3. History of rearing honey bee larvae in the laboratory

Isolating worker larvae from adult bees in a hive and rearing them to the adult stage has challenged many early researchers who were mainly interested in honey bee physiology and caste determination. Bertholf (1927) kept three-day-old larvae in the laboratory alive for more than two days by feeding them different concentrations of sucrose solution, and Velich (1930) reared them to adults. The first important report of hand-feeding honey bee larvae in the laboratory was published by Rhein (1933). He found differences in the jelly fed to workers and queens (royal jelly) and successfully used royal jelly to rear two- to three-day-old worker larvae to adults. He also recognized the problem of worker-queen intermediates. Due to the focus of research, the consistent production of females belonging to the worker caste was one of the main challenges during these early days of in vitro rearing of larvae. Presently, the in vitro rearing of larvae is being used as a routine method in pesticide testing and for many applications in honey bee physiology and pathology (Table 1). Hence, the reduction of individual feedings and, thus the work load for the experimenter, is appreciated and presents a challenge to those further developing these methods. In contrast to the numerous progressive feedings of larvae in a colony, larvae in the laboratory are only fed daily or even less frequently, preferably without influencing survival rate (Aupinel et al., 2005; Kaftanoglu et al., 2010).

After initial attempts, many researchers developed methods to rear honey bee larvae in the laboratory by modifying or optimizing the basic diet, which consisted of royal jelly diluted with an aqueous solution of glucose and fructose. Fraenkel and Blewett (1943) discovered the importance of yeast in artificial insect diets for Drosophila larvae, but this was later replaced with B vitamins and polyunsaturated fatty acids (Vanderzant, 1974). The use of yeast was also applied to honey bee larval diets, and yeast extract is still in use (Michael and Abramovitz, 1955; Peng et al., 1992; Aupinel et al., 2005) but has been omitted by others without a reduction in individual survival rates (Genersch et al., 2005; Genersch et al., 2006). Yeast has been demonstrated not to be a differentiating agent for female castes, but it may be beneficial because of phagostimulatory or nourishing effects (Rembold and Lackner, 1981; Vandenberg and Shimanuki, 1987). Since in vitro rearing can also be completed successfully without the addition of yeast extract, the role of yeast needs to be questioned critically. In addition to using yeast, other attempts have been made to enrich the larval diet, for example with a readily available vitamin formulation (Multibionta; Merck; Herrmann et al., 2008) which did not affect larval weight or mortality significantly.

Michael and Abramovitz (1955), amongst others, developed self-feeding dishes and fed an aqueous honey solution containing 10 % dehydrated yeast extract. They were the first to inoculate larvae in the laboratory with European foulbrood. Rearing larvae on diluted royal jelly is “an art”, as stated by Weaver (1974), and yielded only small numbers of individuals and also queen-worker intermediates. More information on early in vitro rearing of honey bee larvae and the success of different researchers can be found in Jay (1964). These early experiments formed the basis for later protocols, produced valuable scientific information and permitted biological testing of different qualities of royal jelly (Weaver, 1955; Smith, 1959; Mitsui et al., 1964; Asencot and Lensky, 1984). The use of worker jelly (which is also disproportionately labour intensive to harvest compared to royal jelly) was doomed, because survival is low and larvae do not pupate (Rhein, 1933; Herrmann et al., 2008). Survival on worker jelly can be increased when sugars are added, and Asencot and Lensky (1976; 1984; 1988) believed that sugars together with juvenile hormone could be the queen determinant. More recent results regarding the importance of sugars for pupation are available in Kaftanoglu et al. (2011). In turn, to find the caste-determining factor, Rembold et al. (1974) reared young worker larvae in thimbles fed a basic food derived from royal jelly via a very extensive alcohol extraction. This diet produced more adult workers, but also more queens and intermediate forms, than did diluted royal jelly. Shuel and Dixon (1986) prepared a complex diet composed of soluble and insoluble protein extracts from royal jelly enriched with solutions containing vitamins, minerals and other compounds.

Wittmann and Engels (1981) suggested investigating the effects of plant protection products on in vitro-reared honey bee larvae, and Davis et al. (1988) were among the first to administer carbofuran and dimethoate dissolved in royal jelly to honey bee larvae. At the same time, Czoppelt and Rembold (1988) assessed the toxicity of parathion to larvae. Aupinel et al. (2005) imitated the age-dependent increase of sugars and dry matter in the larval food in the colony (Brouwers et al., 1987) by gradually increasing glucose, fructose and yeast extract in the diet of larvae (Fig. 1b). They also replaced beeswax cups or thimbles with readily available plastic cups (cupula - used mainly by beekeepers for queen rearing purposes) placed in 48-well microtiter plates (Fig. 2a). This method allows for adoption in testing insecticides in different laboratories (Aupinel et al., 2009).

Two principal in vitro rearing methods are employed which we will discuss in this paper. In the first method, one larva is reared per cup and the exact amount of diet a larva consumes daily is administered, following the protocol of Rembold and Lackner (1981) with modifications by Vandenberg and Shimanuki (1987). This amount of diet is estimated to be 160µl in total during larval development (Aupinel et al., 2005) or 164µl for Africanized honey bees (Silva et al., 2009). The second approach administers excess diet, and consequently the larvae have to be transferred to new dishes regularly. At the start of this process several worker (Peng et al., 1992; Genersch et al., 2005; 2006) or drone (Behrens et al., 2007; 2010) larvae are reared in one culture plate well (Fig. 2b). With increasing age, the larvae are either progressively isolated by daily grafting on fresh food (Genersch et al., 2005; 2006) or still grafted in groups to new petri dishes (Kaftanoglu et al., 2010). During ad libitum feeding, the individual uptake of diet and compounds / pathogens is not restricted to a certain volume (and hence, dosage) as in single-cup protocols. Therefore, it is only possible to determine the LC50 (median lethal concentration) for a given substance or pathogen; however, the LD50 (median lethal dose) can be roughly estimated (when necessary) by assuming the average consumption to be the abovementioned 160µl per larvae. In many cases it will be sufficient to determine the lethal concentration of a substance or pathogen, because under natural conditions bees and larvae will be exposed to matrices containing a certain concentration of a substance or pathogen. Administering a certain dosage, as is done in medical treatments of humans and other vertebrates, will not be the issue with honey bees, because drugs are administered to bees not in a given dosage per bee but rather dissolved in sugar solution fed to the bee colony. However, it has to be taken into account that all calculated dosage estimates are derived from the assumed consumption, and larvae will consume more if more food is provided. Therefore only individual larval rearing with limited feedings and control of complete consumption allows accurate dosage calculations (Aupinel et al., 2005).

Many of the methods previously published resulted in high mortalities and were labour intensive. However, progress has been made and some of the protocols are sufficient to consistently rear worker honey bee larvae with little or no mortality in the laboratory and can therefore be used for mass rearing and the application of routine testing of compounds (Aupinel et al., 2005; Hendriksma et al., 2011a). This might suggest that no more research efforts to develop new diets would be necessary. However, researchers need to adapt existing diets adequately for their studies or develop new diets to understand the nutritional requirements of honey bee larvae in detail. There is still no chemically defined diet available, as there are for other insects, and it is unlikely that such a diet will be produced for honey bee larvae (Vanderzant, 1974; Shuel and Dixon, 1986). So far, all published diets have been composed of crude materials (royal jelly), imitating the natural food to a great degree and thus are categorized as a third type of artificial diets as described by Vanderzant (1974).

Table 1. Research topics using laboratory-reared honey bee larvae and selected references.

Research topic:


Caste differentiation

Rhein, 1933; Weaver, 1955; Smith, 1959; Shuel and Dixon, 1960; 1986; Mitsui et al., 1964; Rembold et al., 1974; Weaver, 1974; Asencot and Lensky, 1976, 1984, 1988; Shuel et al., 1978; Rembold and Lackner, 1981; Vandenberg and Shimanuki, 1987; Wittmann and Engels, 1987; Patel et al., 2007; Kucharski et al., 2008; Kamakura, 2011; Shi et al., 2011; A. De Souza D, Wang Y, Kaftanoglu O, De Jong D, V. Amdam G, S. Gonçalves L, et al. (2015) Morphometric Identification of Queens, Workers and Intermediates in In Vitro Reared Honey Bees (Apis mellifera). PLoS ONE 10(4): e0123663. doi:10.1371/journal.pone.0123663

diploid drones

Woyke, 1963; Herrmann et al., 2005

Larval pathogens

Michael and Abramovitz, 1955; Peng et al., 1992; 1996 Brødsgaard et al., 1998, 2000; Genersch et al., 2005, 2006; Behrens et al., 2007, 2010; Jensen et al., 2009; Forsgren et al., 2010; Vojvodic et al., 2011a, 2011b, 2012; Vasquez et al., 2012; Foley et al., 2012; Eiri DM, Suwannapong G, Endler M, Nieh JC (2015) Nosema ceranae Can Infect Honey Bee Larvae and Reduces Subsequent Adult Longevity. PLoS ONE 10(5): e0126330. doi:10.1371/journal.pone.0126330


Wittmann and Engels, 1981; Davis et al., 1988; Czoppelt and Remboldt, 1988; Aupinel et al., 2005; 2007a; 2007b; 2009; Medrzycki et al., 2010; Da Silva Cruz et al., 2010; Gregorc and Ellis, 2011; Hendriksma et al., 2011a; Gregorc et al., 2012; Tavares et al., 2015

Transgenic plants

Malone et al., 2002; Brødsgaard et al., 2003; Hendriksma et al., 2011b, 2012; Steijven et al., 2016

Fig. 1.
Comparison of: a. protein and; b. sugar content of brood food of young (1-3 d) and old (4-6 d) worker larvae (Data summarized from Kunert and Crailsheim, 1987) and laboratory diet, both percentages of dry weight. Protein decrease and sugar increase are due to the increase in sugar added to diets (Aupinel et al., 2005). Range of content is determined by variation in royal jelly protein (Sabatini et al., 2009) and sugar (Brouwers, 1984) content. Protein and sugar in yeast were excluded from calculations.

Figure 1


Fig. 2.: a. 48-well microtiter plate containing 47 plastic queen cups on dental rolls. Larvae are about 6-7 days old and reared in the same cup from first instar on (Photo: Pierrick Aupinel). b. 3 groups of 10 each first instars (upper panel) which will be progressively isolated in new wells lined with filter paper, in the end resembling the 3 engorged larvae in one cup each (lower panel, Photo: Elke Genersch).

Figure 2


Robert Brodschneider
Robert Brodschneider says:
Nov 17, 2015 02:13 PM

New publications for Table 1:
Line "Caste differentiation": A. De Souza D, Wang Y, Kaftanoglu O, De Jong D, V. Amdam G, S. Gonçalves L, et al. (2015) Morphometric Identification of Queens, Workers and Intermediates in In Vitro Reared Honey Bees (Apis mellifera). PLoS ONE 10(4): e0123663. doi:10.1371/journal.pone.0123663

Line "Larval pathogens": Eiri DM, Suwannapong G, Endler M, Nieh JC (2015) Nosema ceranae Can Infect Honey Bee Larvae and Reduces Subsequent Adult Longevity. PLoS ONE 10(5): e0126330. doi:10.1371/journal.pone.0126330

Ralf Bünemann
Ralf Bünemann says:
Nov 23, 2015 03:50 PM

Thanks Robert, the new lines have been added.

Robert Brodschneider
Robert Brodschneider says:
Feb 08, 2016 04:50 PM

Please add "Tavares et al., 2015" to Line "Toxicity" and Steijven et al., 2016 to line "Transgenic plants"

Ralf Bünemann
Ralf Bünemann says:
Feb 10, 2016 09:56 AM

OK, added!

Robert Brodschneider
Robert Brodschneider says:
Mar 31, 2016 02:45 PM

For Caste differentiation please add "Mao et al., 2015"

Robert Brodschneider
Robert Brodschneider says:
Oct 04, 2016 09:09 AM

add "Fourrier et al., 2015" for table in line "Toxicity"

Fourrier J, Deschamps M, Droin L, Alaux C,
Fortini D, Beslay D, et al. (2015) Larval Exposure to
the Juvenile Hormone Analog Pyriproxyfen Disrupts
Acceptance of and Social Behavior Performance in
Adult Honeybees. PLoS ONE 10(7): e0132985.

Robert Brodschneider
Robert Brodschneider says:
Oct 04, 2016 09:14 AM

add "Buttstedt et al., 2016" to line caste differentiation!

A. Buttstedt, C. H. Ihling, M. Pietzsch & R. F. A. Moritz Nature 537, http://dx.doi.org/10.1038/nature19349 (2016).

Robert Brodschneider
Robert Brodschneider says:
Jan 18, 2017 01:24 PM

For further editions of this chapter the papers of Schmehl et al., JAR (2016) and Fine et al., SciRep (2017) should be included.