides [27,28]. Those proteins constitute the main group of major royal jelly proteins. The MRJPs represent between 82 and 90% of the total proteins of larval jelly [19]. Some regions of MRJPs can be focused on amino acids rich in nitrogen, thus high levels of nitrogen would be stored in MRJPs. The availability of nitrogen can be critical to the rapid growth of young larvae, as well as for the development of the queen [29]. These observations support the hypothesis that MRJPs have an important role in the nutrition of bees [19].

Major royal jelly protein-3 can be visualized on denaturing SDS-PAGE electrophoresis in head extracts of worker nurses (10–14 days old) or royal jelly (**Figure 2**). The polymorphism was estimated by [30].

**Figure 2** Denaturating SDS-PAGE electrophoresis showing MRJP3 polymorphism in extracts of head of *A. mellifera* nurse. A, B, C, D, E = alleles. Source: Baitala et al. (2013).

MRJPs genes encoding a group of proteins that have a common evolutionary origin with Yellow proteins of *Drosophila melanogaster* [31]. Genome of *Drosophila* encodes at least seven family members of Yellow proteins [32], whose loci are involved in the larval pigmentation [33], unlike the MRJPs that have nutritional function of larvae.

The genes encoding key proteins of royal jelly began to be identified in studies [34] and [35]. After these pioneering studies, several studies have been published in order to identify and characterize new genes encoding the MRJPs proteins [19,31,36,37]. The availability of the complete genome of *Apis mellifera* [38] made it possible to identify new genes encoding proteins of the family MRJP [29].

Since the first study were identified nine proteins in MRJPs *A. mellifera* (MRJP1, MRJP2, MRJP3, MRJP4, MRJP5, MRJP6, MRJP7, MRJP8, and MRJP9) besides an incomplete polypeptide, MRJPψ, encoded by a pseudogene. The genes encoding these proteins are located on chromo‐ some 11 [29]. Classification of *A. mellifera* MRJPs has been performed based on the N-terminal sequences of purified protein and cDNA sequences available in the cDNA library.

Analysis using PCR and DNA sequencing showed that the different alleles of the gene encoding MRJP3 protein differ in length as a result of a varying number of repeating basic units in a region of the *Mrjp3* gene [31,36]. The authors attributed the polymorphism of these proteins is a consequence of the presence of a region with varied number of repetitive sequences in tandem (microsatellite). These markers are comprised of a variable number of identical sequences having from 15 to 100 base pairs, in tandem and repeated up to 50 times. The molecular differences in four types of MRJP3 have shown that the polymorphism of these proteins is linked to the size variability, which is determined genetically by bees from the same colony [19].

The *Mrjp3* is a polymorphic locus that has been identified by DNA sequencing five alleles and PCR analysis identified at least 10 alleles of different sizes [36]. This study also revealed a Mendelian inheritance and high variability of the genomic locus of MRJP3.

Although *A. mellifera* [19,29,31,35–37] and other bees of the genus *Apis* [39–43] having the MRJPs are characterized, data in the literature on the use of MRJPs as molecular markers for selection associated with the improvement of royal jelly production are still scarce.

## **4.** *Apis mellifera* **queens' selection using MRJP3 marker**

The genetic improvement has the aim to increase the frequencies of desirable genes of the loci of economic importance to be selected in a population [44]. Thus, the genetic breeding of bees has the goal to increase the frequency of the number of colonies that produce above the average generation from which the selection was made. Selection of honeybees with superior genotypes involves the use of improved queens replacement techniques, instrumental insemination and molecular marker-assisted selection [45].

Selection of queens is carried out by genetic evaluation, which depends on the estimation of the components of (co)variance and genetic parameters for identification of genetically higher bees. Royal jelly production evaluated by Bayesian inference had a heritability estimate of 0.27% acceptance, 0.10 for the production of royal jelly per colony and 0.55 per dome [5]. The analyses performed by these authors showed that selection of queens can increase the production of royal jelly by colony, larval acceptance and production of royal jelly by the dome, and the external factors can modify the gene expression of individuals.

However, there are few data in the literature associating molecular markers for the production of royal jelly. One of the first molecular studies carried out to obtain DNA markers related to production of royal jelly was performed by [46]. These authors reviewed a total of 96 alleles produced for 10 microsatellite loci and according to the observed allele frequency for some alleles, it was possible to identify seven alleles that can be used as markers bees producing large quantities of royal jelly. The use of molecular markers, particularly microsatellites, can contribute to detect polymorphisms that might be useful to identify colonies of bees with high productivity of royal jelly.

High variability of MRJPs proteins and especially the MRJP3 to contain microsatellite region shows a great potential to use MRJP family proteins as markers for selection of producing queens for improving the production of royal jelly. Use of MRJPs as molecular markers in studies of population genetics and as selection markers associated with the improvement of royal jelly production is still scarce. Some researches have shown that this molecular marker is efficient to be used in the selection of royal jelly-producing queens.

Africanized honeybees selected for royal jelly production showed high allelic variability for the locus *Mrjp3* (**Figure 3**), showing the potential of this marker for selection [6]. In this research analyses of multiple linear regressions with EPD (expected progeny differences) values for royal jelly production were performed. The variance analyses indicated that the *Mrjp3* repetitive region influenced the genetic value of queen's offspring for royal jelly production. The determination coefficient (R2) for the significant alleles of the repetitive region of *Mrjp3* indicated that 36.85% of the EPD variation is explained by the variation of *C*, *D* and *E* alleles. Authors concluded that the three alleles present a considerable genetic effect on the variation of royal jelly production.

**Figure 3** Molecular marker MRJP3. Number = *A. mellifera* DNA. A, B, C, D, E, F, G = *Mrjp3* alleles. M = molecular weight marker. Source: Baitala et al (2010).

Continuing the process of selection and the Program of Improvement of *Apis mellifera* Royal Jelly Producing (PIAMRJP), State University of Maringá, Brazil, alleles of the locus *Mrjp3* descendants queens, those selected by [6], were evaluated in 2011 [47]. Results showed that the royal jelly-producing queens had a high degree of genetic diversity and excess homozygous alleles. The highest frequencies were estimated for *Mrjp3 D* and *E* alleles 0.3357 and 0.3107, respectively, showing that the selection process of queens royal jelly producing these alleles are being maintained and only the *C* allele had a low frequency of 0.0321.

Results obtained by [47] confirm those obtained by [6], the locus *Mrjp3* and their alleles *C*, *D* and *E* influence the genetic value for producing royal jelly; however, the real role of MRJP3 these bees has not yet been identified. The sequencing of *Mrjp3* of *A. mellifera* Africanized alleles in PIAMRJP was performed [48]. Homology and identity of these sequences were compared with the sequences deposited in the database for *A. mellifera* (**Figure 4**). Alleles *Mrjp3* detected showed high identities with alleles deposited in BLAST system. Alleles *Mrjp3 C, D* and *E* are being maintained in the genome of the selected matrices queens.

High similarity among the *Mrjp*3 alleles analyzed and those described in other studies show that the *Mrjp3* locus is conserved among species and subspecies of *Apis*. Similar results were obtained by [40]. These authors found that there are high similarity sequences and intron-exon have the same structure between four species *A. mellifera, A. cerana, A. dorsata* and *A. florea.*

The selection of royal jelly-producing queens may be promoting a selection of these reproduction bees, can alter the genetic characteristics of a given population, can be influenced by the process of transmission of these genes generation to generation [49]. However, it is important to maintain a degree of genetic variation, which results in a larger potential response to selective improvement [50].

In addition to the continuous genotyping of royal jelly–producing queens to locus *Mrjp3*, we developed a study to see if the mitochondrial DNA (mtDNA) of Africanized bees *A. mellifera* maintained in the breeding program have African or European origin. This research was performed by [51], using matrices producing royal jelly.

Mitochondrial DNA was analyzed using the molecular marker PCR-RFLP with specific primers and restriction enzymes to European and African honey bees. Analyses were performed with workers' daughters of royal jelly-producing queens in 2013, seven years after the beginning of the PIAMRJP started in 2006. After this period of selection and analysis of genetic parameters, alleles *C, D and E* are being maintained in queens, evidencing the role in royal jelly production. Queens selected for royal jelly production showed predominance of African mtDNA; therefore, genes of maternal origin are African. Use of microsatellite markers and mtDNA can be used in bee improvement programs to ensure the genetic origin of queens and verify the efficiency of Program of Improvement of *Apis mellifera* Royal Jelly Producing [51].

The employment of molecular markers in selection programs and improvement of honey bees for royal jelly production is efficient because it allows keeping genotypes of interest to ensure the highest productivity of the hives. The microsatellite marker MRJP3 has shown good results as a tool to verify the genotypes of producing matrices, facilitating identification and mainte‐ nance of the hives in the apiary of the Program of Improvement of *Apis mellifera* Royal Jelly Producing.


**Figure 4** Alignment of sequences similar to the *Mrjp3 C* allele performed using ClustalW2 (EMBL-EBI); sequences in‐ clude *A. mellifera* major royal jelly protein mRNA, complete cds (GU434675.1); *A. mellifera* major royal jelly protein 3 (*Mrjp3*), mRNA (NM\_001011601.1); *A. mellifera carnica* major royal jelly protein 3 *(Mrjp3)* gene, complete cds (AY663104.1); and PREDICTED: *A. mellifera* major royal jelly protein 3-like (LOC727045), partial mRNA **(**XM\_001122757.2). "\*" = nucleotides identical in all of the aligned sequences. Source: Casagrande-Pozza (2011).

## **Author details**

Maria Claudia Colla Ruvolo-Takasusuki1\*, Arielen Patricia Balista Casagrande Pozza1 , Ana Paula Nunes Zago Oliveira1 , Rejane Stubs Parpinelli2 , Fabiana Martins Costa-Maia3 , Patricia Faquinello4 and Vagner de Alencar Arnaut de Toledo2

\*Address all correspondence to: mccrtakasusuki@uem.br

1 Biotechnology, Genetics and Cell Biology Department, State University of Maringá, Av. Colombo, Maringá, PR, Brazil.

2 Animal Science Department, State University of Maringá, Av. Colombo, Maringá, PR, Bra‐ zil.

3 Federal University of Technology – Paraná, Dois Vizinhos, PR, Brazil

4 Federal University of Goiás – Rodovia GO, Ceres, GO, Brazil

## **References**


## **Impacts of Pesticides on Honey Bees**

Francisco Sanchez-Bayo and Koichi Goka

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/62487

#### **Abstract**

This chapter focuses on the detrimental effects that pesticides have on managed honey bee colonies and their productivity. We examine first the routes of exposure of bees to agrochemicals used for crop protection and their application to crops, fate and contami‐ nation of water and plants around the fields. Most of the time, the exposure of bees to pesticides is through ingestion of residues found in the pollen and nectar of plants and in water. Honey bees are also exposed to pesticides used for the treatment of *Varroa* and other parasites. The basic concepts about the toxicity of the different kinds of pesticides are explained next. Various degrees of toxicity are found among agrochemicals, and emphasis is given to the classic tenet of toxicology, "the dose makes the poison," and its modern version "the dose and the time of exposure makes the poison." These two factors, dose and time, help us understand the severity of the impacts that pesticides may have on bees and their risk, which are analysed in the third section. Sublethal effects are also considered. The final section is devoted to some practical advice for avoiding adverse impacts of pesticides in beekeeping.

**Keywords:** residues, toxicity, exposure, sublethal effects, risk management

## **1. Introduction**

For centuries, beekeepers have been aware of the environmental conditions that help prosper their honey bee colonies: a diversity of flowers from trees, shrubs, the so-called weeds and even crop plants. A healthy, diverse floral environment has always been the recipe for a healthy, bumper honey production. Perhaps the only problems they faced were the occasional infec‐ tion by microorganisms, diseases and parasites that could kill the bees and their colonies [1] or the unpredictable vagaries of weather that could affect flower production on particular bad years.

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

In the past few decades, however, beekeepers have had to cope with a new threat to their business: agrochemical pesticides, which are scattered over large areas of crops, fruit groves, forests and other environments for the control of insect pests, weeds, vermin and plant diseases. There was no doubt, from the beginning, that chemical insecticides could represent a serious threat to bees for the simple reason that bees are insects and, therefore, susceptible to any poison designed to kill insect pests. Consequently, strict toxicity testing was and still is required before such chemicals can be registered for use in crop protection [2, 3], at least in developed countries. Despite these regulations, the number of managed honey bee (*Apis mellifera* L.) colonies in the United States declined from 6 million in 1947, when DDT was introduced in agriculture, to less than 3 million in 2010 [4]. Similar trends have been observed in Europe, where the number of apiaries declined 14% in Scandinavia and 25% in central Europe between 1985 and 2005, although they increased 13% in the Mediterranean countries in order to counteract the lower production in the north [5].

But, what about other pesticides, such as herbicides and fungicides? Could they also affect honey bee productivity? If the target of such chemicals is not the insects, many argued, they are probably safe to the bees. Research conducted in the past few years in countries with a long history of pesticide usage suggests differently. It is now acknowledged that the extensive and prolonged used of herbicides leads to a reduced diversity of flowering plants [6, 7] that inevitably affect the bees' colonies [8] and their productivity. Moreover, the combination of some fungicides with insecticides has been revealed more deadly to the bees than either chemical alone [9]. Lately, the indiscriminate use of acaricides in apiaries for the control of parasites, such as *Varroa destructor*, has added one more threat for the beekeepers, as these chemicals are also toxic—although to a lesser degree—to the honey bees. Not surprisingly, the colony collapse disorder (CCD) has been linked by some authors not only to parasites and diseases but also to pesticide usage [10].

In these circumstances, a new management approach is needed for successful beekeeping. Production of honey and wax is no longer dependent on the availability of flowers in the surrounding environment, but rather appears to be intimately linked to the quality of food that the bees collect. It is now clear that pesticide-contaminated flowers affect the health of the honey bee colonies to the extent that their productivity declines [11]. In order to better manage this situation, we must first understand how bees are exposed to pesticides and what are the consequences of such exposure for the health of the individual bees, the colony and their overall productivity.

## **2. Exposure of bees to agrochemicals**

Most insecticides are applied as sprays over the crop canopy, but sprays of herbicides and fungicides are usually applied directly on the soil before the planting of crops. In all these cases, droplets and dust from the applications can fall directly on the bees that fly across the treated fields or nearby because wind can carry the tiny droplets and dust particles hundreds of metres away from the crop [12]. A single droplet of insecticide may be sufficient to kill a bee because the spray solutions contain concentrated doses of these chemicals—this is the most common cause behind the bee incidents reported in the literature [13, 14]. Granular pesticides that are incorporated into soil (e.g., herbicides) have no direct exposure to bees.

The so-called systemic insecticides are usually applied as seed coatings. The treated seeds are introduced into the soil using pneumatic drilling planters, and the friction of the seeds in the machinery produces dust particles that are heavily loaded with the insecticides. These poisonous particles can also cause a great deal of mortality among bees, if they happen to be in the surroundings [15]. Systemic insecticides applied this way are taken up by the crop plants as they grow and their residues are present in all parts of the treated plant, including the flowers, pollen and nectar [16]. Not only the crop plants but also the weeds and bushes that grow in the vicinity are affected [17, 18] because they also take up small amounts of residues that spread through the soil through lateral water flow [19] or are contaminated through dust/ spray drift. In addition, some plants can produce guttation drops in the early hours of the morning (e.g. maize, strawberries), and systemic insecticides appear in such drops in elevated concentrations [20] that are capable of killing the bees.

Most of the time, the exposure of bees to pesticides is through ingestion of residues found in the pollen and nectar of contaminated plants, whether from the crop plants or from the weeds around the fields [21]. It is important to realise that bees forage everywhere they can and search for the most suitable flowers that produce pollen and nectar in abundance. Thus, some crops are more attractive than others; for example, the yellow flowers of canola (rape seed oil), sunflowers and many weeds that grow in and around the crops are more attractive to bees than the flowers of potato plants. Pesticide residues in pollen and nectar are taken by the forager bees to their colonies and remain in the beebread and honey for quite some time [22, 23]. These residues are then fed to the larvae and the queen, which are affected in similar ways as the forager bees.

In addition to food, bees also drink water to keep their body temperature under control [24]. Pesticide residues in soil eventually move into the water and appear in the streams, creeks and ponds of agricultural areas and beyond, which are thus contaminated with a mixture of agrochemicals [25]. Some water contamination is also due to drift from spray applications, particularly from insecticides [26, 27]. Honey bees, bumblebees and wild bees like to drink from puddles, irrigation ditches, ponds and streams, and if these waters are contaminated with pesticide residues, the forager bees ingest them as well [28].

Apart from the pesticides used in agricultural production, honey bees are also exposed to the acaricides used for the control of *Varroa* and other parasites. In this case, bees come in contact with the high residue levels present on the waxy cells of the comb [29], affecting mainly the developing larvae [30] and presumably the adult honey bees and the queen.

Given the enormous variety of agrochemicals used in crop production, it is not surprising that, to date, residues of 173 different compounds have been found in apiaries [21]. It should be realised that through the various routes of exposure to pesticides in the environment (**Figure 1**), bees are not threaten by one or two chemicals alone but by cocktails of many agricultural compounds.

**Figure 1.** Routes of exposure of bees to agricultural pesticides.

## **3. Toxicology of pesticides**

Pesticides are toxic chemicals with specific mode of action, meaning they are designed to specifically control a target group of organisms by interfering with particular metabolic pathways. Thus, insecticides and acaricides kill insects and mites by disrupting their neuronal activity, their moulting process or other specific metabolism of these arthropods; herbicides and algicides kill plants and algae by disrupting their photosynthetic capacities or the synthesis of essential organic compounds and fungicides kill fungi by inhibiting the formation of their cell membranes or another metabolism specific of these organisms. There are other kinds as well, like rodenticides that kill small mammals, bird repellents, etc. The term biocide is reserved for broad-spectrum poisons that kill any organism, mainly microbes, but also large animals.

The toxicity of each kind of pesticide, however, is not exclusive to the target group of organ‐ isms: other species that share similar metabolism are affected as well, although usually to a lesser degree. The potency of a pesticide to any species is defined by the dose of toxic chemical that is lethal to 50% of individuals of that species (LD50), and such dose varies from species to species. Doses lower than the LD50 are considered 'sublethal', but they can also cause mortality on a certain proportion of the species population, i.e., 20 or 30% of individuals may die. In general, sublethal doses cause toxic effects that do not kill the organisms but still affect their normal functioning and health. For example, exposure of bees to sublethal doses of neurotoxic insecticides may cause stress [31], paralysis or abnormal behaviours without killing the bees [32].

By their very nature, insecticides are the most toxic compounds to bees, whereas herbicides are largely innocuous (**Table 1**). Beekeepers should be wary of any insecticide application in the vicinity of their hives because spray drift could certainly inflict a heavy toll on the bees. Pesticide applicators are aware of this danger and, in many countries, are required to inform beekeepers before they apply insecticides to a crop [33]. Also, while acaricides are less toxic to bees than to the target parasites, excessive amounts of their residues in the combs may have unpleasant consequences for the health of the bees [34].



†Source: Footprint database (IUPAC). http://sitem.herts.ac.uk/aeru/iupac/

**Table 1.** Toxicity of common pesticides to bees (LD50 at 48 hours) by contact or oral exposure and their persistence in soil (half-life)

All animals, including bees, are endowed with detoxification mechanisms that transform and eliminate most toxic chemicals. Currently, the majority of organic pesticides are degradable either in the organisms themselves or in the environment. The exception is the organochlorine pesticides (e.g. insecticides like DDT and lindane), which are very persistent and recalcitrant. Because they were applied in large quantities in the past decades, their residues are still present —although at low levels—in the soils of many countries, even if nowadays are banned from use in agriculture. Due to their low solubility in water, organochlorine residues are not taken up by the plants growing in contaminated soils, and so they do not appear in the pollen or nectar of the flowers.

The persistence of pesticides is evaluated by their half-life (*t*1/2), which is defined as the time required for half the amount of a chemical to disappear from a medium, that is, water, soil, air or biological tissues. Half-lives longer than 90 days indicate that the pesticide may accumulate, since more than 5% of the amount applied will remain in the environment after 1 year [35]. Residues of persistent pesticides found in pollen or nectar (Table 1) will, therefore, remain in the beebread throughout the entire season of honey production.

Systemic insecticides, such as neonicotinoids (e.g. imidacloprid) and fipronil, are more toxic and persistent than the majority of organophosphorus (e.g. malathion), carbamates (e.g. car‐ bofuran) and pyrethroids (e.g. cypermethrin) (Table 1). Given their high solubility in water, their residues also appear in water bodies of agricultural areas and the rivers they drain into [36, 37]. As they are applied consistently as seed dressings, their residues may remain in the soil for years and are taken up by the crop and weeds, ending up in the nectar and pollen of all plants in the treated landscape [16]. This poses a risk to bees, not only because of their high toxicity and availability but also due to their particular mode of action. For example, neonicotinoids show delayed toxicity at low doses, so apart from various sublethal effects they cause [38], they end up killing the bees if they are exposed to the residues for a long period [39]. Both neonicotinoids and fipronil also produce immune suppression on honey bees [40, 41] and, consequently, they predispose bees to *Nosema* infections [42] and out‐ breaks of viral diseases that are commonly transmitted by *Varroa* mites [43, 44]. As a result, colonies feeding on honey and pollen contaminated with these neurotoxic insecticides may succumb to the combined effects of chemicals and diseases [45].

The toxicity of certain insecticides can be enhanced in the presence of ergosterol-inhibiting fungicides (e.g. propiconazole, myclobutanil), which act as synergists. Indeed, this type of compounds inhibits the detoxification system in bees [46, 47], so the insecticide and acari‐ cide residues are not metabolised or eliminated as fast as they should. Furthermore, the tox‐ icity of insecticides and acaricides used for *Varroa* control is often additive or synergistic [9]. Since the food that forager bees collect is usually contaminated with a mixture of both insec‐ ticides and fungicides, and because most managed apiaries are treated with acaricides, the combined toxicity and synergism of all these chemicals pose a real threat to the health and survival of honey bee colonies and all other species of wild bees.

Sublethal exposure to pesticides, including fungicides and some herbicides, often produce stress in animals, because the organisms try to metabolise and get rid of the toxic chemicals quickly using large amounts of energy. Apart from stress, bees experience other negative ef‐ fects when exposed to sublethal doses of pesticides. For example, under conditions of chron‐ ic exposure, honey bee larvae fed on pollen contaminated with chlorpyrifos produced very few queens [48]. Wild bees (*Osmia bicornis*) exposed to sublethal levels of thiamethoxam and clothianidin had their reproductive success reduced by 50% [49], while honey bee queens experienced unusually high rates (60%) of supersedure [50]; bumble bees (*Bombus terrestris*) colonies exposed to sublethal levels of thiamethoxam failed to perform and produced 85% less queens than normal [51]. Sublethal doses of neonicotinoid insecticides also cause disori‐ entation and memory loss in forager bees [38], contributing to less efficiency in the collection of pollen by bumble bees [52]. Sublethal doses of the acaricide coumaphos also produce ab‐ normal mobility in the exposed honey bees [53]. Undoubtedly, all these effects disturb the performance of the individual bees and that of the colony [54].

Finally, the indirect effects caused by herbicides cannot be ignored. Herbicides are not toxic to bees, but they disturb the environment in which bees and other pollinators live. Plant bio‐ diversity, and its associated arthropod communities, have certainly decreased in areas that have been treated with herbicides for many years [55, 56]. The lack of certain plant species, mainly weeds, implies an impoverishment of the natural environment that sustains pollina‐ tors, including honey bees. Consequently, bees find more difficult to collect the variety of pollen that is required for a healthy bee diet [57]. Poor bee nutrition due to scarcity of flow‐ ers is the indirect result of continuous herbicide applications in crops and forestry areas over many decades.

## **4. Risk of pesticides to bees**

Having explained above the routes of exposure to pesticides and their various impacts on bees, an evaluation of the actual risks that current pest control products and acaricides used for treating hives pose to honey bees is needed. The main risk derives from the acute toxicity of the chemicals to the bees, which produce their mortality in the short or middle term. Other risks include sublethal effects that may harm the performance of hives and the long-term viability of the colonies, as mentioned above.

Risks are typically estimated as probabilities of harm and are based on the acute toxicity and the frequency with which a chemical may affect the bees. Three scenarios can be considered: (1) risks from spraying of pesticides over agricultural fields; (2) risks posed by ingestion of agrochemical residues found in pollen, honey and water, which are collected and ingested by the forager bees and transported to the hive, where they are processed into honey and beebread and fed to the other bees, the larvae and the queen; and (3) risks from exposure to combs treated with acaricide products.

#### **4.1. Risk from exposure to sprays**

For the first scenario, the only data required are the concentrations of the active ingredients in the spray solutions applied and their acute toxicity, i.e., LD50 values for each chemical, since the probability of a bee being sprayed on can be considered 100% if the bee flies direct‐ ly through the spray cloud in the field, or if a hive is placed downwind and within the nor‐ mal range of spray drift by aerial or ground-rig applications, i.e., less than 1 km. This kind of risk is estimated using the typical hazard quotient HQ

$$HQ = \frac{\text{Exposure} \left(\mu \text{g}\right)}{LD50 \left(\mu \text{g} \mid be \text{e}\right)}\tag{1}$$

where the exposure term can be determined by the concentration of active ingredient in the spray droplets and the volume received by the bees, according to the following expression

$$HQ = \frac{\text{Concentration} \left(\mu \text{g} \mid ml\right) \times vol.droplets \left(ml\right)}{LD50 \left(\mu \text{g} \mid bee\right)}\tag{2}$$

In this case, the HQ can be indicative of high risk when its value is 1 or more, since 50% or more bees exposed would die; moderate risk is when HQ values are between 0.1 and 1 and low risk when it is less than 0.1, as fewer than 10% (similar to a natural mortality rate) of bees would be threaten.

Estimates of risks are typically done by considering the spray drift [58, 59] and the exposure to the flying bees [60]. For example, to compare the risk posed by different products under the same conditions, the spray drift volume can be fixed, e.g. 500 droplets for a bee crossing the spray cloud in a few minutes, at 5×10−4 μl for a standard droplet would result in 0.25 μl/


bee. **Table 2** shows a comparison of the risk that commonly applied pesticides would have in such situations.

**Table 2.** Risk of common agricultural pesticides to honey bees that fly across a spray cloud (ppm) and receive a total dose of 0.25 μl/bee

The example in Table 2 reveals that the insecticides fipronil, thiamethoxam, bifenthrin, lambda-cyhalothrin and chlorpyrifos are the most dangerous to bees when sprayed to agricultural crops. The microencapsulated formulation of lambda-cyhalothrin is particularly hazardous because bees can carry the microcapsules containing the concentrated chemical to the hive. In general, dust particles of neonicotinoid-treated seeds and spray droplets of pyrethroids, organophosphorus and carbamate insecticides pose moderate or high risks, whereas other insecticides and acaricides present low risks in comparison. The fungicides shown here, and possibly most others applied as foliar sprays, pose low or negligible risks to bees by direct contact with spray droplets. This evaluation is in agreement with the reported incidents of pesticides on bees in the United Kingdom [63] and Canada [64]. Obviously, the most toxic insecticides are the most dangerous to bees.

#### **4.2. Risks by oral exposure**

For the second scenario, ingestion of contaminated food, data on the concentration and frequency of residues in each media are essential, as well as information on the dietary intake of pollen, honey and water by each caste of bees, that is, foragers, nurses, larvae and queen. Oral exposure to contaminated food is considered the typical exposure of bees in the hive. The risk expression in this case would take the form

$$Risk = \frac{Frequency\left(\%\right) \times residue\,dose\left(\mu\text{g}\right)}{LD50\left(\mu\text{g} \mid base\right)}\tag{3}$$

where the residue dose of a given pesticide can be estimated for different bees as the product of the concentration of residues in pollen, honey or water by the total intake of a particular caste of bee [21]. In turn, total intakes are estimated from daily intakes and the life span of bees, which vary from 5–6.5 days for larvae, 8–10 days for brood attendants and nurses, to 30 or more days for foragers [65]. The food intake by queens is hard to estimate, as they are fed royal jelly (a particular combination of pollen and honey), can live several years and vary their intake —which is unknown—throughout the reproductive and winter seasons. For the toxicity, oral LD50s are used in this case. The risk estimated by expression (3) can be interpreted as the probability that a given pesticide residue has of causing 50% mortality among the bees that ingest the contaminated pollen or nectar.

In recent years, a number of studies have reported the residue levels of agricultural pesticides found in pollen [66, 67] and nectar of flowers [68, 69], in water bodies of agricultural areas [28], as well as in beehive matrices, such as beebread, honey and wax [70, 71]. Based on these reports, we estimated the average and maximum residues for each pesticide as well as their frequency of appearance in those matrices. This information allowed us to calculate the risks that bees encounter when feeding on such contaminated food or drinking sources. A summary of results for the compounds that pose the highest risks by oral exposure of combined food and drink is shown in **Table 3**.


**Table 3.** Average pesticide residue levels in food and water (ppb) and their risk by oral exposure to worker honey bees and larvae. The time to reach the oral lethal dose (T50, days) is also shown for a comparison

Despite the high risk of some chemicals, namely neonicotinoids, most insecticide residues in pollen and honey present a moderate risk to bees (1 to 5%), especially those of pyrethroid and organophosphorus compounds. Overall, 21 of the 113 pesticide residues in food for which toxicity data are available pose some kind of risk to honey bees, but only 8% of the chemicals are of concern. Residues in water are more variable from place to place: the data shown in Table 3 are from one survey in Canada where only neonicotinoids, fungicides and herbicides were found—the risks posed by the latter two groups were negligible nonetheless, so they are not shown in the table.

#### **4.3. Risks by contact exposure**

Apart from oral exposure, bee larvae may also be in contact with residues deposited on the walls of the comb cells, in particular, the acaricides used for controlling *Varroa*. Although the highest loads of pesticide residues in a hive are found in the wax [23], the availability of such chemicals is thought to be minimal except for the fumigated acaricides. The risk of the latter products to bee larvae should be estimated not as oral intake, as some authors do [30], but rather as contact exposure. The expression (3) can be used, with the maximum residue dose in this case estimated as 5 mg of active compound per cell for a single larva and the contact LD50 instead of the oral one. The results of the risk analysis for a number of acari‐ cides to honey bee larvae are shown in **Table 4**.


**Table 4.** Average acaricide residue levels in comb wax (ppb) and their risk to larvae of honey bees. The time to reach the lethal dose (T50, days) is also shown for a comparison

As it can be seen, the risks of acaricides to bee larvae are below 1% for all individual chemi‐ cals, but increases dramatically for synergistic mixtures, such as tau-fluvalinate with amitraz or coumaphos. Except for the latter mixtures, the overall risk to bee larvae of the individual products is very low or negligible compared to that of the same compounds by oral inges‐ tion of contaminated food and water (Table 3).

#### **4.4. Novel approaches to risk assessment**

Another way of estimating risks, particularly for oral exposures, is by calculating the time that would take for a bee to reach the LD50 of a given pesticide, based on the daily intake of contaminated food and water. This estimate is made using the expression

$$T50 \left(days \right) = \frac{LD50 \left(\mu \text{g} / base\right)}{Daily\text{ intake} \left(\mu \text{g}\right)} \tag{4}$$

where T50 is the time to reach the median lethal dose (LD50), also termed median time to death. As it can be expected, there is a good correlation between the T50 values estimated using equation (4) and the risk values calculated using equation (3)—see Tables 3 and 4.

Neonicotinoid insecticides, however, can cause delayed mortality due to their agonistic mode of action [39]. This particularity means that their acute oral LD50s, which are usually estimated for exposures of 48 hours, are insufficient to estimate accurate risks of these insecticides, because the actual dose that causes the death of the bees decreases as the time of exposure increases [72]. Consequently, the mathematical function that relates the median time to death (T50) with the median lethal dose (LD50) is used to estimate the risk, as follows

$$
\ln T\ $0 \left( days \right) = a + b \ln L \text{n} \, L\$ 0 \left( \mu\text{g} \,/\, bce \right) \tag{5}
$$

where *a* (intercept) and *b* (slope) are empirical parameters specific to each chemical and species tested (in this case honey bees). The approach estimates the cumulative mortality with exposure time with more precision than the standard equation (4), as explained in a previous study [21].

## **4.5. Risk from synergistic mixtures of pesticides**

The above tables help determine the pesticides that pose the greatest danger to bees, whether by exposure to spray droplets or dust, oral ingestion of contaminated food and water or contact with chemicals used for mite control in the hives. It is clear that the majority (92%) of pesticides registered for agricultural production do not pose significant or measurable risks to honey bees, but this is only when considering the exposure to individual compounds.

Recent developments, however, indicate that combination of certain chemicals, in particular insecticides and acaricides with fungicides or mixtures of acaricides, is more toxic to bees than the individual compounds on their own. The additive and synergistic effects of those mixtures have already been mentioned above, and estimation of the risks they pose needs to be calculated using the same approaches but modifying the toxicity of the insecticide or acaricide by a synergistic factor [21]. These factors are calculated experimentally for several combina‐ tions of fungicides with insecticides and/or acaricides [73], and some examples are shown in **Table 5**.


**Table 5.** The synergistic effect of some fungicides with insecticides or acaricides and their risks to honey bees

Although the increases in risk are obvious, only the interaction of the pyrethroid insecticide cyhalothrin with propiconazole points to a moderate concern for bee larvae; even the risk of thiacloprid appears to be low under these circumstances. However, the risk of certain acaricide mixtures, such as tau-fluvalinate with amitraz or coumaphos, used in *Varroa* treatments, can be very high for the larvae (see Table 4).

## **5. Management in order to avoid pesticide impacts**

The various risks estimated above give us some clues about the type of exposure most dangerous to the different castes of bees in the hives. Spray drift is the main cause of incidents involving mortality of forager worker bees [63, 74], whereas ingestion of contaminated pollen, nectar and water is at the root of the CCD malady that affects many apiaries of the world [45], affecting mainly the nurse workers and the queen in particular [49, 51]. In addition, the acaricides used in *Varroa* treatment pose a significant risk mainly to the bee larvae, and consequently to the long-term sustainability of the colonies. Awareness of these threats can help beekeepers and farmers draw specific management plans to avoid pesticide impacts.

Beekeepers should be aware of the landscape environment on which their managed bees forage, bearing in mind that a large proportion of the land in developed and developing countries is used for agricultural production where pesticides of all kinds are used on a regular basis. Since usage of these plant protection products cannot be stopped, as they are necessary for agricultural production, a rational approach must look at minimising the risks of such agrochemicals to bees.

Chemical companies are obliged by law to state on the labels whether their products are dangerous to bees or not. If so, they must specify the risks they pose and the specific actions to take, such as "DO NOT spray any plants in flower while they [the bees] are foraging." However, label warnings are ineffective unless there is proper communication among the applicators, farmers and beekeepers. It is the responsibility of the former to ensure that beekeepers are informed of any spraying operations, so that hives are moved to a safe location during the spraying season. Moving hives usually takes more than 24 hours, so farmers must notify their neighbouring beekeepers with sufficient time in advance. Only thus damage by drift to the hives can be avoided.

Bees are generally active between sunrise and an hour or two before sunset, and most honey bees forage within a 2–4 km radius of their hive, although may travel as far as 7 km or more in search of pollen and nectar when their local sources are scarce [75]. Therefore, pesticide risk to bees can be reduced by spraying the crops in the evening, when bees are not foraging.

Despite all precautions, if an area in which the crop or weeds were in flower has been sprayed inadvertently, the farmer should notify the affected beekeepers in order for them to take appropriate action. This should ensure the managed bees are kept out of that sprayed area for a while. As well as the cropping areas, damage may occur when pesticides drift over the neighbouring vegetation that is foraged by bees, including hedges, road-side weeds and trees, such as fruit trees, eucalypts, etc. For example, coolibah trees (*Eucalyptus microtheca*) grow on plains along many river courses in the cotton growing areas of Australia and are a primary source of nectar and pollen for wild and honey bees. The Australian cotton industry has produced a best management practices manual in which, among other recommendations, indicates to the cotton growers how to deal with this issue and be aware of the possible damage to beekeepers. "With good communication and good will," says the manual, "it is possible for apiarists and cotton growers to work together to minimize risks to bees, as both the honey industry and cotton industry are important to regional development." [33].

In summary, awareness of the problems that pesticides have for bees should prompt appro‐ priate actions by all parties involved in order to minimise the chemical impacts on bees and the productivity of the apiarist industry. Such actions must aim, first of all at managing the use of agrochemicals in ways that do not harm other producers of the land. In addition, farmers should minimize the contamination of the surrounding landscapes, including water bodies, with pesticides, because not only honey bees but a large array of pollinator species (e.g. butterflies, bumblebees, hoverflies, etc.) may also be affected.

## **Author details**

Francisco Sanchez-Bayo1\* and Koichi Goka2


## **References**

