University of Florida Thesis Or Dissertation Formatting

BEHAVIORAL AND MORPHOLOGICAL EFFECTS OF IN VITRO LARVAL REARING ON ADULT HONEY BEES, APIS MELLIFERA L.

ASHLEY NICOLE MORTENSEN

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2017

To my loving family that has supported me through all of life’s inspirations

ACKNOWLEDGMENTS

I am grateful to Daniel Schmehl and Hudson Tomé for their guidance troubleshooting the in vitro rearing protocol; Cameron Jack and Ping Li Dai for assistance in rearing many honey bee larvae; Brandi Simmons, Emily Helton, and

Branden Stanford for ordering and preparing project supplies; Logan Cutts and Liana

Teigan for managing healthy honey bee colonies; Vince Alderman for his diligent construction of project supplies; Tomas Bustamonte, Branden Stanford, and Jon Novak for assistance measuring and entering data; Edzan Van Santan for statistical consulting, and my advisory committee, Jamie Ellis, Christine Miller, Andrea Lucky, and Emily

Miller-Cushon, for their support and guidance.

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ...... 4

LIST OF TABLES ...... 7

LIST OF FIGURES ...... 8

ABSTRACT ...... 9

CHAPTER

1 RATIONALE AND SIGNIFICANCE ...... 11

2 A HONEY BEE COLONY’S NATURAL BROOD SURVIVAL RATE PREDICTS ITS IN VITRO LARVAL SURVIVAL RATE ...... 15

Introduction ...... 15 Materials and Methods...... 16 Brood Isolation ...... 17 Colony Rearing Assay ...... 17 In vitro Rearing Assay ...... 18 Statistical Analysis ...... 19 Results ...... 19 Discussion ...... 20

3 COMPARATIVE MORPHOLOGY OF ADULT HONEY BEES REARED IN VITRO OR IN THEIR PARENT HIVE ...... 28

Introduction ...... 28 Materials and Methods...... 30 Dry Bee Weight ...... 31 External Morphology ...... 32 Statistical Analysis ...... 32 Results ...... 33 Discussion ...... 33

4 EFFECTS OF ARTIFICIAL REARING ENVIRONMENT ON THE BEHAVIOR OF ADULT HONEY BEES ...... 39

Introduction ...... 39 Materials and Methods...... 43 Queen Recognition ...... 45 Brood Rearing ...... 47 Trophallaxis ...... 48 Sucrose Responsiveness ...... 49

Statistical Analysis ...... 50 Results ...... 51 Queen Recognition ...... 51 Brood Rearing ...... 51 Trophallaxis ...... 51 Sucrose Responsiveness ...... 52 Discussion ...... 52

5 FIELD-BASED BEHAVIORAL OBSERVATIONS OF ADULT HONEY BEES REARED IN VITRO ...... 67

Introduction ...... 67 Materials and Methods...... 70 Results ...... 73 Discussion ...... 74

6 CONCLUSION ...... 84

LIST OF REFERENCES ...... 90

BIOGRAPHICAL SKETCH ...... 103

LIST OF TABLES

Table page

1-1 Comparison of the environmental conditions experienced by in vitro- and colony-reared honey bees...... 14

2-1 Means comparisons of the rearing environment by time point interaction effect observed for survival percentage...... 25

3-1 Summary of the results of each quantitative morphometric trait measured...... 38

4-1 Summary of the distribution function, link function, and random effects for each response group/variable by social behavior...... 62

4-2 Summary of parameters tested to assess the effect of developmental environment (parental colony or in vitro) on adult worker bee behavior...... 66

5-1 Categories used when recording adult honey bee behavior during observational periods...... 79

5-2 Age at which honey bees were participating in each behavior category, shown by rearing environment...... 82

5-3 Comparisons of the mean age at which each behavior category was performed by honey bees that were reared by their parental colony or in vitro. . 83

6-1 Summary of parameters tested to assess the effects of developmental environment (parental colony or in vitro) on adult worker honey bees...... 88

LIST OF FIGURES

Figure page

2-1 The average percent survival by rearing environment (parental colony or in vitro) and time point (day 11 or at adult emergence) for developing honey bees...... 24

2-2 Predictive relationship between day 11 survival percentages and adult emergence survival percentages by rearing environment...... 26

2-3 Predictive relationship between the day 11 survival percentages in the parental hive and the adult emergence survival percentages of larvae grafted from those parental hives and reared in vitro...... 27

3-1 External features measured for morphological comparisons between adult honey bees that were reared in vitro or in their parental colony...... 37

3-2 External morphological characteristics used to categorize individuals as worker honey bees, queens, or queen/worker intercastes...... 37

4-1 Images of the queen recognition assay...... 57

4-2 Composite image of the petri dish cages in round two of the brood rearing assay...... 58

4-3 Images of the trophallaxis assay...... 59

4-4 Images of the dissected stomach contents of the donor bee and ten recipient bees from one perti dish...... 60

4-5 Images of the sucrose sensitivity and responsiveness assay...... 61

4-6 The average total counts of queen cell visits...... 63

4-7 The average seconds of contact/bee for honey bees reared in their parental colony (C; 146.8 ± 30.2 sec, 18) or in vitro (IV; 40.7 ± 30.1)...... 64

4-8 The average sucrose responsiveness scores for bees reared in vitro or by their parental colonies...... 65

5-1 Histograms of the age of each honey bee at its last observation, shown by rearing environment (parental colony, or in vitro)...... 80

5-2 Age distribution of honey bees performing each task category by rearing environment (colony-reared, or in vitro-reared)...... 81

6-1 Adult worker honey bee that had been reared in vitro crawling out of her well at a daily inspection...... 89

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

BEHAVIORAL AND MORPHOLOGICAL EFFECTS OF IN VITRO LARVAL REARING ON ADULT HONEY BEES, APIS MELLIFERA L.

Ashley Nicole Mortensen

December 2017

Chair: James D. Ellis Major: Entomology and Nematology

Rearing honey bee, Apis mellifera L., larvae in vitro is a popular risk assessment tool because many uncontrollable factors (e.g. weather conditions, food availability, etc.) that bias field studies can be eliminated in the laboratory. However, few investigators have explored how in vitro rearing may affect the resulting honey bees’ morphology, behavior, and/or physiology. Furthermore, modern in vitro rearing techniques still suffer variable mortality rates. To address these gaps in knowledge, I analyzed the developmental survival rates of bees in colonies and in vitro, and assessed the morphology and behavior of adult workers that had been reared in vitro compared to bees that had been reared by their parental colony. I found that brood survival in a colony is predictive of the survival rate of larvae from that colony in vitro. Furthermore, bees reared in vitro have decreased dry body weight and forewing size, however hind wing length, head width, and basitarsus length are unchanged. Some behaviors that were assessed in the laboratory (trophallaxis, queen recognition, and cell capping behavior) were unchanged in in vitro-reared bees, but sucrose responsiveness, and brood tending behaviors are decreased in bees that were reared in vitro. Finally, in vitro-

reared bees have a lower mean age at which they perform some behavior categories

(grooming, head in a cell, and non-productive behaviors such a standing or walking), but the mean age at which they perform other behaviors (attending the queen, washboarding, wax manipulation, ventilation, guarding, and foraging) is unchanged.

Bees reared in vitro suffered reduced longevity in the hive. Therefore, significant changes detected in behavior categories that span the entire lifetime (grooming, head in a cell, and non-productive behaviors) may be an artifact of a shorter lifespan rather than a change in the expression of those behaviors. Further examination of the physiology and behavior of bees that have been reared in vitro, in a broader array of assays in the laboratory and in the field, promise to offer more insight into the extent to which rearing environment affects adult honey bees.

CHAPTER 1 RATIONALE AND SIGNIFICANCE

In vitro larval rearing has the potential to be a valuable risk assessment tool for honey bee, Apis mellifera L., research. Modern management systems and human- mediated transport of honey bees worldwide has resulted in colonies being exposed to a myriad of stressors such as parasites, pathogens, and pesticides (Ruttner 1988; Ellis and Munn 2005; Dietemann et al 2006; Neumann and Carreck 2010; Mullin et al 2010;

Medrzycki et al 2013). Standard methods for assessing the risk of exposure to these stressors to adult honey bees in the field and in the laboratory have been well vetted

(OEPP/EPPO 2010; Medrzycki et al 2013). However, there is a critical need for a risk assessment protocol for determining the effects of stressors on immature honey bee stages (Hendriksma et al 2011; Crailsheim et al 2013).

Evaluation of brood development in response to stressors is difficult within a hive because nurse bees readily abort larvae that have been experimentally manipulated

(Spivak and Gilliam 1998; Ibrahim and Spivak 2006). Furthermore, experiments conducted within a honey bee hive are biased by many uncontrolled factors such as resource availability, season, climate, and colony genetics (Hendriksma et al 2011).

Therefore, in vitro rearing promises to be the ideal method by which to assess the risk associated with larval exposure to potential stressors.

Rudimentary larval rearing techniques were first used to study caste differentiation in honey bees in 1933 (Crailsheim et al 2013). Wittmann and Engels

(1981) were first to suggest the use of in vitro rearing protocols to determine the toxicity of pesticides to worker bee larvae. Since 1981, several rearing protocols have been developed that offer moderate survival success for in vitro-reared individuals

(Vandenberg and Shimanuki 1987; Peng et al 1992; Aupinel et al 2005; Silva et al 2009;

Crailsheim et al 2013). However, inconsistent survival rates among untreated individuals have continued to be problematic when interpreting results of in vitro larval risk assessments (Aupinel et al 2010). Recently, Schmehl et al. (2016) improved upon these techniques so that adult emergence rates in untreated larvae are consistently well above the OECD minimum survival requirement of >70% adult emergence (OECD

2015).

Fundamental understanding of how in vitro rearing may alter the resulting bees is critical to our interpretation of risk assessments utilizing the method (Huettel 1976;

Hendriksma et al 2011; Crailsheim et al 2013). There are substantial differences in the environment in which in vitro- and colony-reared larvae develop (Table 1-1; Seeley

1985; Schmickl and Crailsheim 2002; Tautz et al 2003; Crailsheim et al 2013; Schmehl et al 2016). Notable social development occurs during the immature stages of other arthropods and this can affect their behavior as adults (Hebets 2003; Strodl and

Schausberger 2012). It is plausible that honey bees need social interactions/stimuli as immatures in order to function normally as adults.

There has been little investigation into the effects of in vitro rearing on the physiology and behavior of the resulting adult honey bees (Brodschneider et al 2009;

Kaftanoglu et al 2010). Brodschneider et al. (2009) noted that in vitro-reared workers had similar flight performances to those of naturally reared workers. However, naturally reared workers obtained height maximum flight speeds, had higher dry body weights, and increased forewing surface area than did in vitro-reared workers. These findings highlight that rearing environment does affect adult honey bee morphology and

physiology. Furthermore Tautz et al. (2003) demonstrated that the temperature at which pupal bees are incubated affects the learning and memory of the resulting adult bees. It is imperative that we determine to what extent the in vitro rearing environment may alter the physiology, morphology, and behavior of adults honey bees.

The goal of the research detailed within is to determine if in vitro larval rearing alters the physiology and behavior of the resulting adult honey bees. I will address this research goal with the following comparative studies:

1. Survival of workers reared in vitro to that of colony-reared workers,

2. Morphology of workers reared in vitro to that of colony-reared workers,

3. Behavior in laboratory assays of workers reared in vitro that of colony-reared workers, and

4. Behavior in colonies of workers reared in vitro to that of colony-reared workers.

These studies provide insight into how the developmental environment of honey bee larvae may impact the morphology, physiology, and behavior of the resulting adult workers. I hypothesize that adult workers that are reared in vitro will develop into seemingly normal adults that differ slightly from colony-reared adults in physiology, behavior, and key morphometric parameters. If larvae reared in vitro develop into normal, functional adults, the in vitro protocol will be confirmed as a powerful risk assessment tool for determining the impacts of stressors on immature honey bees

(Hendriksma et al 2011; Crailsheim et al 2013). Any potential abnormalities in in vitro- reared workers may support the need to view data collected using this assay cautiously.

Table 1-1. Comparison of the environmental conditions experienced by in vitro- and colony-reared honey bees. Environmental Colony-Reared In Vitro-Reared Condition only exposed to nurse bees for up up to 2,785 interactions social interactions to 24 hours after hatching from the with nurse bees1 egg2 royal jelly for three days artificial, made of a variable diet then bee bread and composition of royal jelly, glucose, worker jelly3 fructose, yeast, and water2

ad libitum from nurse five bolus feedings from a pipette at feeding bees3 fixed times2

maintained in a dark hive complete darkness within the light w/ diffuse sunlight from incubator accented w/ daily the hive entrance3 inspections in artificial light2 physical environment horizontal wax comb3 vertical 48-well tissue culture plate2

the individual pupates in a larvae are transferred to a new well pupation wax cell that is covered on a 48-well plate and incubated at with a wax capping3 a reduced humidity2

1 (Schmickl and Crailsheim 2002) 2 (Schmehl et al 2016) 3 (Seeley 1985)

CHAPTER 2 A HONEY BEE COLONY’S NATURAL BROOD SURVIVAL RATE PREDICTS ITS IN VITRO LARVAL SURVIVAL RATE

Introduction

Increased honey bee colony losses (Neumann and Carreck 2010) and dramatic pollinator decline (Lebuhn et al 2013) are occurring worldwide. Parasites, pathogens, poor nutrition, queen quality, and pesticides are considered significant colony stressors and likely contributing causes to the global losses. A substantial amount of research has focused on determining the extent to which these stressors affect honey bee health

(Chauzat et al 2006; Higes et al 2008; Le Conte et al 2010; Mullin et al 2010; vanEngelsdorp et al 2010; Martin et al 2012; Steinhauer et al 2014).

Pesticides and pathogens generally are regarded as primary factors affecting honey bee colony health (Neumann and Carreck 2010; Mullin et al 2010; Medrzycki et al 2013). Standard methods for investigating pesticide and pathogen effects on adult honey bees at the field and laboratory level have been well-vetted (OEPP/EPPO 2010;

Medrzycki et al 2013), though potential effects of pesticides on developmental stages

(larvae and pupae) often are overlooked. Some field evaluations do include parameters such as total brood area (Delaplane et al 2013). However, experiments conducted within a honey bee colony are biased by many uncontrolled factors such as, resource availability, season, climate, and colony genetics (Hendriksma et al 2011). Much of this bias can be overcome using in vitro rearing techniques for honey bees (Hendriksma et al 2011; Crailsheim et al 2013). Furthermore, in vitro rearing is useful in studying honey bee development and caste differentiation (Woyke 1963; Rembold and Lackner 1981;

Asencot and Lensky 1984; Brouwers 1984; Crailsheim et al 2013; Buttstedt et al 2016).

Rudimentary in vitro rearing techniques were first used to study caste differentiation in honey bees in 1933 (Crailsheim et al 2013). In 1981, an in vitro rearing protocol was suggested as a potential risk assessment tool that could be used to test the toxicity of pesticides to worker bee larvae (Wittmann and Engels 1981). However, early in vitro rearing techniques were plagued by pour survival of grafted larvae to adult emergence and inconsistent caste determination of the emerging adults. Since that time, several in vitro rearing protocols that offer moderate survival success have been developed (Vandenberg and Shimanuki 1987; Peng et al 1992; Aupinel et al 2005;

Crailsheim et al 2013). Recently, Schmehl et al. (2016) improved upon these protocols, consistently achieving adult emergence rates well above the OECD minimum of >70% survival (OECD 2015).

Despite improved in vitro rearing protocols, variability in survival success continues to be reported within and between laboratories (Aupinel et al 2010). Larvae that are grafted for in vitro risk assessments often come from a small number of source colonies (i.e. three or fewer colonies). Furthermore, the same source colonies are not always used between replicates or experiments over time. I suspected that the source colony from which one-day-old larvae are collected influences the survival percentages seen in experiments conducting in vitro rearing. To test my prediction, I designed a study to evaluate the survival rates of larvae from numerous source colonies in two rearing environments: their parental hive and in vitro.

Materials and Methods

Between June of 2015 and May of 2016, the survival rates of larvae reared in their parental hives were calculated and compared to the survival rates of larvae from the same parental hives reared in vitro. All colonies were housed in the University of

Florida apiary in Gainesville, Florida. Each colony was headed by a European-derived queen (Sheppard 1989). All hives were a single,10-frame Langstroth hive body, and managed according to regional best management practices throughout the duration of the study.

Brood Isolation

To initiate a trial, a queen was confined to a patch of empty brood comb (patch

A) as described by Schmehl et al. (2016). Twenty-four hours later (day 1), the queen was relocated to a new frame and confined to another patch of empty brood comb

(patch B) for an additional 24 hours. After the second confinement, the queen was released back into the hive and the confinement cages replaced onto both brood patches to ensure that no additional eggs were laid in those areas. For each colony, brood patch A was randomly assigned to either the colony-reared or in vitro-reared groups. B brood patch B was assigned the opposite rearing environment of brood patch

A for (e.g., if brood patch A was assigned colony-reared the brood patch B was assigned in vitro-reared and vise versa).

Colony Rearing Assay

On day 4, the confinement cage was removed and a piece of clear acetate

(21.59 cm × 28 cm) was overlaid on the brood patch (Human et al 2013). A section of brood from within the patch was outlined and empty cells were denoted on the acetate sheet. The total number of cells that contained larvae was determined, the acetate sheet removed, and the frame returned to the parental hive. Pre-pupal survival was assessed on day 11 by replacing the acetate sheet over the section of brood that was profiled on day 4 and noting any cells that had previously contained a larva, but were then empty. The number of cells that still contained viable brood was determined, the

acetate sheet removed, and the frame returned to the parental hive. On day 18, the frame containing the brood patch was collected from the hive and transported to the laboratory. The acetate sheet was again overlaid on the same section of brood that was profiled on day 4 and 11 and any cells that had previously contained brood, but were now empty were noted. Screen push-in cages were then placed over the brood patch and the frame placed in an incubator maintained at 35C and ~50% R.H. for 3 days

(Human et al 2013). On day 21, the frame was removed from the incubator, the push-in cage removed from the frame, and the total number of cells from which bees had emerged was counted. Any cells that had not emerged were opened manually to confirm that the individual inside was not viable. In total, brood survival percentages were calculated at day 11 for 25 colonies and at adult emergence for 14 colonies.

In vitro Rearing Assay

In vitro larval rearing was performed as described by Schmehl et al. (2016). In short, frames containing a patch of 1-day-old larvae were collected from the colony on day 4 and transported to the laboratory. The larvae were transferred from the comb to prepared sterile tissue culture plates containing 20 ul of artificial diet and maintained in an incubator at 35C and ~94% R.H. Each larva was fed 20, 30, 40, and 50 ul of artificial diet on days 6, 7, 8, and 9 respectively. The larvae were then transferred to another prepared sterile tissue culture plate for pupation once they had consumed all of the diet (between days 10 and 12). The pupation plates were incubated at 35C and

~75% R.H. Each larva was visually inspected every day. Any dead individuals were immediately removed from the plate at each inspection and the total number of surviving individuals was recorded. Adults were counted and removed from the pupation

plate as they emerged (between days 21 through 23). In vitro rearing survival percentages were calculated at day 11 and at adult emergence for 25 colonies.

Statistical Analysis

Survival percentages were analyzed using generalized linear mixed models methodology as implemented in SAS PROC GLIMMIX (SAS/STAT 14.1; SAS Institute,

Cary, NC) using the binomial distribution function with the default logit link function.

Rearing environment (colony, in vitro), time (day 11, adult emergence) and the interaction between the two were considered fixed effects. Experimental repeat (4 levels), colony (experimental repeat) and the interaction of treatment with colony

(experimental repeat) were considered random effects, the latter serving as the proper error term to test treatment. The residual covariance structure was modeled using various structures but none were able to improve the Generalized Chi-Square / df ratio fit statistics of 0.92. This indicates a good fit of the random model and no indication of over-dispersion. Interaction means were generated using the LSMEANS command in the abovementioned PROC and pairwise contrasts performed. Means and standard errors were back-transformed using the ilink option of the LSMEANS command. Final survival data were regressed on initial survival proportion (day 11) within a generalized linear mixed models environment. Predicted mean survival values were generated and plotted against initial survival.

Results

Percent survival means and standard deviations were as follows: 84.8% (±16.5,

25) to day 11 in the parental hive, 78.6% (±14.5, 14) to adult emergence in the parental hive, 94.4% (±6.5, 25) to day 11 in vitro, and 82% (±11.9, 25) to adult emergence in vitro (Fig. 2-1). There was a statistically detectable difference in survival percentage

between rearing environments (in vitro or colony) based on the time point (day 11 or adult emergence) (F= 8.5, df= 85, p=0.004). The mean survival percentage decreased over time within rearing environment (colony: p= 0.002, in vitro: p> 0.001). Furthermore, there was higher survival to day 11 in vitro than in the parental hive (p=0.013), but there was not a statistically detectable difference between in vitro and colony survival percentages at adult emergence (p=0.996; Table 2-1).

Day 11 colony survival percentage is predictive of adult emergence percentage within each rearing environment (Fig 2-2). The responses of adult emergence to day 11 survival percentage for both colony-reared, and in vitro-reared bees have the same slope (5.1 ± 0.6, 95% CI (3.9, 6.4), p< 0.001). The intercept of the colony-reared day 11 and adult emergence survival percentage relationship is -2.8 ± 0.5 (95% CI (-3.9, -1.7), p< 0.001), and the intercept of the in vitro-reared day 11 and adult emergence survival percentage relationship is -3 ± 0.5 (95% CI (-4.2, -1.9), p< 0.001). Furthermore, colony survival to day 11 is predictive of the in vitro adult emergence rate (slope= 3.3 ± 0.7,

95%CI (1.9, 4.7), p< 0.001; intercept= -1.2 ± 0.6,95%CI (-2.4, 0), p=0.048) (Fig. 2-3).

Discussion

Individual variation is well documented throughout the natural world (Houle 1992;

Lynch and Walsh 1997; Nettle 2006). Furthermore, colony-level variation in honey bees for parameters such as adult behavior and morphology is also well documented (Breed and Rogers 1991; Meixner et al 2013; Pirk et al 2013; Scheiner et al 2013; De Souza et al 2015). Correspondingly, multiple source colonies are recommended for use in laboratory risk assessments to account for colony-level differences in stress responses

(OEPP/EPPO 2010; Crailsheim et al 2013; OECD 2015). While the effect of colony- level variation is recognized in the physiological response of bees to stressors such as

pathogens and pesticides, the impact of colony-level variation on survival to adult emergence has been disregarded in in vitro rearing protocols. Moreover, colony-level variation in survival has been overlooked throughout the extensive effort to improve survival rates in in vitro rearing protocols (Peng and Jay 1977; Vandenberg and

Shimanuki 1987; Aupinel et al 2005; Silva et al 2009; Crailsheim et al 2013; Schmehl et al 2016).

Based on my results, I recommend that any potential source colonies be prescreened to determine if they will be ideal or undesirable for inclusion in experiments involving in vitro rearing. One method by which prescreening could be accomplished is to graft larvae from each colony and rear them in vitro. At adult emergence, the survival rate for each colony can be calculated and only colonies that score well above the

OECD guidelines of >70% adult emergence (OECD 2015) should selected for use in future studies that season.

Our data demonstrate that there is a predictive relationship between the day 11 survival percentages of colony-reared individuals to the adult emergence survival percentage of bees in vitro-reared from that parental colony. Day 11 survival assessments for potential source colonies are simple, low cost, and require very little time compared to rearing larvae from multiple colonies to identify colonies with high in vitro emergence rates (Human et al 2013; Schmehl et al 2016).

The predicted in vitro adult emergence rate of individuals grafted from a source colony that has 80% day 11 brood survival is also 80% with a 90% confidence interval of 71%-88%. Therefore, colonies that have day 11 survival percentages of ≥ 80% are recommended as source colonies for in vitro rearing because they are most likely to

foster in vitro survival percentages that met or exceed the OECD requirements of at least 70% survival to adult emergence (OECD 2015).

In addition to the ease and cost effectiveness of assessing day 11 brood survival in the colony, I suggest that brood survival in the colony at day 11 is a more accurate indicator of adult emergence rates than the day 11 brood survival in vitro. At day 11, brood survival was significantly higher in vitro than in the parental colony. However, there was no difference in adult survival between bees reared in the colony or in vitro.

This is primarily due to a window of mortality that occurs during the in vitro process between the pupal transfer and pupation stages. I suspect this discrepancy occurs because nurse bees detect and abort larvae that are subtly abnormal readily in the colony (Spivak and Gilliam 1998; Ibrahim and Spivak 2006), whereas larvae reared in vitro are maintained until they fail to pupate or die during or soon after pupation.

The correlation between parental colony and in vitro survival rates suggests that genetics may play a roll in brood survival. Larvae that are reared in vitro are removed from their colony within 24 hr of hatching from their eggs. Therefore, any potential environmental influences (e.g., pathogens) must affect eggs or larvae that are less than one day old in order to impact in vitro-reared larval survival. Considering that genetic factors may play a roll in in vitro survival rates, care should be taken throughout the research season to ensure that source colonies do not replace their queen via swarming, supersedure, or beekeeper requeening. If a requeening event does occur, the survival rate should be reassessed for that colony before using that queenline in an in vitro rearing experiment.

The results of this study emphasize the importance of source colony selection for any research involving in vitro rearing protocols. Underlying variation in colony brood survival can limit the ability of researchers to detect subtle effects of focal stressors in critical risk assessments. I offer a simple and cost effective strategy by which colonies that will have high in vitro survival can be identified by determining the survival percentage to day 11 of the brood in the hive. Further investigations are needed to determine if and/or for how long individual colonies maintain their respective brood survival percentages over time within a season and/or across multiple seasons.

Figure 2-1. The average percent survival by rearing environment (parental colony or in vitro) and time point (day 11 or at adult emergence) for developing honey bees. There is a significant effect of rearing on survival to day 11 (p=0.014) but not to adult emergence. Error bars depict one standard error from the mean.

Table 2-1. Means comparisons of the rearing environment by time point interaction effect observed for survival percentage. Data are: the rearing environment, the time point at which survival was assessed, the rearing environment by time point unique identifier, the mean ± SE survival percentage of each rearing environment by time point, the indication of which means are being compared, and the p value of each comparison. Rearing Survival Percentage Means Time Point Identifier p value Environment Mean ± SE Comparison colony day 11  89 ± 2.53 , 0.002 colony emergence  84.9 ± 3.4 , 0.013 in vitro day 11  93 ± 17.6 , > 0.001 in vitro emergence  84.9 ± 3.3 , 0.996

Figure 2-2. Predictive relationship between day 11 survival percentages and adult emergence survival percentages by rearing environment. Data were transformed back from the logit scale to proportion to communicate biologically relevant values. Day 11 colony survival percentage is predictive of the adult emergence colony survival percentage (top graph) and day 11 in vitro survival percentage is predictive of the adult emergence in vitro survival percentage (bottom graph).

Figure 2-3. Predictive relationship between the day 11 survival percentages in the parental hive and the adult emergence survival percentages of larvae grafted from those parental hives and reared in vitro. Data were transformed back from the logit scale to proportion to communicate biologically relevant values.

CHAPTER 3 COMPARATIVE MORPHOLOGY OF ADULT HONEY BEES REARED IN VITRO OR IN THEIR PARENT HIVE

Introduction

The environment in which an individual develops can dramatically affect body condition and size. For example, temperature and nutrition have been demonstrated to affect body size in a variety of vertebrate and invertebrate species (Partridge et al 1994;

Metcalfe and Monaghan 2001; Angilletta et al 2004; Cassidy et al 2014; Scofield and

Mattila 2015). Many free-living organisms develop in variable and stochastic environments where it is likely that they will encounter non-ideal developmental conditions (Feeny 1970; Awmack and Leather 2002). In contrast, honey bee larvae develop in a relatively constant environment within the broodnest of their parental colony (Seeley 1985; Winston 1991; Moritz and Southwick 1992; Schmickl and

Crailsheim 2004).

While the honey bee colony, as a whole, may experience environmental variation

(i.e. temperature fluctuations or pollen limited periods, Jeffree and Allen 1957), the brood typically are sheltered from such stressors. In fact, a honey bee colony will reduce the number of larvae being reared in times of limited pollen availability (Hellmich and Rothenbuhler 1986) or poor weather conditions (Schmickl and Crailsheim 2002).

Moreover, nurse bees will cannibalize younger larvae to maintain sufficient protein intake so they can produce adequate volumes of worker jelly to feed older larvae in times of severe resource limitation (Schmickl and Crailsheim 2001).

Artificial rearing of honey bee larvae is an important risk assessment tool

(Hendriksma et al 2011; Crailsheim et al 2013; OECD 2015) to assess the risk of numerous potential stressors, including pesticides and pathogens, on developing honey

bees. However, the developmental environment of larvae reared in vitro is vastly different in nutrition and physical atmosphere than that of larvae reared in their parental colony (Seeley 1985; Winston 1991; Crailsheim et al 2013).

Differences between the nutrition of larvae reared in vitro or by their parental colony is of particular importance when considering potential developmental differences of larvae reared in both environments. Empirical field studies have been conducted to demonstrate that the body size of adult honey bees is affected by the quality of diet they received while larvae (Kunert and Crailsheim 1988; Daly et al 1995; Mattila and Otis

2006). A colony produces smaller adult bees when pollen access is restricted via pollen trapping and/or removal of bee bread from the hive (Scofield and Mattila 2015).

Conversely, larger-than-normal adult honey bees are produced when overfeeding of the larvae is stimulated (Allsopp and Crewe 1993). Furthermore, female caste- differentiation (queen versus worker) is nutritionally mediated in honey bees (Weaver

1966; Page and Peng 2001). Not only can overfeeding result in larger adult workers, but overfeeding can also stimulate female larvae to develop into queens or queen/worker intercastes rather than workers (Boot et al 2006; Buttstedt et al 2016).

Early in vitro rearing methods typically called for multiple larvae to be reared in a single well while provided an artificial diet ad libitum (Rembold et al 1974; Rembold and

Lackner 1981; Shuel and Dixon 1986; Vandenberg and Shimanuki 1987). These methods often resulted in high mortality rates and created large workers and queen/worker intercaste individuals. More recently, several methods have been developed that individually feed larvae a controlled volume of diet (Aupinel et al 2005;

Silva et al 2009; Schmehl et al 2016). However, these methods have the potential to limit the nutritional resources of the developing larvae.

Multiple investigators have indicated that the adult honey bees reared via modern in vitro rearing methods are workers, not queens or queen/worker intercastes

(Kucharski et al 2008; Kaftanoglu et al 2011; De Souza et al 2015). However, very few have investigated potential differences in the morphological development of workers reared in vitro compared to workers reared by a colony (Herbert et al 1988;

Brodschneider et al 2009; Kaftanoglu et al 2010), and none have analyzed the morphology of bees resulting from the most recent methodological update by Schmehl et al. (2016).

We must fully understand the physiology of the individuals that result from in vitro rearing for the technique to be a robust tool in future research on honey bee health and development. The first step in achieving that goal is to determine the degree to which in vitro rearing affects the morphology of resulting adult bees. To address this knowledge gap, I conducted a study to compare the external morphology and dry body weight of adult workers that were reared in vitro to adult workers that were reared by their parent colony.

Materials and Methods

The dry body weights, external morphology, and ovariole number of adult worker honey bees that were reared in vitro or by their parent colony were analyzed. Individuals were collected from 14 European-derived honey bee colonies between September 2015 and May 2016. All colonies were housed in the University of Florida apiary in

Gainesville, Florida, and maintained in a single ten-frame Langstroth hive body throughout the duration of the study.

Larvae were reared in vitro as described by Schmehl et al. (2016). The mated, laying queen of each colony was caged on a patch of empty brood comb and returned to the hive for 24 hr. The queen was then relocated to a new patch of empty brood comb on another frame, caged, and returned to the hive for an additional 24 hr.

Following the tandem caging events, the queen was released back into the hive, the exclusion cages were returned to the brood patches to ensure no additional eggs were laid in the combs, and the frames were returned to their parental colonies. The starting day of each caging event was regarded as day 0 for each brood patch. The brood patches were then randomly assigned to the in vitro or colony-reared treatment groups so that each hive had one frame assigned to both treatments.

On day 4, brood patches assigned to the in vitro treatment were removed from their parent hive and transported to the laboratory and grafted into prepared sterile tissue culture plates and maintained in an artificial rearing environment as described by

Schmehl et al. (2016). Brood patches assigned to the colony-reared treatment group remained in their parent hive until day 18, at which point they were transported to the laboratory and housed in an incubator at 35C and ~50% R.H. for 3 days (Human et al

2013). As adults emerged, 15 individuals were randomly collected from both treatment groups for each colony and individually placed into 0.5 ml microcentrifuge tubes and immediately preserved at -80C.

Dry Bee Weight

Dry weights were calculated for five emerged bees from both treatment groups for each of the 14 colonies. Each microcentrifuge tube containing a frozen adult bee was weighed using a Mettler Toledo AL204 analytical balance. The microcentrifuge

tubes were then opened, placed into microcentrifuge tube racks, and incubated at 60C

(Henderson 1992). After seven days of incubation, 20% of the tubes containing dried bees were randomly selected, weighed, and returned to the incubator. The following day, the same subset of tubes containing dried bees were reweighed to confirm that no additional weight loss had occurred since the previous day, thus indicating the specimens were dry (Human et al 2013). The remaining tubes of dried bees were weighed. After each tube containing a dry bee was weighed, the dry bee was removed from each tube and the empty tube was weighed again. Dry bee weight was determined for each bee as follows: dry bee weight = weight of tube containing dry bee – weight of empty tube.

External Morphology

Five individuals from each rearing environment per colony were dissected and the following external features were measured: forewing length, forewing width, hind wing length, head width, and hind leg basitarsus length (Fig. 3-1; Meixner et al 2013; De

Souza et al 2015). Additionally, the following worker-specific features were categorized as worker, queen, or queen/worker intercaste for each bee: mandibular notch, pollen brush, corbicula (Fig. 3-2; Shi et al 2011; Buttstedt et al 2016). Bees were individually dissected and each body part was photographed via a Leica M205 microscope with a

Leica MC170 camera and Leica Application Suite (Leica Microsystems Limited, version:

4.8.0, build: 154). Images were analyzed with custom-built, assistive measuring software.

Statistical Analysis

Response variables were analyzed using mixed models methodology as implemented in SAS PROC MIXED (SAS/STAT 14.1; SAS Institute, Cary, NC). Rearing

environment (parental colony, in vitro) was the only fixed effect. Colony and colony x treatment interactions were random effects, the latter serving as the proper error term to evaluate treatment differences. Because of heterogeneity of variances, I employed the group option in the random statement to calculate group-specific standard errors.

Results

Statistically detectable differences were observed between some, but not all, of the features that were quantified for bees from both rearing environments. There was a statistically detectable decrease in forewing length and width and dry body weight of bees reared in vitro compared to bees reared in their parental hive. However, there were not statistically detectable differences between the treatment groups for hind wing lengths, head widths, or basitarsus widths of bees reared in vitro or in their parental hive. Results of the quantitative trait analyses are detailed in Table 3-1. Only worker- specific qualitative features (un-notched mandibles, pollen brushes, and corbiculae) were observed on all individuals.

Discussion

The present study confirmed that honey bees that have been reared in vitro via the Schmehl et al. (2016) method develop into adults without gross morphologic deformations. Furthermore, all of bees expressed worker-specific characteristics (un- notched mandibles, pollen brushes, and corbiculae). These findings support the use of the Schmehl et al. (2016) artificial rearing method in risk assessments.

There was not a statistically detectable affect of rearing environment on hind wing length, head width, or basitarsus length. However, statistically detectable effects of rearing environment on forewing length and width and dry body weight were identified.

This may suggest that basitarsus, head, and hind wing size may be under more

canalizing pressure and that body weight and forewing size may be more phenotypically plastic.

The reduction in forewing size that I observed in in vitro-reared bees aligns with previous findings of decreased forewing surface area in bees reared in vitro via the

Aupinel et al. (2005) method (Brodschneider et al 2009). Brodschneider et al. (2009) also reported decreased hind wing surface area in bees that were reared in vitro, though hind wing size did not differ between the two groups in the present study.

Furthermore, Herbert et al. (1988) found that European bees that were reared in vitro via the Vandenberg et al. (1987) method were more likely to be misclassified as

Africanized by USDA ID (Daly et al 1982): presumably due to reduction in forewing and/or hind wing size (Herbert et al 1988). All three studies utilized different rearing methods and different measures of forewing size. However, all three studies identified reduced forewing size in artificially reared bees. Hind wing length was not significantly reduced in my bees that had been reared in vitro and it is not clear if hind wing size was a contributing factor in the misclassification of in vitro-reared bees as Africanized in

Herbert et al. (1988) analysis.

Brodschneider et al. (2009) and Herbert et al. (1988) both reported similar fresh body weights of bees reared in vitro or in a hive. Brodschneider et al. (2009) also measured dry body weights of those bees and identified statistically detectable differences between bees reared in the hive and bees reared in vitro, similar to what I found in this study. Brodschneider et al. (2009) dissected body regions to identify that the difference in dry body weight was due to reduced thoracic mass. They hypothesized that the decreased thoracic mass was cause by reduced muscle development.

However, despite reduced wing surface area, and reduced musculature, the flight ability of in vitro-reared bees was not impacted when fed a low concentration of sucrose and flight speed was only slightly decreased when fed a high concentration of sucrose

(Brodschneider et al 2009).

There are significant differences in the nutrition, physical environment, and social environment of bees reared in vitro compared those reared by their parental colony

(Seeley 1985; Winston 1991; Crailsheim et al 2013; Schmehl et al 2016). Therefore, it is not possible to say that observed differences are caused exclusively by nutritional differences. However, we do also see reduced dry body weights and reduced wing size of adult bees reared during time periods of nutritional stress in field studies (Kunert and

Crailsheim 1988; Daly et al 1995; Mattila and Otis 2006). This suggests that the in vitro diet likely contributes to the morphological variation that I observed in my in vitro-reared bees compared to my colony-reared bees.

External morphometry offers quick and simple metrics by which to quantify differences between treatment groups in risk assessments. However, investigations of dry body weight and morphological characters are limited in what inferences can be made about the biological relevance of any potential abnormalities. Morphometric data are a valuable first step in comparisons, but they should be considered in conjunction with other biologically relevant assays because changes in morphology do to not guarantee changes in behavior (Gordon 2016).

Field studies have demonstrated that small differences in the rearing environment can modify honey bee behavior and possibly physiology (Tautz et al 2003;

Jones et al 2005; Becher et al 2009). Simple survival rates and/or morphological

analysis are not sufficient to declare that bees reared in vitro are normal or abnormal.

Many other parameters of bee behavior and physiology remain unexplored in regards to how artificial rearing may affect the resulting bees. Further investigation is needed to explore how in vitro rearing my impact other facets of honey behavior and physiology.

Figure 3-1. External features measured for morphological comparisons between adult honey bees that were reared in vitro or in their parental colony. Specific measurements are A) hind leg basitarsus length, B) head width, C) forewing length, D) forewing width, and E) hindwing width. Photo courtesy of author.

Figure 3-2. External morphological characteristics used to categorize individuals as worker honey bees, queens, or queen/worker intercastes. The worker characteristics shown are A) pollen brush, located on the basitarsus of the hind legs, B) corbicula, located on the tibia of the hind legs, and C) a smooth (un-notched) mandible. Photo courtesy of author.

Table 3-1. Summary of the results of each quantitative morphometric trait measured. Data are: 1) the features that were measured and the unit of measure for each feature, 2) the averages and standard errors of each feature for colony- reared and in vitro-reared individuals, and 3) the p value of the statistical comparison between row means. Featured Measured Colony-Reared In Vitro-Reared p Statistical (unit of measure) (mean ± SE) (mean ± SE) value Significance

Dry Weight (g) 0.021 0.016 0.003 *

Forewing Length (mm) 9 ± 0.07 8.48 ± 0.09 >0.001 *

Forewing Width (mm) 3.14 ± 0.03 2.99 ± 0.03 0.002 *

Hind Wing Length (mm) 4.83 ± 0.03 4.60 ± 0.13 0.108

Basitarsus Length (mm) 2.11 ± 0.06 2.19 ± 0.2 0.716

Head Width (mm) 3.82 ± 0.05 3.74 ± 0.03 0.175

CHAPTER 4 EFFECTS OF ARTIFICIAL REARING ENVIRONMENT ON THE BEHAVIOR OF ADULT HONEY BEES

Introduction

There are varying degrees of sociality observed throughout the animal kingdom.

Honey bees have achieved the highest level of social organization, eusociality.

Eusociality is defined by three characteristics: 1) overlapping generations (multiple generations exist within the colony), 2) cooperative brood care (adults within the colony care for larvae that are not their own offspring), and 3) reproductive division of labor (a non-reproductive worker caste cares for the immature offspring of the reproductive caste; Wilson 1971). There are elaborate social structures within eusocial colonies, and the fitness of the colony is dependent upon proper execution of a suite of behaviors by the individual members (Seeley 1985; Winston 1991).

The queen is the only reproductive individual within the honey bee colony. The remainder of an established colony is comprised of the queen’s daughters (workers) and sons (drones). Honey bee workers perform all brood care, hive maintenance, and foraging. Each worker will perform a series of colony tasks, in a predictable order, throughout its lifetime (Seeley 1985; Huang and Robinson 1996; Schmickl and

Crailsheim 2004). Generally, the youngest workers perform tasks within the brood nest

(e.g. cell cleaning and capping, brood and queen tending). These bees then progress to food handling and storage related activities (e.g. receiving and storing pollen and nectar, comb building, cleaning debris). The oldest workers in the colony take on tasks outside of the hive (e.g. ventilation, guarding, and foraging).

The importance of social interactions during juvenile development to adult behavior has been well documented in vertebrate species (Freedman et al 1961; Einon

and Morgan 1977; Bateson 1979; Bailey Jr et al 2001; Knudsen 2004). Furthermore, juvenile social interactions influence adult behavior in other species of arthropods

(Hebets 2003; Strodl and Schausberger 2012). However, the extent to which the developmental environment may affect adult behavior is not well understood in honey bees.

Historical investigations of how the developmental environment may affect adult honey bees are limited to crude manipulations of the colony’s food stores, artificial pupal incubation, or experiments where brood are cross-fostered in non-parental colonies. For example, it has been demonstrated that suboptimal larval nutrition (Scofield and Mattila

2015) and pupal incubation temperatures (Tautz et al 2003; Jones et al 2005; Becher et al 2009) impact foraging behavior of the resulting adult workers. Additionally, cross- fostering studies have been used to assess morphological and behavioral differences between subspecies of A. mellifera (Allsopp and Crewe 1993; Schneider et al 2003;

Alaux et al 2009) and caste differentiation (Weaver 1957).

There have been significant improvements made in our ability to rear honey bee workers artificially since the technique was first implemented in 1933 (Crailsheim et al

2013). Historically, investigators using artificial rearing techniques have struggled with poor adult survival and incomplete caste differentiation in their studies (Rembold et al

1974; Rembold and Lackner 1981; Shuel and Dixon 1986; Vandenberg and Shimanuki

1987; Aupinel et al 2005; Kaftanoglu et al 2010; Kaftanoglu et al 2011). However, recent refinements to the methodology consistently produce viable adult workers (as reviewed by Crailsheim et al 2013), thereby offering the opportunity to examine the impact that developmental environment may have on honey bee behavior.

Vast differences exist between the physical and social environments of honey bees that develop in vitro and those that develop in their parental hive. Notably, in vitro- reared larvae are maintained in a solitary environment, without exposure to adult bees

(Schmehl et al 2016). Normally, honey bee brood develop in a dynamic social environment in the hive. A larva reared within a hive may be tended by nurse bees as many as 2,785 times during its development (Schmickl and Crailsheim 2002). In contrast, in vitro-reared bees are fed an artificial diet by a pipette tip once a day for five days in the laboratory (Aupinel et al 2005; Silva et al 2009; Schmehl et al 2016). Larvae that are reared in vitro are dramatically deprived of physical and chemical interactions with adult nurse bees.

To date, in vitro rearing has been used to study the effects of pesticides, pathogens, and transgenic plants on immature honey bees (as reviewed by Crailsheim et al 2013). In contrast, only one study has been conducted to assess how in vitro larval rearing may affect adult honey bee physiology (Brodschneider et al 2009), and no behavioral analyses have been conducted. To address this knowledge gap, I conducted a multifaceted study to evaluate the behavior of in vitro-reared honey bees in a variety of controlled laboratory assays relating to the defining characteristics of eusociality. The behaviors selected for this study were queen recognition, brood rearing, trophallactic food sharing, and sucrose responsiveness. These behaviors were determined to be of particular interest based on their fundamental importance to the social structure of a eusocial colony (Wilson 1971; Seeley 1985; Winston 1991)

The workers’ ability to recognize the queen is critical to stabilizing the social structure of the colony. Worker recognition of queen mandibular pheromone stimulates

care of the queen, suppresses the production of new queens, and promotes cohesion of the swarm during colony reproduction (Butler 1954; Velthuis and van Es 1964; Winston et al 1989; Winston and Slessor 1992). Queen recognition by workers is the foundation of the reproductive division of labor in the colony. Interruption of the workers’ ability to recognize the queen would have catastrophic affects and likely result in colony failure.

Parental care is broadly known to increase the survival rate and condition of offspring (Alcock 2009). Honey bee workers forego their own reproduction to rear the queen’s offspring, their brothers and sisters. Brood rearing in the hive is intensive: attending nurse bees and larvae communicate through chemical and tactile cues to optimize larval care. Reduced efficiency in brood rearing decreases colony fitness, honey production, and pollination efficiency (Free 1967; Woyke 1984; Delaplane et al

2013).

Food transfer in the honey bee colony is not limited to brood care. Adult bees also share food with other adults via trophallaxis. Trophallactic food sharing is extensive among honey bees within a hive. Therefore, a colony is regarded as having a

“communal stomach” (Wilson 1971; Moritz and Southwick 1992). In that, food is shared so efficiently that the contents of the stomachs of all of the bees in the hive can be cumulatively regarding as the stomach of the entire colony. The communal stomach concept was demonstrated in an experiment where six foragers where trained to a feeder of radiolabeled sugar syrup. Fifty percent of the adult bees in the colony were radioactive within 24 hr because they had been fed the radiolabeled diet (Nixon and

Ribbands 1952). Degradation of trophallactic interactions could have dramatic effect on the task allocation of a colony (Gordon 2016). Trophallaxis is thought to be one of the

mechanisms by which that the colony assesses its own condition. The frequency at which adult bees are begging to be fed via trophallaxis is thought to communicate how much food is readily available in the hive. If the frequency of trophallactic requests increases this may signal to younger nurse bees that there is a food shortage in the hive, and stimulate precocial development of those nurse bees into foragers (Robinson

1992; Schulz et al 2002).

Sucrose responsiveness is integral to foraging success. The sucrose content of a nectar source determines how much nectar a forager bee will collect on a foraging trip

(von Frisch 1967; Schmid-Hempel et al 1985), how quickly that nectar will be received by house bees at the hive (Seeley 1996), how likely that forager is to perform a recruitment dance for that floral patch (von Frisch 1967; Seeley 1996), and how likely that forager is to return to that patch (Frisch 1967). The foraging efficiency of the colony could be dramatically impaired if its foragers’ abilities to differentiate between high and low sucrose concentrations were compromised, and decreased foraging efficiency is likely to have a negative impact on colony health (Wray et al 2011).

I hypothesized that in vitro rearing may reduce resulting adult worker honey bees’ ability to preform all or some of the focal behaviors commonly performed by bees that have been reared in their parental hive. For example, bees reared in vitro may fail to recognize a queen, or show a reduced response to a queen when compared to bees that were reared by their parental colony.

Materials and Methods

All trials were conducted March - July 2017 in Gainesville, Florida, USA. All source colonies were headed by a European-derived queen, and maintained in hives consisting of a single, ten frame Langstroth hive body. In March 2017, the in vitro

rearing success rate was determined for each of ten colonies by rearing 96 individuals to adult emergence as described by Schmehl et al. (2016). Any colony that had >80% adult emergence was selected for use in the laboratory behavior assays.

A new cohort of bees was reared for each behavioral assay. Each cohort was generated from four or five source colonies. The queen of each source colony was caged on a frame of empty brood comb (day 0). After a 24-hour confinement, each queen was released back into her colony and the cage replaced over the brood patch to ensure that no additional eggs were laid within the patch. On day 4 the brood patches, now containing 4 - 28-hr-old larvae, were transported to the laboratory and larvae grafted from them into prepared sterile tissue culture plates. The grafted larvae were fed a series of artificial diets on days 4, 6, 7, 8, and 9, and individually transferred to a new sterile tissue culture plate for pupation when they had consumed all of their diet (days

10 or 11; Schmehl et al 2016). Larvae were incubated at 35C and ~94% R.H. Pupae were incubated at 35C and ~75% R.H.

On day 21, a frame that contained emerging adult bees was collected from the hive of each source colony used in the grafting cohort. The emerging brood frames were transported to the laboratory and individually caged in an incubator at 35C and

~50% R.H. (Human et al 2013). As adult bees emerged from the colony frames and in vitro plates, they were collected and transferred to cup cages (Williams et al 2013;

Fleming et al 2015). Bees were grouped on the cup cages by their source colony and rearing environment (in vitro-reared or colony-reared). The cups were randomly distributed in an incubator maintained at 35C and ~30% R.H. Bees in each cup cage had ad libitum access to water, sucrose solution (50:50 sucrose weight: water volume),

and pollen patty (50:50 pollen weight:sucrose solution volume; Williams et al 2013;

Fleming et al 2015). Each cup was assessed daily and any dead individuals were removed.

Queen Recognition

Queen recognition was quantified based on the methodology described by

Kaminski et al. (1990). Age dependence has not been documented in queen response assays (Allen 1960; Kaminski et al 1990). Therefore, I used eight-day-old workers for the queen recognition assay because this falls within the average age span of queen attendants documented in observation hives [5.5 days (Seely 1979) and 10.7 (Allen

1960)].

Four source colonies were used for this bioassay. Two replicates of both rearing environments, in vitro-reared and colony-reared, were conducted for each source colony (16 total experimental units). Groups of ten, eight-day-old workers were transferred from their respective cup cages into petri dishes (100 mm × 10 mm) modified with a 1 cm diameter hole in the center of the bottom of the dish. Each petri dish also was equipped with a feeder through the lid of the dish to provide the bees access to sucrose solution (50:50 sucrose weight: water volume). The petri dishes of bees were transferred to a dark room that was illuminated with red light and were allowed to acclimate for 1-1.5 hr.

Eight mated, European-derived queens were obtained from a local queen breeder. Each queen was restrained by wrapping a 1 cm × 3 cm piece of paraffin film around the thorax to facilitate observation of worker contact throughout the assay.

Observational assays were divided into two trial groups. Trial group “A” consisted of one replicate of each rearing environment/source colony combination and trial group “B”

consisted of the other replicate of each rearing environment/source colony combination.

A restrained, mated queen was introduced into each of the petri dishes containing ten worker bees. Activity in all eight petri dishes was recorded for 5 min (Fig. 4-1). The queens were removed from the group A petri dishes and randomly introduced into the group B dishes, at which point the activity in the group B dishes was recorded for 10 min.

Still images of the video were inspected at 30 second intervals (0:30, 1:00, 1:30,

2:00, 2:30, 3:00, 3:30, 4:00, 4:30, and 5:30 min). The number of bees touching the queen was counted at each 30 sec time point, and the time point counts were totaled to calculate the total touch count for each petri dish (Allen 1960; Kaminski et al 1990).

Additionally, the total bee seconds of contact and average seconds of contact/bee were calculated using BORIS [v. 2.999 - 2016-10-28 (Friard and Gamba 2016)] behavioral analysis software. Videos were played back in BORIS and time stamped when 0 – 10 bees were touching the queen. Those timestamps were then used to determine the seconds that the queen was being contacted by 0 - 10 bees during the 5 min observation. Bee seconds were calculated for each bee number category (0 - 10 bees) by multiplying the time that a queen was being contacted by a given number of bees (0

– 10 bees) by the number of bees. For example, if a queen was contacted by two workers (number of bees) for a total 30 sec during the 5 min observation, the bee seconds for the “2 bees” category would be 60 sec (two workers × 30 seconds). The bee seconds for each number of bees were then totaled to determine the total bee seconds of contact for that petri dish. The average seconds of contact/bee was

calculated by dividing the total bee seconds by the number of bees in the dish (ten bees in each dish).

Brood Rearing

Brood rearing behavior was evaluated as described by (Shpigler and Robinson

2015). The brood rearing assay was conducted twice due to a limitation in the number of queen cells available during the first round. Five source colonies we used in the first round, and four of those five colonies were used again in the second round. In both rounds, two replicates of both rearing environments were conducted for each source colony (for a total of 36 experimental units).

Brood rearing cages were constructed as described by (Shpigler and Robinson

2015). Deep petri dishes (100 mm × 20 mm) were oriented vertically. Wax foundation was affixed to the back of each dish. Furthermore, a 1 cm hole was cut into the top of dish. The hole was made to accommodate a queen cell. The sides of the dishes were equipped with feeders for the purpose of delivering sucrose solution (50:50 sucrose weight: water volume) to the bees. Pollen patty (50:50 pollen weight:sucrose solution volume) was availible ad libitum at the bottom of each dish (Fig. 4-2). Ten, seven-day- old bees were transferred from the cup cages to the pertri dishes and returned to the incubator to acllimate for 24 hr.

This assay also requires that a four-day-old queen cell be inserted into the top of each dish. Queen cells were obtained by grafting one-day-old female larvae into wax queen cups and introducing those larvae into a starter hive (Connor 2009; Büchler et al

2013). Four-day-old queen larvae were collected from the starter hive and transported to the laboratory when workers in the the perti dishes were eight days old and had acclimated for 24 hr. A queen cell was introduced into the top of each petri dish and the

behavior of the workers toward the cell was recorded for 5 min under red light in the incubator. The queen cell remained in the petri dish overnight (Fig. 4-2), except for a subset of the petri dishes in round one in which there were not enough queen cells for every dish to care for a queen cell overnight. The next day, each petri dish that had a queen cell was categorized as having capped or not capped that queen cell.

The total amount of time that each queen cell was tended (i.e. a worker had her head inside the queen cell) was quantified using BORIS (Friard and Gamba 2016) behavioral analysis software. Videos were played in BORIS and time stamped when any bee in the petri dish had its head inside the queen cell. The timestamps were used to determine the total time that each queen was tended by the workers in the dish.

Trophallaxis

Trophallactic food sharing was quantified as described by Moritz and Hallmen

(1986). One-day-old workers were selected as the donor bees and eight-day-old bees were selected as the recipient bees. Moritz et al. (1986) demonstrated that one-day-old bees frequently share food via trophallaxis, while eight-day-old bees are less likely to engage in trophallaxis. Therefore, by pairing a one-day-old donor with ten, eight-day- old, recipients I increased the likelihood that the food exchange quantified in the assay is due to primary trophallaxis between the donor and each recipient, rather than secondary sharing amongst the recipients.

All donor/recipient pairings were composed of bees from the same source colony and a full factorial design was used to compare trophallactic exchange between (1) an in vitro-reared donor and in vitro-reared recipients, (2) an in vitro-reared donor and colony-reared recipients, (3) a colony-reared donor and in vitro-reared recipients, and

(4) a colony-reared donor and colony-reared recipients. Each donor/recipient × rearing

environment pairing was repeated twice for each of the five source colonies, for a total of 40 experimental units.

Each one-day-old donor bee was collected from its respective cup cage, marked on the thorax with a white paint pen, and individually restrained onto a piece of

Styrofoam via dissecting pins crossed over the gaster (Fig. 4-3). Ten, eight-day-old recipients were collected from their respective cup cages and placed in a petri dish (100 mm × 10 mm) without access to food and water. Donors and recipients were transported to a dark room with red light and starved for 1-1.5 hours. Following the starvation period, each donor was fed 20 ul of sucrose solution (50:50 sucrose weight: water volume) that contained 0.25% rhodamine B (Sigma-Aldrich CAS: 81-88-9). After feeding, the donor bees were immediately introduced into their assigned dish of recipient bees.

The recipients and donors were allowed to interact for 1 hr (Fig. 4-3), at which point the petri dish of bees was frozen at -80C. Four days later, the digestive system was dissected from each bee and squashed onto Whatman qualitative filter paper

(grade 1). The stomach contents were visually inspected for florescence in a dark room with an ultraviolet light box (Fig. 4-4). The number of recipient bees that had florescent stomach contents was totaled and divided by the number of recipients in the cage to generate the proportion of bees fed via trophallaxis for each petri dish.

Sucrose Responsiveness

Sucrose responsiveness was quantified as described by Page et al. (1998),

Rueppell et al. (2006), and Scheiner et al. (2013). Four source colonies were utilized in this assay. Six to eight individuals were tested per rearing environment from each

source colony, for a total of 60 experimental units. Workers that were 24 days old were used as this age coincides with the age at which these bees would be foraging if they were in a hive (Seeley 1982).

Adult bees were individually restrained (Fig 4-5; Ciarlo et al 2012) and starved for

1-1.5 hr. Bees were not anesthetized for restraint because anesthetization has can affect sucrose responsiveness (Pankiw and Page 2003; Frost et al 2011). Each bee was presented with a series of sucrose solutions (0.1%, 0.3%, 1%, 3%, 10%, and 30% sucrose weight: water volume). For each sucrose trial, a droplet of solution was formed on the tip of 1 ml syringe and touched to the antenna of the bee. The bee was recorded as having responded to that sucrose concentration if she extended her proboscis following antennal stimulation (Fig 4-5). Water was presented to each bee in the same manner as the sucrose solution exposures between sucrose solution exposures

(Rueppell et al 2006). There was a 2-3 minute period between each sucrose trial. The sucrose responsiveness was scored for each bee by totaling the number of times the bee extended its proboscis in response to the sucrose stimulation.

Statistical Analysis

Data were analyzed using generalized linear mixed models methodology as implemented in SAS PROC GLIMMIX (SAS/STAT 14.1; SAS Institute, Cary, NC) using the appropriate distribution and associated default link functions. The rearing environment (parental colony, in vitro) was the only fixed effect in the model (Table 4-1).

Random effects are detailed in Table 4-1; the residual error term was used to test the treatment effect. Means and standard errors were back-transformed using the ilink option of the LSMEANS command except for response variables analyzed using the lognormal distribution. In these instances, calculated means were exponentiated and

the standard error of back-transformed means calculated as SElog * back-transformed mean (delta rule).

Results

Queen Recognition

There were no statistically detectable effects of the rearing environment (parental colony or in vitro) on the average seconds of contact/bee (F=2.2, df= 1, p= 0.17) or the total touch count (F= 2.4, df= 10, p= 0.15). The average seconds of contact/bee for bees reared in their parent colony were 207.4 sec (±11.2, 8) and average seconds of contact/bee for bees reared in vitro were 185.4 sec (±11.2, 8). The average total touch counts for in vitro and colony-reared bees were 47.1 (±3.7, 8) and 55 (±4.2, 8), respectively.

Brood Rearing

There were statistically detectable effects of rearing environment on brood rearing behavior. There was a statistically detectable decrease in the average number of visits made to queen cells by bees reared in vitro (1.44 ± 0.6, 18) than for bees reared in their parental hive (5.12 ± 2.2, 18) (F= 20.4, df=61.3, p<0.01, Figure 4-6).

Furthermore, the average time that each queen cell was tended by colony-reared workers (146.8 ± 30.2 sec, 18) was higher than that of in vitro-reared workers (40.7 ±

30.1)(F=35.9, df= 58.3, p< 0.01, Fig. 4-7). However, there was not a statistical difference between cell capping behavior of bees reared in vitro (85.7%, 14) or those reared in their parent colony (92.9%, 14) (F= 0.05, df= 27, p= 0.83).

Trophallaxis

There were no statistically detectable effects of the rearing environment (parental colony or in vitro) of donors (F= 0.3, df= 36, p= 0.56) or recipients (F= 1.3, df= 36, p=

0.26) on the frequency of trophallactic food sharing. The average proportion of recipients fed by colony-reared donors was 0.93 (±0.03, 20) and the average proportion of recipients fed by workers reared in vitro was 0.89 (±0.03, 20). An average of 93%

(±3.03, 20) of colony-reared recipients were fed while 87.1% (±4, 20) of recipients reared in vitro were fed.

Sucrose Responsiveness

There was a statistically detectable effect of rearing environment on sucrose responsiveness (F= 6.6, df= 41.4, p= 0.01; Fig 4-8). The odds ratio indicates that colony-reared bees are twice as responsive to sucrose as bees reared in vitro, and that bees reared in vitro are twice as likely not to respond to any of the offered sucrose concentrations as bees that were reared in their parental colony. The average sucrose responsiveness score of colony- and in vitro-reared bees was 0.53 (±0.11, 31), and 0.34

(±0.11, 29), respectively.

A summary of the results for each experiment is presented in Table 4-2.

Discussion

This is the first study through which the development of honey bee behavior via artificial rearing has been assessed systematically. Queen recognition, brood rearing, trophallactic food exchange, and sucrose responsiveness are fundamental behaviors in honey bee colonies and interruption of these behaviors could dramatically impact colony fitness (Seeley 1985; Winston 1991). There were no statistical differences in trophallactic food sharing or queen recognition behaviors between bees reared in vitro and those reared in their parental colony. However, statistically detectable differences in the brood rearing behavior and sucrose responsiveness between bees reared in vitro or in their parental colony were identified.

Previous investigators have identified a myriad of environmental factors that influence trophallactic behavior, including climate, colony food stores, and season

(Schulz et al 2002). Additionally, the genetic relationship of the workers can affect how likely they are to share food via trophallaxis (Crailsheim 1998). It is important to note that the trophallaxis assay preformed here simply quantified the proportion of individuals that had any fluorescent sucrose in their crop, not the quantity of fluorescent sucrose that was shared with each bee. Therefore, any potential differences in the volume of food shared would not have been reflected in these data (Crailsheim 1998).

The amount of time bees spent tending a queen cell was reduced for bees reared in vitro compared to those reared by their parental colony, though cell-capping performance was not different between the two bee groups. Modified care of offspring by individuals that received modified parental care as immatures has been documented in other model organisms (e.g., rhesus monkeys and rats; Maestripieri 1999, Caldji et al

2000). Honeybees reared in vitro only receive parental care from their nest mates for up to 24 hours before being relocated to the artificial rearing environment. Our data highlight a potential effect of the social rearing environment on brood rearing behavior of adult honey bees. Alternatively, the variation in physical and nutritional environments that honey bee larvae are exposed to in vitro could result in physiological changes that make the resulting adults physically incapable of caring for brood (e.g., reduced hypopharyngeal glad development).

We did not observe nay difference in cell capping behavior of bees that were reared in vitro. However, this brood rearing assay utilizes four-day-old queen larvae.

Four-day-old queen larvae have already been supplied with a large provision of royal

jelly on which they will feed throughout the remainder of their larval development (Wang et al 2016). Nurse worker bees typically provide more food to a queen larva <24 hr before the queen’s cell is capped. However, a four-day-old queen larva can survive to pupation in the absence of additional provisioning by nurse bees (Shpigler and

Robinson 2015). Therefore, queen larvae that received reduced care from in vitro- reared workers were still viable individuals and the in vitro-reared bees responded appropriately by capping the queen cells.

In each assay, the rearing environment manipulation did not occur until larvae were approximately one day old. All larvae reared in vitro had been exposed to the physical environment of their parental colony for 3 days as an egg and up to one day as larvae prior to being transferred to the in vitro rearing environment (Schmehl et al 2016).

During that time, larvae would have been exposed to colony pheromones and interacted with nurse bees. Therefore, it is not possible to state definitively that environment has no effect on trophallaxis, queen recognition, or cell capping because in vitro-reared larvae were exposed to their natural rearing environment for a short period. However, my findings do add to the literature that suggest genetics play a large role in some honey bee behaviors (Robinson and Page 1989; Breed et al 1990; Guzmán-Novoa and

Page 1993).

Adult environment has been demonstrated to influence guarding and foraging behavior (Schulz et al 2002; Hunt et al 2003). My data suggest that the rearing environment of immature bees can also affect brood rearing and sucrose responsiveness. There is a systemic disregard for the juvenile rearing environment in apicultural research. Previously published works overlook the environment of the

developmental stages when investigators are attempting to determine the genetic origins of honey bee behavior. It is common for investigators to begin environmental manipulations, such as cross-fostering, after adult emergence (Boch and Rothenbuhler

1974; Winston and Katz 1981; Robinson et al 1989; Breed and Rogers 1991; Rueppell et al 2006; Hunt et al 2007; Alaux et al 2009). Our results underscore the impact that juvenile environment can have on adult bee behavior. I argue that the immature stages of honey bee development are dynamic and that future behavioral research will benefit from accounting for the influence that the rearing environment can have on adult behavior. For instance, experiments incorporating cross-fostering assays could be initiated during the egg stage (Calderone and Page Jr 1988; Calderone and Page

1992), rather than immediately prior to, or after, adult emergence.

In addition to providing insights into the origin of basic honey bee behaviors, my findings also relate to the relevance of in vitro rearing as a tool for assessing the risk of certain stressors, such as pesticides and pathogens, to honey bee health. There is a critical need for a robust risk assessment protocol for immature honey bee stages

(Hendriksma et al 2011; Crailsheim et al 2013) and the OECD has incorporated in vitro rearing methodology into their risk assessment protocols (OECD 2015). However, until now, the extent to which the in vitro rearing method may affect the behavior of the resulting bees has not been explored. Based on my findings, trophallaxis and queen recognition behaviors are good candidates for behavioral tests that can be added to larval risk assessments to expand our understanding of how honey bee stressors may affect the function of the resulting bees.

The physical, social, and nutritional environments of developing larvae differ dramatically between artificial and natural rearing conditions (Seeley 1985; Winston

1991; Crailsheim et al 2013; Schmehl et al 2016). In the present study, I am unable to determine which, or what combination, of these environmental differences most likely influenced the behavior of the bees that were reared in vitro. Further examination of the physiology and behavior of bees that have been reared in vitro in a broader array of assays in the laboratory and in the field promise to offer more insight into the extent to which rearing environment affects adult honey bees.

Overall, this study was designed with a reductionist approach, in that I examined the behavior of individual bees, rather than those behaviors that manifest at the colony level. However, the biological unit of honey bees is the colony, not the individual (Moritz and Southwick 1992). Biologically relevant behaviors are those that emerge at the colony level from the coordination of these fundamental individual behaviors (Gordon

2014; Jandt and Gordon 2016). The mechanism(s) by which collective behaviors materialize in social insects is not well understood. Therefore, I am unable to infer to what extent the behavioral changes that I observed in my in vitro-reared bees would affect the productivity of a colony composed exclusively of in vitro-reared bees. There is a need to evaluate the impact that rearing environment may have on the fitness of a honey bee colony further.

Figure 4-1. Images of the queen recognition assay. A) The eight petri dishes of trial group A during the recoding period. Workers can be seen clustering around the restrained queen in each dish. B) One petri dish of workers and a restrained queen under white light conditions following the recording periods. The restrained queen is visible on the right side of the image. The feeder containing sucrose solution can be seen in the lower left portion of the image. The 1 cm hole in the bottom lid of the dish is used for introducing the queen at the start of the trial and is visible in the top of the image. Photo courtesy of author.

Figure 4-2. Composite image of the petri dish cages in round two of the brood rearing assay. The feeder containing sucrose solution is inserted through the side of each cage. Pollen patty is available at the bottom of each cage. A queen cell in inserted through the top of each cage. Photo courtesy of author.

Figure 4-3. Images of the trophallaxis assay. Top: Eight donor bees that have been individually restrained for their starvation period via dissecting pins crossed over their gasters. Bottom: A petri dish of recipients during their interaction period with a donor bee. The donor bee (located toward the center of the image) is sharing food with one of the recipients, and can be identified by the white dot on her thorax. Photo courtesy of author.

A B

Figure 4-4. Images of the dissected stomach contents of the donor bee and ten recipient bees from one perti dish under white (A) and ultraviolet (B) light conditions. The donor bee’s stomach contents are at the top center of the each image. Only seven of the ten recipients have pink stomach contents visible under white light conditions (A). However, nine of the same ten stomach contents clearly fluoresce into ultraviolet light (B). Photo courtesy of author.

A B

Figure 4-5. Images of the sucrose sensitivity and responsiveness assay. The worker has been restrained so that she retains free movement of her head and forelegs. A) 30% sucrose solution is being presented to the forager by touching a droplet formed on the end of a 1 ml syringe to the antennae of the worker. B) The forager responds to the test solution by extending her proboscis. Photo courtesy of author.

Table 4-1. Summary of the distribution function, link function, and random effects for each response group/variable by social behavior. Response Distribution Link Social Behavior Random Effects Group/Variable Function Function queen recognition bee seconds of negative log colony, replicate contact binomial queen recognition total touch identity colony, replicate lognormal count

brood rearing total count of negative colony, experimental log queen cell visits binomial repeat, replicate

brood rearing total time colony, experimental visiting the lognormal identity repeat, replicate queen cell

brood rearing colony, experimental cell capping binary logit repeat, replicate

trophallaxis proportion fed binomial logit colony

sucrose overall binomial logit colony responsiveness responsiveness

Figure 4-6. The average total counts of queen cell visits. There was a statistically detectable decrease in the average number of visits made to queen cells by bees reared in vitro (IV; 1.44 ± 0.6, 18) than for bees reared in their parental hive (C; 5.12 ± 2.2, 18) (F= 20.4, df=61.3, p<0.01, Figure 4-6). Error bars depict one standard error from the mean.

Figure 4-7. The average seconds of contact/bee for honey bees reared in their parental colony (C; 146.8 ± 30.2 sec, 18) or in vitro (IV; 40.7 ± 30.1) were significantly different from one another (F=35.9, df= 58.3, p< 0.001). Error bars depict one standard error from the mean.

Figure 4-8. The average sucrose responsiveness scores for bees reared in vitro or by their parental colonies. A sucrose responsive score of 0 indicates that the individual never extended its proboscis in response to a sucrose stimulus. A sucrose responsive score of 6 indicates that the individual extended their proboscis at every sucrose stimulus (0.1%, 0.3%, 1%, 3%, 10%, and 30% sucrose weight: water volume). There was a statistically detectable effect of rearing environment on sucrose responsiveness (F= 6.6, df= 41.4, p= 0.01). Error bars depict one standard error from the mean.

Table 4-2. Summary of parameters tested to assess the effect of developmental environment (parental colony or in vitro) on adult worker bee behavior. Columnar data are: 1) the social behavior being tested (social behavior), 2) the specific parameter of each behavior being quantified (parameter measured); 3) indication of whether the mean for in vitro-reared bees was increased (∧) or decreased (∨) compared to colony-reared bees in there was a statistically significant difference (Directionality Difference); and 4) the p-value of the statistical test used to detect statistical significance. Directionality Social Behavior Parameter Measured p-value Difference queen recognition bee seconds of queen contact 0.152

queen recognition total queen touch count 0.171

brood rearing total count of cell visits ∨ <0.001 brood rearing total time tending the cell ∨ <0.001 brood rearing cell capping 0.829

trophallaxis proportion donors fed 0.564 sucrose responsiveness proportion of recipients fed 0.262

queen recognition sucrose responsiveness ∨ 0.014

CHAPTER 5 FIELD-BASED BEHAVIORAL OBSERVATIONS OF ADULT HONEY BEES REARED IN VITRO

Introduction

For social systems to succeed, individuals must work cooperatively towards the benefit of the group. This is especially true for obligately social, eusocial, insects

(Wilson 1971). A eusocial colony is viewed as a single “superorganism” rather than as a group of communal individuals (Moritz and Southwick 1992). Within the superorganism, each individual is somewhat analogous to a single cell within a mammal, and groups of individuals work in a coordinated manner to ensure that the colony functions normally.

The successful completion of colony level tasks requires the execution of a suite of behaviors by the individual members (Seeley 1985; Winston 1991).

Cooperative behaviors observed in social insects are often regarded as simple instinctual responses to positive and negative feedback (Moritz and Southwick 1992;

Beekman et al 2006; Seeley et al 2006; Schmickl et al 2009; Gordon 2016). However, learning and cognition also play a role in the behavior of individuals within eusocial colonies (Menzel 2012). For example, foraging honey bees learn pertinent cues though operant (Craig et al 2014) and classical conditioning (Giurfa and Sandoz 2012).

Furthermore, early social experiences have been demonstrated to affect adult behavior in other invertebrates such as wolf spiders, Schizocosa uetzi Stratton (Hebets 2003), and predatory mites, Phytoseiulus persimilis Athias-Henriot (Strodl and Schausberger

2012).

Honey bees are ideal model organisms to use to explore how social interactions that occur while the bees are immature impact their behavior as adults for several reasons. First, honey bee colony husbandry is well-understood, making it easy to

maintain colonies and bees for experimental purposes. Furthermore, standardized research methods that can be used to study honey bee behavior, morphology, and physiology have been published. Finally, normal adult behavior is known, providing a baseline to which the impacts of experimental manipulation on test bee behavior can be compared (Frisch 1967; Michener 1974; Seeley 1985; Page et al 2006; Robinson et al

2006; Dietemann et al 2013).

The majority of the investigations into honey bee behavior have focused exclusively on adult bees (Frisch 1967; Michener 1974; Menzel and Muller 1996; Seeley

1996; Scheiner et al 2013). Moreover, investigations into the genetic origins of honey bee behavior often overlook the potential impacts of developmental environment on behavior and begin experimental manipulations, such as cross-fostering, after adult emergence (Boch and Rothenbuhler 1974; Winston and Katz 1981; Robinson et al

1989; Breed and Rogers 1991; Rueppell et al 2006; Hunt et al 2007; Alaux et al 2009).

Few researchers have explored how a bee’s developmental conditions could alter it adult behavior. The investigations of developmental environment that have been completed highlight that worker behavior is not fixed. For example, immature bees that were subjected to suboptimal larval nutrition (Scofield and Mattila 2015) or abnormal pupal incubating temperatures (Tautz et al 2003; Jones et al 2005) were less capable of communicating forage locations to nest mates. However, suboptimal larval nutrition and suboptimal pupation temperatures may result in physiological changes that render the resulting adults incapable of performing adult behaviors rather than modifying the behavior itself. Moreover, these developmental experiments were conducted in a living honey bee colony. Experiments that occur within a hive are biased by many

uncontrolled factors such as colony strength, weather conditions, and food availability

(Hendriksma et al 2011). For example, Mattila et al. (2006) concluded that developmental nutrition played a role in adult behavior and physiology, but noted that they could not control for other environmental factors that likely influenced their findings.

Rearing honey bee larvae in the laboratory offers the opportunity to explore the extent to which the developmental environment of immature bees may affect their behavior as adults. Until recently, it was not possible to consistently rear immature worker bees to adulthood in an artificial environment (Crailsheim et al 2013). However, improvements to the artificial rearing methodology now allows for consistent survivability of immature bees to adulthood (Aupinel 2005, Schmehl et al 2016). This improved in vitro rearing protocol offers the opportunity to rear immature honey bees in a controlled environment that is free of social interactions and other potential biases encountered within honey bee hives (Crailsheim et al 2013; Schmehl et al 2016). Rearing bees in vitro provides a more controlled assessment of the effect of juvenile environment on adult behavior than do studies conducted in a living colony.

Vast differences exist between the environments that artificially and colony- reared larvae experience during their development, especially differences in the level of care they receive. Brood care within the colony is intensive: larvae may be tended by nurse bees as many as 2,785 times during their development (Schmickl and Crailsheim

2002). In contrast, larvae reared in vitro are deprived of all in-hive interactions including physical contact nurse bees, chemical communication with other colony members, and progressive feeding (Crailsheim et al 2013; Schmehl et al 2016).

It is possible that rearing environment can influence the resulting adult bee’s somewhat predictable progression through a series of age-related tasks (temporal polyethism – Seeley 1985). In general, young workers perform jobs in the central area of the hive where the brood resides. Young workers’ jobs include cleaning brood cells, feeding and tending the brood, and tending the queen. As they age, workers shift to performing tasks in the outer regions of the hive. These jobs include building comb, receiving nectar and pollen, storing nectar and pollen, processing honey and ventilating the hive. The oldest bees perform tasks outside of the hive, such as guarding the hive, removing dead bees from the hive, and foraging (Seeley 1985; Winston 1991).

If rearing environment does play a role in adult honey bee behavior, I would expect that bees reared in vitro would not perform their age-related tasks predictably as colony-reared workers would. Abnormal bees may be delayed or precocious in their progression through typical hive tasks, or potentially even fail to perform some tasks altogether. In the present study, I compared the behavior of in vitro-reared and colony- reared bees that were introduced into the same established colony to determine the extent to which the rearing environment affects the behavior of adult honey bees.

Materials and Methods

This experiment occurred March - August 2017 in Gainesville, Florida, USA. Four colonies were selected as source colonies based on their in vitro rearing survival rates

(>80% adult emergence) at the time of the study. All source colonies were maintained in hives consisting of single ten-frame, Langstroth hive bodies, were headed by a

European-derived queens and were managed according to local best management practices.

Larvae from each of the four source colonies were reared in vitro as described by

Schmehl et al. (2016). In short, the queen of each source colony was caged on a frame of empty brood comb for 24 hours. After this, the queen was released back into her colony and the exclusion cage replaced onto the brood patch to ensure that no additional eggs were laid in the patch after the designated laying period. Three days later (day 4), the frame was transported to the laboratory and the one-day-old larvae were grafted into prepared sterile tissue culture plates. A series of artificial diets were fed to the larvae on days 4, 6, 7, and 8, and 9. The larvae were transferred individually to a new sterile tissue culture plate for pupation once they had consumed all of their diet

(day 10 or 11). The larval plates were incubated at 35C and ~94% R.H. and the pupal plates were incubated at 35C and ~75% R.H.

On day 21, a frame of emerging adult bees was collected from each source colony used in the grafting cohort and transported to the laboratory. The frames were caged individually in an incubator at 35C and ~50% R.H. (Human et al 2013) to allow adult bees to emerge.

A subset of 15 emerging adult bees from both treatments (in vitro-reared and colony-reared) for each source colony were individually identified using queen marking discs (Human et al 2013) and placed into cup cages (Williams et al 2013; Fleming et al

2015) by their treatment group. Bees in each cup cage had access to ad libitum water, sucrose solution (50:50 sucrose weight:water volume), and pollen patty (50:50 pollen weight:sucrose solution volume; Williams et al 2013; Fleming et al 2015). The cups were incubated at 35C and ~30% R.H.

An observation hive consisting of three deep frames of drawn comb (one frame of pollen/honey/nectar and two frames of brood) was established prior to the emergence of the experimental bees. A 5 cm2 grid was drawn on the glass of both sides of the observation hive. The observation hive was headed by a European-derived queen and stocked with a mix of workers from three other colonies in the apiary. Neither the queen nor the workers in the observation hive were related to those from the source colonies.

The observation hive was maintained in a dark room and viewed exclusively under red light.

The marked bees were transported to the apiary and introduced into the observation hive after dark on the day they had emerged as adults. Beginning the next morning, behavioral observations were conducted at 08:00, 14:00, and 20:00 every day for 28 days. At each observation, both sides of the observation hive were systematically inspected for marked bees for approximately 15 min (Frumhoff and Baker 1988). When a marked bee was found, its location (x, y coordinate based on the 5 cm2 grid) and behavior were recorded. The entrance of the hive was monitored for 15 min at the end of each observation period. Any marked bees seen at the entrance were identified and their behavior recorded. Twenty-eight behaviors were identified over the 84 observation periods. The 28 behaviors were grouped into ten behavioral categories for analysis

(Table 5-1).

The distribution of days for each behavior category was graphed as box plots depicting the mean median, interquartile range, with whiskers depicting 1.5 times the interquartile range. Mixed models procedures as implemented in SAS PROC GLIMMIX

(SAS/STAT 14.1; SAS Institute, Cary, NC) was used to model the response (days).

Treatment (rearing environment), behavior group, and their interaction were considered fixed effects. Colony and the interaction of colony with the previously mentioned two- way interaction were the random effects. Treatment by behavior group interaction means were calculated using the LSMEANS within the above named procedure with the slicediff option to assess simple effect differences for each behavior group.

The maximum day of last observation per bee was calculated and its distribution graphed as a stacked histogram by rearing environment. A simple t-test was used to compare the means of the maximum day of last observation per bee by rearing environment.

Results

Every bee that was introduced was seen at least once and the average number of times a colony-reared (19 ± 1.7 times/bee) or in vitro-reared bee (20.1 ± 1.7) was observed did not differ significantly (F= 0.1898, df= 119, p= 0.66). However, there was a statistically detectable difference in the age at last observation between bees reared by their parental colony or in vitro (t(125)= 3.9, p< 0.001). The mean age at last observation was lower for bees reared in vitro (12.5 ± 0.9) than for those reared in their parental colony (17.5, ± 0.9; Figure 5-1).

Age distributions of bees performing each behavior category are illustrated by rearing environment in Figure 5-2. The number of times bees were observed performing a behavior in each behavior category, the median age at which bees were performing behaviors in each behavior category, and the quartiles of the median are detailed by rearing environment in Table 5-2. Mean comparisons by rearing environment for the age at which bees were observed performing a behavior in a given behavior category are detailed in Table 5-3. There was a statistically detectable reduction in the mean age

at which bees performed each of the following behavioral categories: non-productive

(p= 0.02), grooming (p= 0.01), and head in cell (p= 0.01).

Discussion

In vitro-reared bees were observed engaging in every behavior in which colony- reared bees engaged. I also observed in vitro-reared bees engaging in behaviors that are not explicitly obvious based on my reporting as behavioral categories. These included guarding a small hive beetle that had been confined in a beetle prison (Ellis

2005), attending a waggle dance, performing a waggle dance, and returning to the hive with full pollen loads. These observations highlight that in vitro-reared bees are capable of preforming a myriad of common honey bee behaviors. Additionally, in vitro-reared bees appear to be responding appropriately to colony level cues that coordinate task allocation within age-related polyethism (Seeley 1985; Gordon 2016).

Rearing environment did not noticeably impact the acceptance of experimental bees into the observation hive. Every bee that was introduced into the observation hive was seen again during at least one observation period. However, I did observe a statistically detectable reduction in lifespan of bees that were reared in vitro compared to bees that had been reared by their parental colony. Brodschneider et al. (2009) also reported that the visible number of bees that had been reared in vitro via the Aupinel et al. (2005) method and introduced into an established colony decreased more quickly over time than did bees that had been reared in a hive. Furthermore, Tautz et al. (2003) noted that bees that had been incubated at suboptimal temperatures during their pupal development had shorter lifespans compared to those incubated at optimal temperatures during pupation, and Mattila et al. (2006) reported that bees reared during seasonal times of low pollen stores suffered reduced lifespans. The consistencies in

shorter survival times across bees reared in vitro and bees reared in suboptimal conditions suggest that the physical environment experienced by, and the nutrition provided to bees reared in vitro compared to bees reared in the hive may play a role in the physiology of resulting bees as adults (Tautz et al 2003; Mattila and Otis 2006;

Brodschneider et al 2009).

I observed a statistically significant effect of rearing environment on the mean age at which bees were observed performing behaviors in three behavior categories: non-productive, grooming, and head in cell. However, I observed bees from both rearing environment groups performing behaviors in each of these behavior categories across the entire sampling period (day 1 to day 28). The mean ages for colony-reared bees performing these tasks were higher than those for in vitro-reared bees performing the tasks. I believe this is due to the longer lifespan of bees reared by their parental colony compared to those reared in vitro rather than a direct response to the environment in which they developed.

Cumulatively, the results from the present study and the laboratory brood rearing assay (chapter 4) suggest that there may be a biologically significant change brood rearing behavior based on rearing environment. In my laboratory assay of brood rearing behavior, I noted a statistically detectable decrease in brood rearing behavior of the bees that had been reared in vitro. In the present study, brood rearing behavior is included in the “head in a cell” behavioral category. I also observed a statistically detectable effect of rearing environment (possibly explained by reduced longevity of in vitro-reared bees) on the “head in a cell” behavior category. It is possible that bees reared in vitro have reduced brood rearing abilities, but many in-hive behaviors,

including brood rearing, are difficult to quantify (Tautz et al 2003). It is hard to know what is happening when a bee has her head in a brood cell. Researchers typically use the duration of the visit as an indication of the task that the bee is performing (i.e. longer visits are interpreted as feeding; Shpigler and Robinson 2015). I observed bees reared in both environments with their heads in brood cells. Some of those visits were long enough to be classified as feeding visits per Shpigler and Robinson (2015). However, I do not have the data needed to assess how much, if any, food was provided to larvae or the quality of that food.

The ages at which bees were observed performing behaviors in the remaining behavior categories (washboarding, trophallaxis, attending queen, wax manipulation, ventilation, guarding, and foraging) were not affected by the rearing environment of the bees. However, my data are limited in that I am only able to confirm that the bees I observed were performing a task. I am not able to make a qualitative assessment of the performance of these tasks. For example, in vitro-reared bees did manipulate wax.

However, I cannot say if their execution of this task was correct or appropriate.

Tautz et al. (2003) could not discern any obvious differences in movement within the hive, movement into and out of the hive, or arrival and feeding at experimental feeding stations between bees incubated at a variety of temperatures while pupating.

Furthermore, bees that had been incubated at suboptimal temperatures during pupation also performed waggle dances at the same frequency as those performed by bees incubated at normal temperatures. However, the dances of bees reared suboptimally were not as persistent and the dances contained fewer waggle runs than did those of bees reared optimally. Moreover, Brodschneider et al. (2009) found that in vitro-reared

bees had slightly reduced flight speeds compared to those of colony-reared bees. They suspected that this might affect the foraging efficiency of in vitro-reared bees. I observed bees reared in vitro and in the colony performing waggle dances in the hive. I also consistently observed in vitro-reared and colony-reared bees leaving and returning to the hive. However, I did not determine the frequency or duration of those flights or waggle dances.

While my results add to our understanding of how adult behavior might be influenced by rearing environment, these data do not address questions regarding collective behaviors that emerge at the colony level such as brood and/or honey production, swarming, or nest construction (Gordon 2014; Jandt and Gordon 2016). As superorganisms, it is these colony-level behaviors that truly determine the fitness of a honey bee colony (Moritz and Southwick 1992). Further research, potentially incorporating methods such as establishing colonies composed exclusively of in vitro- reared bees, would offer additional insight into the origins of honey bee behaviors.

Our findings support the use of in vitro rearing as a tool for honey bee health risk assessments and basic examinations of bee behavior and physiology. I did detect differences in the longevity and the execution of broad behavioral categories between bees reared in vitro or in their parental colony. However, statistical differences were not detected in the age at which bees performed more specific behaviors such as guarding, ventilating, and foraging. Consequently, the more specific behaviors would be better focal behaviors for studies where investigators seek to determine the impact of certain stressors on honey bee behaviors. The value of a robust risk assessment is that the results can be used to infer how the agent being tested (e.g. a pesticide) is likely to

impact a colony. Researchers will be able to make more informed inferences regarding colony fitness costs by incorporating colony observations.

Table 5-1. Categories used when recording adult honey bee behavior during observational periods. Data are the code assigned to each category (category code), the category name given to each behavioral grouping (behavior category), and the observed behaviors that were grouped into each category (behaviors observed). Category Behavior Category Behaviors Observed Code 1 non-productive walk, stationary, beard, antennate grooming self, grooming nest mate, being 2 grooming groomed 3 washboard wash boarding engaged in trophallaxis (either as the donor or the 4 trophallaxis recipient) head in an empty cell or a cell containing brood, 5 head in a cell nectar, or pollen 6 attending queen feeding, grooming, or antennating the queen trimming cappings, shaping comb, capping brood, 7 wax manipulation or capping honey 8 ventilation fanning wings at the entrance 9 guarding guarding the hive entrance leaving or returning from the hive, attending or 10 foraging performing a waggle dance

Figure 5-1. Histograms of the age of each honey bee at its last observation, shown by rearing environment (parental colony, or in vitro). The mean maximum age of colony-reared and in vitro-reared bees are noted on both graphs with vertical red (colony-reared) and blue (in vitro-reared) lines. There was a statistically detectable difference in the age at last observation between bees reared by their parental colony or in vitro (t(125)= 3.9, p< 0.001).

Figure 5-2. Age distribution of honey bees performing each task category by rearing environment (colony-reared, or in vitro-reared). Data for each behavior category are the 25th and 75th percentiles (boxes), the median age (vertical line in the box), the mean age (diamond), and whiskers depicting 1.5 times the interquartile range. Exact values of the parameters are detailed in Tables 5-2 and 5-3.

Table 5-2. Age at which honey bees were participating in each behavior category, shown by rearing environment. Data are the behavior category (as defined in Table 5-1), the rearing environment, the number of times the behavior was observed, the median, and quartiles (25th percentile and 75th percentile) age of the bees performing each task.

Behavior Rearing Number of 25th 75th Median Category Environment Observations Percentile Percentile

non-productive colony 394 9 5 16 non-productive in vitro 404 7 4 14 grooming colony 258 8 5 14 grooming in vitro 202 6 3 10 washboard colony 259 6 3 10 washboard in vitro 373 6 3 10 trophallaxis colony 54 10 5 14 trophallaxis in vitro 38 6 3 12 head in a cell colony 50 12 9 18 head in a cell in vitro 98 10 5 14 attending queen colony 4 2 2 6.5 attending queen in vitro 3 9 2 11 wax manipulation colony 17 8 7 10 wax manipulation in vitro 16 7 4.5 9 ventilation colony 4 5.5 3.5 10 ventilation in vitro 1 9 9 9 guarding colony 19 13 12 16 guarding in vitro 5 15 14 23 foraging colony 100 16.5 14 20.5 foraging in vitro 64 19 13.5 22

Table 5-3. Comparisons of the mean age at which each behavior category was performed by honey bees that were reared by their parental colony or in vitro. Data are the behavior category (as defined in Table 5-1), the mean age and SE at which bees were observed performing each task if they were colony- reared or in vitro-reared. A p value ≤ 0.05 notes that the row means are significantly different. Colony-Reared In Vitro-Reared Behavior Category p value Mean ± SE (N) Mean ± SE (N) non-productive 10.3 ± 0.5 (394) 9.0 ± 0.5 (404) 0.02 grooming 9.2 ± 0.5 (258) 7.5 ± 0.6 (202) 0.01 washboard 6.9 ± 0.5 (259) 6.5 ± 0.5 (373) 0.43 trophallaxis 10.1 ± 0.9 (54) 7.7 ± 1.0 (38) 0.06 head in a cell 12.6 ± 0.9 (50) 10 ± 0.7 (98) 0.01 attending queen 4 ± 2.9 (4) 6.8 ± 3.3 (3) 0.52 wax manipulation 8.3 ± 1.4 (17) 6.6 ± 1.5 (16) 0.39 ventilation 6.7 ± 2.9 (4) 9 ± 5.7 (1) 0.72 guarding 14.1 ± 1.4 (19) 17.4 ± 2.6 (5) 0.25 foraging 17.5 ± 0.7 (100) 17.9 ± 0.8 (64) 0.67

CHAPTER 6 CONCLUSION

Cumulatively, the experiments presented in this dissertation demonstrate that honey bees reared in vitro do not suffer gross morphological of behavioral abnormalities compared to their colony-reared counterparts. Many parameters, such as survival to adult emergence, basitarsus length, queen recognition, and foraging age, were unaffected by rearing environment. However, there were statistically detectable effects of rearing environment on other parameters such as survival to the prepupal stage, hind wing length, brood rearing behavior, and longevity in a living colony (Table 6-1).

In vitro rearing eliminates social interactions during the honey bee’s development after transfer to the laboratory as a one-day-old larva. It is possible that juvenile social interactions play a direct roll in the normal expression of adult behavior, as has been documented in other arthropods (Hebets 2003; Strodl and Schausberger 2012).

However, in vitro rearing could also negatively affect the physiology of bees due to suboptimal nutrition and/or physical attributes of the rearing environment (temperature, orientation, etc.). This reduced fitness could secondarily affect adult behavior.

My in vitro-reared bees emerged at an appriate age, lacked any physical deformaties, and performed simialry to colony-reared bees in several lab-based behavioral assays and when introduced into a hive. However, closer examination of the data reveals that the current artificial diet may not be adequate to produce robust adult bees. I observed reduced dry body weight, forewing length, forewing width, and adult longevity in bees reared in vitro compared to those reared by their parental colony.

Reduction of these characteristics have been described previously in the literature in response to suboptimal rearing conditions, such as pollen deficiency and reduced pupal

incubation temperatures (Kunert and Crailsheim 1988; Daly et al 1995; Tautz et al 2003;

Mattila and Otis 2006).

Worker larvae within the colony are progressively provisioned via trophallaxis with small volumes of diet that varies to meet the physiological needs of each life stage

(Schmickl and Crailsheim 2002; Hrassnigg and Crailsheim 2005). Replicating the natural larval diet has been the primary focus of investigators working to optimize the in vitro rearing protocol since it was first developed in 1933 (Crailsheim et al 2013). Larvae that are underfed or fed an unacceptable diet suffer increased mortality rates, but overfeeding can stimulate female larvae to develop into queens or queen/worker intercastes rather than workers (Boot et al 2006; Crailsheim et al 2013; Buttstedt et al

2016).

Pollen limitation in the hive can cause nurse bees to induce early pupation in smaller larve that have yet to reach the typical nutritional state required to induce pupation (Schmickl and Crailsheim 2001). The most current in vitro rearing protocols

(Aupinel et al 2005; Schmehl et al 2016) may be erring on the side of caution by restricting the nutritional intake of the larvae to ensure that workers are produced rather than queens or queen/worker intercastes. I suggest that future iterations of the in vitro rearing protocol explore slightly increased diet volumes or nutrient content in an effort to minimize nuritional stress. A further imporved diet may reduce the descripancies that I observed in the body wieght, forewing size, and longevity of the bees that were reared in vitro compared to those of bees that are reared in a hive. I recommend the last two days of larval development as the feeding points at which one consider modifying the

larval diet given that larve are less likely to develop into queens after the 4th larval day

(Asencot and Lensky 1984).

Throughout the in vitro-rearing process, the larval and pupal plates are maintained horizontally. Brodschneider et al. (2009) suspected that horizontal orientation would result in wing malformation that would prohibit normal flight and suggested that pupal plates should be maintained vertically. I maintained plates vertically in a preliminary trial. However, I found it to be difficult to handle the plates for daily inspections and it was not possible to remove any dead individuals without disturbing the remaining live pupae on the plate. All of the bees reared for the experiments discussed herein were maintained horizontally throughout their larval and pupal development. Immediately upon adult emergence, I did note a slight curve to the end of the wings of some bees (Fig 6-1). However, the curves were not visible after 24 hr. Furthermore, I observed in vitro-reared bees returning from successful foraging flights during my field observations, suggesting that the bend in the wing self-corrects and does not impede normal worker function.

My data highlight some statistically detectable differences in morphology and behavior of adult honey bees that have been reared in vitro. The impact of rearing environment was assessed at the level of the individual bee; yet, the biological unit of the honey bee is the colony, not the individual (Moritz and Southwick 1992). As such, the most biologically relevant parameters are those that emerge at the colony level from the coordination of the colony members. The mechanism(s) by which collective phenotypes materialize in social insects is (are) not well understood (Wilson 1971;

Moritz and Southwick 1992; Gordon 2014; Jandt and Gordon 2016). Therefore, I am

unable to infer to what extent any statistical differences in morphology and/or behavior that were observed in bees that were reared in vitro would affect the overall fitness of a colony.

When assessing honey bee health, it is imperative that researchers be able to infer how any changes in individual bee development may affect the fitness of a colony.

Future studies that examine colony-level parameters such as brood and honey production in colonies composed of workers that were reared in vitro will provide more insight into how the developmental environment affects the productivity of the collective behavior of worker bees. Furthermore, researchers conducting risk assessments can incorporate some of the simple assays outlined within this dissertation to assess potential behavioral impacts of any stressors of interest in future risk assessment.

In vitro rearing promises to continue to be a valuable tool in honey bee health and development research. Evaluation of brood development is difficult within a hive because nurse bees readily abort larvae that have been experimentally manipulated

While in vitro rearing may itself affect honey bee development, it is presently the best available tool researchers can use to assess larval health and development.

Furthermore, continued refinement of the methodology is likely to lead to the production of adult bees that are more similar to bees reared by their parental colony.

Table 6-1. Summary of parameters tested to assess the effects of developmental environment (parental colony or in vitro) on adult worker honey bees. Columnar data are: 1) The chapter in which each experiment is presented (Chap.), 2) the parameter being tested, 3) the dependent variable quantified; 4) indication of whether the mean for in vitro-reared bees was significantly higher (∧) or lower (∨) than that of colony-reared bees (Direction of Difference); and 5) the p-value of the statistical test used to detect statistical significance. Direction of p- Chap. Parameter Tested Measure Quantified Difference value 2 survival survival to day 11 ∧ 0.013 2 survival survival to emergence 0.996

3 morphology forewing length ∨ >0.001 3 morphology forewing width ∨ 0.002 3 morphology hind wing length 0.108

3 morphology basitarsus length 0.716

3 morphology head width 0.175

3 weight dry body weight ∨ 0.003 4 queen recognition bee seconds of queen contact 0.152

4 queen recognition total queen touch count 0.171

4 brood rearing total count of cell visits ∨ <0.001 4 brood rearing total time tending the cell ∨ <0.001 4 brood rearing cell capping 0.829

4 trophallaxis proportion donors fed 0.564

4 trophallaxis proportion of recipients fed 0.262

4 sucrose responsiveness sucrose responsiveness ∨ 0.014 5 age at performance non-productive ∨ 0.02 5 age at performance grooming ∨ 0.01 5 age at performance washboard 0.43

5 age at performance trophallaxis ∨ 0.06 5 age at performance head in a cell ∨ 0.01 5 age at performance attending queen 0.52

5 age at performance wax manipulation 0.39

5 age at performance ventilation 0.72

5 age at performance guarding 0.25

5 age at performance foraging 0.67

5 age at performance age at last observation ∨ <0.001

Figure 6-1. Adult worker honey bee that had been reared in vitro crawling out of her well at a daily inspection. A slight bend can be seen in the forewings and hind wings where they lay along the back of her abdomen. Photo courtesy of Mike Bentley, PhD.

LIST OF REFERENCES

Alaux C, Sinha S, Hasadsri L, Hunt GJ, Guzmán-Novoa E, DeGrandi-Hoffman G, Uribe- Rubio JL, Southey BR, Rodriguez-Zas S, Robinson GE (2009) Honey bee aggression supports a link between gene regulation and behavioral evolution. Proc Natl Acad Sci U S A 106:15400–15405. doi: 10.1073/pnas.0907043106

Alcock J (2009) Animal behavior: An evolutionary approach, 9th edn. Sinauer Associates, Sunderland MA

Allen MD (1960) The honeybee queen and her attendants. Animal Behav 8:201–208. doi: 10.1016/0003-3472(60)90028-2

Allsopp MH, Crewe RM (1993) The Cape honey bee as a Trojan horse rather than the hordes of Genghis Khan. Amer Bee J (USA)

Angilletta MJ, Steury TD, Sears MW (2004) Temperature, growth rate, and body size in ectotherms: fitting pieces of a life-history puzzle. Integr Comp Biol 44:498–509. doi: 10.1093/icb/44.6.498

Asencot M, Lensky Y (1984) Juvenile hormone induction of “queenliness” on female honey bee (Apis mellifera L.) larvae reared on worker jelly and on stored royal jelly. Comp Biochem Phys Part B: Comparative Biochemistry 78:109–117.

Aupinel P, Fortini D, Dufour H, Tasei J, Michaud B, Odoux J, Pham-Delegue M (2005) Improvement of artificial feeding in a standard in vitro method for rearing Apis mellifera larvae. Bull Insectology 58:107–111.

Aupinel P, Fortini D, Michaud B, Medrzycki P, Padovani E, Przygoda D, Maus C, Charriere JD, Kilchenmann V, Riessberger-Galle U, Vollmann J J, Jeker L, Janke M, Odoux JF, Tasei JN (2010) Honey bee brood ring-test: method for testing pesticide toxicity on honeybee brood in laboratory conditions. Julius-Kühn-Archiv 423: 96-102.

Awmack CS, Leather SR (2002) Host plant quality and fecundity in herbivorous insects. Annu Rev Entomol 47:817–844. doi: 10.1146/annurev.ento.47.091201.145300

Bailey Jr DB, Bruer JT, Symons FJ, Lichtman JW (2001) Critical Thinking about Critical Periods. A Series from the National Center for Early Development and Learning. ERIC

Bateson P (1979) How do sensitive periods arise and what are they for? Animal Behaviour 27:470–486.

Becher MA, Scharpenberg H, Moritz RFA (2009) Pupal developmental temperature and behavioral specialization of honeybee workers (Apis mellifera L.). J Comp Physiol A Neuroethol Sens Neural Behav Physiol 195:673–679. doi: 10.1007/s00359-009-0442-7

Beekman M, Fathke RL, Seeley TD (2006) How does an informed minority of scouts guide a honeybee swarm as it flies to its new home? Animal Behaviour 71:161– 171.

Boch R, Rothenbuhler WC (1974) Defensive behaviour and production of alarm pheromone in honeybees. J Apic Res 13:217–221. doi: 10.1080/00218839.1974.11099783

Boot W, Calis J, Allsopp M (2006) Differential feeding of larvae affects caste differentiation in Apis mellifer. Proc Netherlands Entomol Soc 17:63–39.

Breed MD, Robinson GE, Page RE (1990) Division of labor during honey bee colony defense. Behav Ecol Sociobiol 27:395–401. doi: 10.1007/BF00164065

Breed MD, Rogers KB (1991) The behavioral genetics of colony defense in honeybees: genetic variability for guarding behavior. Behav Genet 21:295–303. doi: 10.1007/BF01065821

Brodschneider R, Riessberger-Gallé U, Crailsheim K (2009) Flight performance of artificially reared honeybees (Apis mellifera). Apidologie 40:441–449. doi: 10.1051/apido/2009006

Brouwers EVM (1984) Glucose/fructose ratio in the food of honeybee larvae during caste differentiation. Journal of Apicultural Research 23:94–101.

Büchler R, Andonov S, Bienefeld K, Costa C, Hatjina F, Kezic N, Kryger P, Spivak M, Uzunov A, Wilde J (2013) Standard methods for rearing and selection of Apis mellifera queens. J Apic Res 52:1–30. doi: 10.3896/IBRA.1.52.1.07

Butler CG (1954) The method and importance of the recognition by a colony of honeybees (A. mellifera) of the presence of its queen. Transactions of the Royal Entomological Society of London 105:11–29. doi: 10.1111/j.1365- 2311.1954.tb00773.x

Buttstedt A, Ihling CH, Pietzsch M, Moritz RFA (2016) Royalactin is not a royal making of a queen. Nature 537:E10–E12.

Calderone NW, Page Jr RE (1988) Genotypic variability in age polyethism and task specialization in the honey bee, Apis mellifera (Hymenoptera: Apidae). Behav Ecol Sociobiol (Print) 22:17–25.

Calderone NW, Page RE (1992) Effects of interactions among genotypically diverse nestmates on task specialization by foraging honey bees (Apis mellifera). Behav Ecol Sociobiol 30:219–226. doi: 10.1007/BF00166706

Caldji C, Diorio J, Meaney MJ (2000) Variations in maternal care in infancy regulate the development of stress reactivity. Biol Psych 48 (12): 1164-1174. dio: 10.1016/S0006-3223(00)01084-2.

Cassidy EJ, Bath E, Chenoweth SF, Bonduriansky R (2014) Sex-specific patterns of morphological diversification: evolution of reaction norms and static allometries in neriid flies. Evolution 68:368–383. doi: 10.1111/evo.12276

Chauzat M-P, Faucon J-P, Martel A-C, Lachaize J, Cougoule N, Aubert M (2006) A survey of pesticide residues in pollen loads collected by honey bees in France. J Econ Entomol 99:253–262.

Ciarlo TJ, Mullin CA, Frazier JL, Schmehl DR (2012) Learning impairment in honey bees caused by agricultural spray adjuvants. PLoS ONE 7:e40848. doi: 10.1371/journal.pone.0040848

Connor LJ (2009) Queen Rearing Essentials, 1st edn. Wicwas Press

Craig DPA, Varnon CA, Sokolowski MBC, Wells H, Abramson CI (2014) An assessment of fixed interval timing in free-flying honey bees (Apis mellifera ligustica): an analysis of individual performance. PLoS ONE 9:e101262. doi: 10.1371/journal.pone.0101262

Crailsheim K (1998) Trophallactic interactions in the adult honeybee (Apis mellifera L.). Apidologie 29:97–112. doi: 10.1051/apido:19980106

Crailsheim K, Brodschneider R, Aupinel P, Behrens D, Genersch E, Vollmann J, Riessberger-Gallé U (2013) Standard methods for artificial rearing of Apis mellifera larvae. J Apic Res 52:1–16. doi: 10.3896/IBRA.1.52.1.05

Daly HV, Danka RG, Hoelmer K, Rinderer TE, Buco SM (1995) Honey bee morphometrics: linearity of variables with respect to body size and classification tested with European worker bees reared by varying ratios of nurse bees. J Apic Res 34:129–145. doi: 10.1080/00218839.1995.11100898

Daly HV, Hoelmer K, Norman P, Allen T (1982) Computer-assisted measurement and identification of honey bees (Hymenoptera: Apidae). Ann Entomol Soc Am 75:591–594.

De Souza DA, Wang Y, Kaftanoglu O, De Jong D, Amdam GV, Gonçalves LS, Francoy TM (2015) Morphometric Identification of Queens, Workers and Intermediates in In Vitro Reared Honey Bees (Apis mellifera). PLoS ONE 10:e0123663. doi: 10.1371/journal.pone.0123663

Delaplane KS, van der Steen J, Guzman-Novoa E (2013) Standard methods for estimating strength parameters of Apis mellifera colonies. J Apic Res. doi: 10.3896/IBRA/1.52.1.03

Dietemann V, Lubbe A, Crewe RM (2006) Human factors facilitating the spread of a parasitic honey bee in South Africa. J Econ Entomol 99:7–13.

Einon DF, Morgan MJ (1977) A critical period for social isolation in the rat. Dev Psychobiol 10:123–132.

Ellis JD (2005) Reviewing the confinement of small hive beetles (Aethina tumida ) by western honey bees (Apis mellifera ). Bee World 86:56–62. doi: 10.1080/0005772X.2005.11417312

Ellis JD, Munn PA (2005) The worldwide health status of honey bees. Bee World 86:88– 101.

Feeny P (1970) Seasonal changes in oak leaf tannins and nutrients as a cause of spring feeding by winter moth caterpillars. Ecology 51:565–581. doi: 10.2307/1934037

Fleming JC, Schmehl DR, Ellis JD (2015) Characterizing the Impact of Commercial Pollen Substitute Diets on the Level of Nosema spp. in Honey Bees (Apis mellifera L.). PLoS ONE 10:e0132014. doi: 10.1371/journal.pone.0132014

Free JB (1967) Factors determining the collection of pollen by honeybee foragers. Animal Behav 5:134–144. doi: 10.1016/S0003-3472(67)80024-1

Freedman DG, King JA, Elliot O (1961) Critical period in the social development of dogs. Science 133:1016–1017.

Friard O, Gamba M (2016) BORIS: a free, versatile open-source event-logging software for video/audio coding and live observations. Methods Ecol Evol 7:1325–1330. doi: 10.1111/2041-210X.12584

Frisch K von (1967) The Dance Language and Orientation of Bees. Belknap Press

Frost EH, Shutler D, Hillier NK (2011) Effects of cold immobilization and recovery period on honeybee learning, memory, and responsiveness to sucrose. J Insect Physiol 57:1385–1390. doi: 10.1016/j.jinsphys.2011.07.001

Frumhoff PC, Baker J (1988) A genetic component to division of labour within honey bee colonies. Nature 333:358–361. doi: 10.1038/333358a0

Giurfa M, Sandoz J-C (2012) Invertebrate learning and memory: Fifty years of olfactory conditioning of the proboscis extension response in honeybees. Learn Mem 19:54–66. doi: 10.1101/lm.024711.111

Gordon DM (2014) The ecology of collective behavior. PLoS Biol 12:e1001805. doi: 10.1371/journal.pbio.1001805

Gordon DM (2016) From division of labor to the collective behavior of social insects. Behav Ecol Sociobiol 70:1101–1108. doi: 10.1007/s00265-015-2045-3

Guzmán-Novoa E, Page RE (1993) Backcrossing africanized honey bee queens to european drones reduces colony defensive behavior. Ann Entomol Soc Am 86:352–355. doi: 10.1093/aesa/86.3.352

Hebets EA (2003) Subadult experience influences adult mate choice in an arthropod: exposed female wolf spiders prefer males of a familiar phenotype. Proc Natl Acad Sci U S A 100:13390–13395. doi: 10.1073/pnas.2333262100

Hellmich RL, Rothenbuhler WC (1986) Relationship between different amounts of brood and the collection and use of pollen by the honey bee (Apis mellifera). Apidologie 17:13–20. doi: 10.1051/apido:19860102

Henderson CE (1992) Variability in the size of emerging drones and of drone and worker eggs in honey bee (Apis mellifera L.) colonies. J Apic Res 31:114–118.

Hendriksma HP, Härtel S, Steffan‐Dewenter I (2011) Honey bee risk assessment: new approaches for in vitro larvae rearing and data analyses. Methods Ecol Evol 2:509–517.

Herbert EW, Sylvester HA, Vandenberg JD, Shimanuki H (1988) Influence of nutritional stress and the age of adults on the morphometrics of honey bees (Apis mellifera L.). Apidologie 19:221–230.

Higes M, Martín-Hernández R, Botías C, Bailón EG, González-Porto AV, Barrios L, Del Nozal MJ, Bernal JL, Jiménez JJ, Palencia PG, Meana A (2008) How natural infection by Nosema ceranae causes honeybee colony collapse. Environ Microbiol 10:2659–2669. doi: 10.1111/j.1462-2920.2008.01687.x

Houle D (1992) Comparing evolvability and variability of quantitative traits. Genetics 130:195–204.

Hrassnigg N, Crailsheim K (2005) Differences in drone and worker physiology in honeybees (Apis mellifera). Apidologie 36:255–277.

Huang Z-Y, Robinson GE (1996) Regulation of honey bee division of labor by colony age demography. Behav Ecol Sociobiol 39:147–158.

Huettel MD (1976) Monitoring the quality of laboratory-reared insects: a biological and behavioral perspective. Environ Entomol 5:807–814.

Human H, Brodschneider R, Dietemann V, Dively G, Ellis JD, Forsgren E, Fries I, Hatjina F, Hu F-L, Jaffé R, Jensen AB, Köhler A, Magyar JP, Özkýrým A, Pirk CWW, Rose R, Strauss U, Tanner G, Tarpy DR, van der Steen JJM, Vaudo A, Vejsnæs F, Wilde J, Williams GR, Zheng H-Q (2013) Miscellaneous standard methods for Apis mellifera research. J Apic Res 52:1–53. doi: 10.3896/IBRA.1.52.4.10

Hunt GJ, Amdam GV, Schlipalius D, Emore C, Sardesai N, Williams CE, Rueppell O, Guzmán-Novoa E, Arechavaleta-Velasco M, Chandra S, Fondrk MK, Beye M, Page RE (2007) Behavioral genomics of honeybee foraging and nest defense. Naturwissenschaften 94:247–267. doi: 10.1007/s00114-006-0183-1

Hunt GJ, Guzmán-Novoa E, Uribe-Rubio JL, Prieto-Merlos D (2003) Genotype– environment interactions in honeybee guarding behaviour. Animal Behav 66:459–467. doi: 10.1006/anbe.2003.2253

Ibrahim A, Spivak M (2006) The relationship between hygienic behavior and suppression of mite reproduction as honey bee (Apis mellifera) mechanisms of resistance to Varroa destructor. Apidologie 37:31.

Jandt JM, Gordon DM (2016) The behavioral ecology of variation in social insects. Curr Opin Insect Sci 15:40–44. doi: 10.1016/j.cois.2016.02.012

Jeffree EP, Allen DM (1957) The annual cycle of pollen storage by honey bees. J Econ Entomol 50:211–212. doi: 10.1093/jee/50.2.211

Jones JC, Helliwell P, Beekman M, Maleszka R, Oldroyd BP (2005) The effects of rearing temperature on developmental stability and learning and memory in the honey bee, Apis mellifera. J Comp Physiol A Neuroethol Sens Neural Behav Physiol 191:1121–1129. doi: 10.1007/s00359-005-0035-z

Kaftanoglu O, Linksvayer TA, Page RE (2010) Rearing honey bees (Apis mellifera L.) in vitro: effects of feeding intervals on survival and development. J Apic Res 49:311–317. doi: 10.3896/IBRA.1.49.4.03

Kaftanoglu O, Linksvayer TA, Page RE (2011) Rearing honey bees, Apis mellifera, in vitro 1: effects of sugar concentrations on survival and development. J Insect Sci 11:96. doi: 10.1673/031.011.9601

Kaminski LA, Slessor KN, Winston ML, Hay NW, Borden JH (1990) Honeybee response to queen mandibular pheromone in laboratory bioassays. J Chem Ecol 16:841– 850. doi: 10.1007/BF01016494

Knudsen EI (2004) Sensitive periods in the development of the brain and behavior. J Cogn Neurosci 16:1412–1425. doi: 10.1162/0898929042304796

Kucharski R, Maleszka J, Foret S, Maleszka R (2008) Nutritional control of reproductive status in honeybees via DNA methylation. Science 319:1827–1830. doi: 10.1126/science.1153069

Kunert K, Crailsheim K (1988) Seasonal Changes in Carbohydrate, Lipid and Protein Content in Emerging Worker Honeybees and their Mortality. J Apic Res 27:13– 21. doi: 10.1080/00218839.1988.11100775

Lebuhn G, Droege S, Connor EF, Gemmill-Herren B, Potts SG, Minckley RL, Griswold T, Jean R, Kula E, Roubik DW, Cane J, Wright KW, Frankie G, Parker F (2013) Detecting insect pollinator declines on regional and global scales. Conserv Biol 27:113–120. doi: 10.1111/j.1523-1739.2012.01962.x

Le Conte Y, Ellis M, Ritter W (2010) Varroa mites and honey bee health: can Varroa explain part of the colony losses? Apidologie 41:353–363.

Lynch M, Walsh B (1997) Genetics and analysis of quantitative traits. Sinauer Associates Incorporated, Sunderland, USA

Maestripieri D (1999): The biology of human parenting: Insights from nonhuman primates. Neurosci Biobehav Rev 23:411– 422

Martin SJ, Highfield AC, Brettell L, Villalobos EM, Budge GE, Powell M, Nikaido S, Schroeder DC (2012) Global honey bee viral landscape altered by a parasitic mite. Science 336:1304–1306. doi: 10.1126/science.1220941

Mattila HR, Otis GW (2006) The effects of pollen availability during larval development on the behaviour and physiology of spring-reared honey bee workers. Apidologie 37:533–546. doi: 10.1051/apido:2006037

Medrzycki P, Giffard H, Aupinel P, Belzunces LP, Chauzat M-P, Claßen C, Colin ME, Dupont T, Girolami V, Johnson R, Le Conte Y, Lückmann J, Marzaro M, Pistorius J, Porrini C, Schur A, Sgolastra F, Delso NS, van der Steen JJM, Wallner K, Alaux C, Biron DG, Blot N, Bogo G, Brunet J-L, Delbac F, Diogon M, Alaoui H El, Provost B, Tosi S, Vidau C (2013) Standard methods for toxicology research in Apis mellifera. J Apic Res 52:1–60. doi: 10.3896/IBRA.1.52.4.14

Meixner MD, Pinto MA, Bouga M, Kryger P, Ivanova E, Fuchs S (2013) Standard methods for characterising subspecies and ecotypes of Apis mellifera. J Apic Res 52:1–28. doi: 10.3896/IBRA.1.52.4.05

Menzel R (2012) The honeybee as a model for understanding the basis of cognition. Nat Rev Neurosci 13:758–768. doi: 10.1038/nrn3357 Menzel R, Muller U (1996) Learning and memory in honeybees: from behavior to neural substrates. Annu Rev Neurosci 19:379–404. doi: 10.1146/annurev.ne.19.030196.002115

Metcalfe NB, Monaghan P (2001) Compensation for a bad start: grow now, pay later? Trends Ecol Evol (Amst) 16:254–260. doi: 10.1016/S0169-5347(01)02124-3

Moritz RFA, Hallmen M (1986) Trophallaxis of worker honeybees (Apis mellifera L.) of different ages. Insectes Soc 33:26–31.

Moritz RFA, Southwick EE (1992) Bees as superorganisms: An evolutionary reality. Springer-Verlag, Berlin

Mullin CA, Frazier M, Frazier JL, Ashcraft S, Simonds R, Vanengelsdorp D, Pettis JS (2010) High levels of miticides and agrochemicals in North American apiaries: implications for honey bee health. PLoS ONE 5:e9754. doi: 10.1371/journal.pone.0009754

Nettle D (2006) The evolution of personality variation in humans and other animals. Am Psychol 61:622–631. doi: 10.1037/0003-066X.61.6.622

Neumann P, Carreck NL (2010) Honey bee colony losses. J Apic Res 49:1–6. doi: 10.3896/IBRA.1.49.1.01

Nixon HL, Ribbands CR (1952) Food transmission within the honeybee community. Proc R Soc Lond, B, Biol Sci 140:43–50.

OECD (2015) Honey Bee (Apis mellifera) Larval Toxicity Test, Repeated Exposure. OECD DRAFT GUIDANCE DOCUMENT

OEPP/EPPO (2010) EPPO statndards PP 3/10 (3): Environmental risk assessmentscheme for plant protection products. Bulletin OEPP/EPPO Bulletin 40:323–331.

Page RE, Fondrk MK, Erber J (1998) The effect of genotype on response thresholds to sucrose and foraging behavior of honey bees (Apis mellifera L.). Journal of Comparative Physiology A: Sensory, Neural, and Behavioral Physiology 182:489–500. doi: 10.1007/s003590050196

Page RE, Peng CY (2001) Aging and development in social insects with emphasis on the honey bee, Apis mellifera L. Exp Gerontol 36:695–711. doi: 10.1016/S0531- 5565(00)00236-9

Page RE, Scheiner R, Erber J, Amdam GV (2006) The Development and Evolution of Division of Labor and Foraging Specialization in a Social Insect (Apis mellifera L.). Current topics in developmental biology volume 74. Elsevier, pp 253–286 Pankiw T, Page RE (2003) Effect of pheromones, hormones, and handling on sucrose response thresholds of honey bees (Apis mellifera L.). J Comp Physiol A Neuroethol Sens Neural Behav Physiol 189:675–684. doi: 10.1007/s00359-003- 0442-y

Partridge L, Barrie B, Fowler K, French V (1994) Evolution and development of body size and cell size in drosophila melanogaster in response to temperature. Evolution 48:1269–1276. doi: 10.1111/j.1558-5646.1994.tb05311.x

Peng Y-SC, Jay SC (1977) Larval rearing by worker honey bees lacking their mandibular glands: I. Rearing by small numbers of worker bees. Can Entomol 109:1175–1180.

Peng Y-SC, Mussen E, Fong A, Montague MA, Tyler T (1992) Effects of chlortetracycline of honey bee worker larvae reared in vitro. J Invertebr Pathol 60:127–133.

Rembold H, Czoppelt C, Rao PJ (1974) Effect of juvenile hormone treatment on caste differentiation in the honeybee, Apis mellifera. J Insect Physiol 20:1193–1202.

Rembold H, Lackner B (1981) Rearing of honeybee larvae in vitro - effect of yeast extract on queen differentiation. J Apic Res 20:165–171.

Robinson GE (1992) Regulation of division of labor in insect societies. Annu Rev Entomol 37:637–665. doi: 10.1146/annurev.en.37.010192.003225

Robinson GE, Evans JD, Maleszka R, Robertson HM, Weaver DB, Worley K, Gibbs RA, Weinstock GM (2006) Sweetness and light: illuminating the honey bee genome. Insect Mol Biol 15:535–539. doi: 10.1111/j.1365-2583.2006.00698.x

Robinson GE, Page RE (1989) Genetic determination of nectar foraging, pollen foraging, and nest-site scouting in honey bee colonies. Behav Ecol Sociobiol 24:317–323. doi: 10.1007/BF00290908

Robinson GE, Page RE, Strambi C, Strambi A (1989) Hormonal and genetic control of behavioral integration in honey bee colonies. Science 246:109–112. doi: 10.1126/science.246.4926.109

Rueppell O, Chandra SBC, Pankiw T, Fondrk MK, Beye M, Hunt G, Page RE (2006) The genetic architecture of sucrose responsiveness in the honeybee (Apis mellifera L.). Genetics 172:243–251. doi: 10.1534/genetics.105.046490

Ruttner F (1988) Biogeography and taxonomy of honeybees. Springer-Verlag

Scheiner R, Abramson CI, Brodschneider R, Crailsheim K, Farina WM, Fuchs S, Grünewald B, Hahshold S, Karrer M, Koeniger G, Koeniger N, Menzel R, Mujagic S, Radspieler G, Schmickl T, Schneider C, Siegel AJ, Szopek M, Thenius R (2013) Standard methods for behavioural studies of Apis mellifera. J Apic Res 52:1–58. doi: 10.3896/IBRA.1.52.4.04

Schmehl DR, Tomé HVV, Mortensen AN, Martins GF, Ellis JD (2016) Protocol for the in vitro rearing of honey bee (Apis mellifera L.) workers. J Apic Res 55:113–129. doi: 10.1080/00218839.2016.1203530

Schmickl T, Crailsheim K (2001) Cannibalism and early capping: strategy of honeybee colonies in times of experimental pollen shortages. J Comp Physiol A Neuroethol Sens Neural Behav Physiol 187:541–547. doi: 10.1007/s003590100226

Schmickl T, Crailsheim K (2002) How honeybees (Apis mellifera L.) change their broodcare behaviour in response to non-foraging conditions and poor pollen conditions. Behav Ecol Sociobiol (Print) 51:415–425.

Schmickl T, Crailsheim K (2004) Inner nest homeostasis in a changing environment with special emphasis on honey bee brood nursing and pollen supply. Apidologie 35:249–263.

Schmickl T, Thenius R, Moeslinger C, Radspieler G, Kernbach S, Szymanski M, Crailsheim K (2009) Get in touch: cooperative decision making based on robot- to-robot collisions. Auton Agent Multi Agent Syst 18:133–155. doi: 10.1007/s10458-008-9058-5

Schmid-Hempel P, Kacelnik A, Houston AI (1985) Honeybees maximize efficiency by not filling their crop. Behav Ecol Sociobiol 17:61–66. doi: 10.1007/BF00299430

Schneider SS, Leamy LJ, Lewis LA, DeGrandi-Hoffman G (2003) The influence of hybridization between African and European honeybees, Apis mellifera, on asymmetries in wing size and shape. Evolution 57:2350–2364.

Schulz DJ, Vermiglio MJ, Huang ZY, Robinson GE (2002) Effects of colony food shortage on social interactions in honey bee colonies. Insectes Soc 49:50–55. doi: 10.1007/s00040-002-8279-x

Scofield HN, Mattila HR (2015) Honey bee workers that are pollen stressed as larvae become poor foragers and waggle dancers as adults. PLoS ONE 10:e0121731. doi: 10.1371/journal.pone.0121731

Seely TD (1979) Queen substance dispersal by messenger workers in honeybee colonies. Behav Ecol Sociobiol 5:391–415. doi: 10.1007/BF00292527

Seeley TD (1982) Adaptive significance of the age polyethism schedule in honeybee colonies. Behav Ecol Sociobiol (Print) 11:287–293.

Seeley TD (1985) Honeybee ecology: a study of adaptation in social life. Princeton University Press, Princeton, N.J

Seeley TD (1996) The Wisdom of the Hive: The Social Physiology of Honey Bee Colonies. Harvard University Press

Seeley TD, Visscher PK, Passino KM (2006) Group decision making in honey bee swarms. Am SciAm Scientist 94:220.

Sheppard WS (1989) A history of the introduction of honey bee races into the United Sates: Part II. Amer Bee J 129:664–667.

Shi YY, Huang ZY, Zeng ZJ, Wang ZL, Wu XB, Yan WY (2011) Diet and cell size both affect queen-worker differentiation through DNA methylation in honey bees (Apis mellifera, Apidae). PLoS ONE 6:e18808. doi: 10.1371/journal.pone.0018808

Shpigler HY, Robinson GE (2015) Laboratory assay of brood care for quantitative analyses of individual differences in honey bee (Apis mellifera) affiliative behavior. PLoS ONE 10:e0143183. doi: 10.1371/journal.pone.0143183

Shuel RW, Dixon SE (1986) An artificial diet for laboratory rearing of honeybees. J Apic Res 25:35–43. doi: 10.1080/00218839.1986.11100689

Silva IC, Message D, Cruz CD, Campos LAO, Sousa-Majer MJ (2009) Rearing Africanized honey bee (Apis mellifera L.) brood under laboratory conditions. Genet Mol Res 8:623–629.

Spivak M, Gilliam M (1998) Hygienic behaviour of honey bees and its application for control of brood diseases and Varroa. Bee World 79:124–134. doi: 10.1080/0005772X.1998.11099394

Steinhauer NA, Rennich K, Wilson ME, Caron DM, Lengerich EJ, Pettis JS, Rose R, Skinner JA, Tarpy DR, Wilkes JT, vanEngelsdorp D, for the Bee Informed Partnership (2014) A national survey of managed honey bee 2012–2013 annual colony losses in the USA: results from the Bee Informed Partnership. J Apic Res 53:1–18. doi: 10.3896/IBRA.1.53.1.01

100

Strodl MA, Schausberger P (2012) Social familiarity modulates group living and foraging behaviour of juvenile predatory mites. Naturwissenschaften 99:303–311.

Tautz J, Maier S, Groh C, Rossler W, Brockmann A (2003) Behavioral performance in adult honey bees is influenced by the temperature experienced during their pupal development. Proc Natl Acad Sci U S A 100:7343–7347. doi: 10.1073/pnas.1232346100

Vandenberg JD, Shimanuki H (1987) Technique for rearing worker honeybees in the laboratory. J Apic Res 26:90–97. doi: 10.1080/00218839.1987.11100743 vanEngelsdorp D, Speybroeck N, Evans JD, Nguyen BK, Mullin C, Frazier M, Frazier J, Cox-Foster D, Chen Y, Tarpy DR, Haubruge E, Pettis JS, Saegerman C (2010) Weighing risk factors associated with bee colony collapse disorder by classification and regression tree analysis. J Econ Entomol 103:1517–1523. doi: 10.1603/EC09429

Velthuis HHW, van Es J (1964) Some functional aspects of the mandibular glands of the queen honeybee. J Apic Res 3:11–16. doi: 10.1080/00218839.1964.11100076

Wang Y, Ma L, Zhang W, Cui X, Wang H, Xu B (2016) Comparison of the nutrient composition of royal jelly and worker jelly of honey bees (Apis mellifera). Apidologie 47:48–56. doi: 10.1007/s13592-015-0374-x

Weaver N (1957) Effects of larval age on dimorphic differentiation of the female honey bee. Ann Entomol Soc Am 50:283–294.

Weaver N (1966) Physiology of caste determination. Annu Rev Entomol 11:79–102. doi: 10.1146/annurev.en.11.010166.000455

Williams GR, Alaux C, Costa C, Csáki T, Doublet V, Eisenhardt D, Fries I, Kuhn R, McMahon DP, Medrzycki P, Murray TE, Natsopoulou ME, Neumann P, Oliver R, Paxton RJ, Pernal SF, Shutler D, Tanner G, van der Steen JJM, Brodschneider R (2013) Standard methods for maintaining adult Apis mellifera in cages under in vitro laboratory conditions. J Apic Res 52:1–36. doi: 10.3896/IBRA.1.52.1.04

Wilson EO (1971) The Insect Societies. The Belknap Press of Harvard University Press, Massachusetts, USA

Winston ML (1991) The Biology of the Honey Bee. Harvard University Press

Winston ML, Katz SJ (1981) Longevity of cross-fostered honey bee workers (Apis mellifera ) of European and Africanized races. Can J Zool 59:1571–1575. doi: 10.1139/z81-214

101

Winston ML, Slessor KN (1992) The essence of royalty: honey bee queen pheromone. Am Sci 80:374–385.

Winston ML, Slessor KN, Willis LG, Naumann K, Higo HA, Wyborn MH, Kaminski LA (1989) The influence of queen mandibular pheromones on worker attraction to swarm clusters and inhibition of queen rearing in the honey bee (Apis mellifera L.). Insectes Soc 36:15–27. doi: 10.1007/BF02225877

Wittmann D, Engels W (1981) Development of test procedures for insecticide-induced brood damage in honeybees. Mitteilungen der Deutschen Gesellschaft fur Allgemeine und Angewandte Entomologie 3:187–190.

Woyke J (1984) Correlations and interactions between population, length of worker life and honey production by honeybees in a temperate region. J Apic Res 23:148– 156. doi: 10.1080/00218839.1984.11100624

Wray MK, Mattila HR, Seeley TD (2011) Collective personalities in honeybee colonies are linked to colony fitness. Animal Behav 81:559–568. doi: 10.1016/j.anbehav.2010.11.027

102

BIOGRAPHICAL SKETCH

Ashley Mortensen is a University of Florida (UF), doctoral candidate in the

Entomology & Nematology Department. She received a B.S. in Animal Science from

Texas A&M University in 2005 and a M.S. in Entomology from UF in 2013. Ashley’s current research aims to determine if key honey bee behaviors are instinctual or learned by examining the role that the juvenile rearing environment plays on adult behavior. In vitro rearing of honey bee brood is an emerging risk assessment tool that has been implemented in compound safety screening requirements for the OECD. Ashley’s research will better inform how investigators interpret the findings of these risk assessments. Ashley is an active contributor to UF’s teaching and Extension programs.

She has developed and taught a beekeeping field techniques course; redesigned, taught, and assisted the department’s online beekeeping course; led introduction to entomology laboratory sections; and has provided numerous guest lectures and field presentations for departmental courses. To share relevant research findings and beekeeping knowledge with a broader audience, Ashley co-created a social medial strategy via Twitter, Instagram, and Facebook (@UFHoneyBeeLab) through which she shares research updates and other honey bee related information to over 7,500 followers across the three sites. Ashley also produces blog and newsletter posts,

Extension reports, web-based presentations, in-person presentations, and workshops.

103