Learning and Task Interference by Corpse-removal Specialists in Honey Bee Colonies

Undertakers are considered to be among the most specialized of pre-foraging honey bee (Apis mellifera L.) workers. In this study we examined a possible benefit and a cost of the corpse-rem oval specialty, the improvement in performance with experience, and interference by individuals attempting to perform the same task in the same location, respectively. Experienced bees removed corpses significantly faster than less experienced bees and also were less likely to drop corpses while exiting the hive (5.5% vs. 14.3% of attempts). Superior performance by experienced undertakers might occur as a consequence of learning, or by greater ability from the outset. Because active undertakers (≥ 3 corpse removals) did not improve with experience over their own careers, learning was not demonstrated. An extreme specialist, Yellow 54, removed a total of 114 corpses (33.8% of experimentally introduced dead bees) from the hive over a 13-day period. This is the longest recorded tenure of undertaking to date and demonstrates how a few individuals can dominate this task in a honey bee colony-Yellow 54 removed corpses significantly faster than other active bees, but she demonstrated no obvious improvement in performance over her undertaking career. This suggests the possibility that active undertakers were more talented than less active undertakers, irrespective of learning. When two undertakers worked together to remove a corpse from the hive, they took longer to complete the task than did single individuals. When multiple undertakers flew together from the hive, they were less likely to clear a nearby obstruction than single undertakers and were more likely to drop the corpse within 1 m of the hive. Thus, mutual interference exacted a measurable cost as a result of the undertaking specialization while learning provided few benefits.     

Spatial effects, sampling errors, and task specialization in the honey bee

Task allocation patterns should depend on the spatial distribution of work within the nest, variation in task demand, and the movement patterns of workers, however, relatively little research has focused on these topics. This study uses a spatially explicit agent based model to determine whether such factors alone can generate biases in task performance at the individual level in the honey bees, Apis mellifera. Specialization (bias in task performance) is shown to result from strong sampling error due to localized task demand, relatively slow moving workers relative to nest size, and strong spatial variation in task demand. To date, specialization has been primarily interpreted with the response threshold concept, which is focused on intrinsic (typically genotypic) differences between workers. Response threshold variation and sampling error due to spatial effects are not mutually exclusive, however, and this study suggests that both contribute to patterns of task bias at the individual level. While spatial effects are strong enough to explain some documented cases of specialization; they are relatively short term and not explanatory for long term cases of specialization. In general, this study suggests that the spatial layout of tasks and fluctuations in their demand must be explicitly controlled for in studies focused on identifying genotypic specialists.     

Antennal olfactory sensitivity in response to task-related odours of three castes of the ant Atta mexicana (hymenoptera: formicidae)

Abstract.  The relationship between scent composition and antennal sensitivity in different castes of Atta mexicana is investigated under laboratory conditions. Extracts of dead ants are analysed by gas chromatography-mass spectrometry to identify the compounds presumably responsible for the specific undertaking behaviour. Oleic acid is identified as one compound that triggers undertaking behaviour. To determine differences in odour reception between workers of different castes (i.e. foragers, undertakers and soldiers), further antennal sensitivity to task-related odours is tested using electroantennographic techniques. Soldiers are the most specialized caste because of their low response to all odours, except odourants related to alarm pheromones. The behavioural specialization of soldiers and their reduced behavioural repertoire are discussed.     

Physiological correlates of division of labor among similarly aged honey bees

Hormone analyses and exocrine gland measurements were made to probe for physiological correlates of division of labor among similarly aged adult worker honey bees (Apis mellifera L.). Middle-age bees (ca. 2 weeks old) performing different tasks showed significant differences in both juvenile hormone (JH) biosynthesis rates and hemolymph titers; guards and undertakers had high JH, and wax producers and food storers, low JH. Guards and undertakers had similar hormone levels to foragers, even though they were 10 days younger than foragers. No differences in JH were detected among young bees (1-week-old queen attendants and nurses) or older bees (3–4 week-old pollen foragers, non-pollen foragers, and soldiers). Hypopharyngeal gland size was inversely correlated with worker age and rate of JH biosynthesis, but soldiers had significantly larger hypopharyngeal glands than did foragers, despite their similar age and JH level. Results from soldiers indicate that exocrine gland development is not always linked with age-related behavior and endocrine development; they also support the recent claim that soldiers constitute a group of older bees that are distinct from foragers. Hormonal analyses indicate that the current model of JH's role in honey bee division of labor needs to be expanded because high levels of JH are associated with several other tasks besides foraging. JH may be involved in the regulation of division of labor among similarly aged workers in addition to its role in age-related division of labor.Abbreviations JH Juvenile hormone - RIA radioimmunoassay - CA corpora allata - HPLC high performance liquid chromatography - TLC thin layer chromatography     

The guard honey bee: ontogeny and behavioural variability of workers performing a specialized task

Guarding is a relatively unstudied aspect of honey bee, Apis mellifera L., worker behaviour. The aim of this study was to characterize quantitatively the ontogeny and individual variability of guarding behaviour, the allocation of workers to the guard population in a colony, and the intercolonial variability of guarding behaviour. Guarding is a discrete task performed by a distinct group of workers that are younger than foragers and older than house bees. Workers that guarded initiated the behaviour between the ages of 7 and 22 days. The mean age of the onset of guarding varied; the minimum mean age of guards for a colony was 13·6 days and the maximum was 16·0 days. Workers varied in the length of time they spent as a guard. Most bees guarded for less than 1 days; however, some guarded up to 6 consecutive days. The more time a bee spent guarding during a day the more likely that bee was to guard for more than 1 day. Bees that guarded for more than 1 day also had longer and more frequent individual guarding bouts. All colonies that were studied had guard populations, but not all workers guarded. A relatively small proportion of any age cohort was observed to guard. The percentage of an age cohort that guarded varied among colonies, as did the size of the guard population. Guarding is a specialized task in that few bees guard, but guarding does not appear to require experience because so few bees remained as guards for very long. There was intercolonial variation in all aspects of the ontogeny of guarding and in allocation of workers to guarding. This variation is discussed in the light of other studies of variation in worker behaviour.     

Machine learning approach for automatic recognition of tomato-pollinating bees based on their buzzing-sounds

Bee-mediated pollination greatly increases the size and weight of tomato fruits. Therefore, distinguishing between the local set of bees–those that are efficient pollinators–is essential to improve the economic returns for farmers. To achieve this, it is important to know the identity of the visiting bees. Nevertheless, the traditional taxonomic identification of bees is not an easy task, requiring the participation of experts and the use of specialized equipment. Due to these limitations, the development and implementation of new technologies for the automatic recognition of bees become relevant. Hence, we aim to verify the capacity of Machine Learning (ML) algorithms in recognizing the taxonomic identity of visiting bees to tomato flowers based on the characteristics of their buzzing sounds. We compared the performance of the ML algorithms combined with the Mel Frequency Cepstral Coefficients (MFCC) and with classifications based solely on the fundamental frequency, leading to a direct comparison between the two approaches. In fact, some classifiers powered by the MFCC–especially the SVM–achieved better performance compared to the randomized and sound frequency-based trials. Moreover, the buzzing sounds produced during sonication were more relevant for the taxonomic recognition of bee species than analysis based on flight sounds alone. On the other hand, the ML classifiers performed better in recognizing bees genera based on flight sounds. Despite that, the maximum accuracy obtained here (73.39% by SVM) is still low compared to ML standards. Further studies analyzing larger recording samples, and applying unsupervised learning systems may yield better classification performance. Therefore, ML techniques could be used to automate the taxonomic recognition of flower-visiting bees of the cultivated tomato and other buzz-pollinated crops. This would be an interesting option for farmers and other professionals who have no experience in bee taxonomy but are interested in improving crop yields by increasing pollination.     

Ethology of Hygienic Behaviour in the Honey Bee Apis mellifera L. (Hymenoptera: Apidae): Behavioural repertoire of Hygienic bees

Hygienic behaviour performed by middle‐aged worker bees is an important intranidal task in colonies of the honey bee Apis mellifera (L.). It comprises detecting diseased brood in the larval and pupal stages and removing all such infected brood, thereby decreasing the incidence of infection. Hygienic behaviour consists of two task‐components: uncapping cells and removing the cell contents. The aim of this study was to observe bees performing hygienic behaviour to determine their age at performance of the behaviour and to describe their behavioural repertoire. The bees performing hygienic behaviour were middle‐aged bees, younger than foragers. In the colonies where the behaviours of individual bees were observed, all bees performing the hygienic behaviour were seen to exhibit both the components, though at different frequencies. One behavioural class performed the task of uncapping cells at higher frequencies than the task of removing cell contents, while another class performed both tasks to the same extent. While these two classes had higher frequencies of the tasks comprising the hygienic behaviour but lower frequencies of other common behaviours in their repertoire, a third class of bees included those that performed all behaviours in their repertoire at similar frequencies. There was no difference in the ages of the bees in these three behavioural classes. These results suggest that there is no evidence of task partitioning among bees performing the hygienic behaviour. The segregation observed could, however, be based on their response thresholds to the stimulus and/or on their ability to discriminate the various cues emanating from the dead brood.     

Pupal developmental temperature and behavioral specialization of honeybee workers (<Emphasis Type="Italic">Apis mellifera</Emphasis> L.)

Honeybees (Apis mellifera) are able to regulate the brood nest temperatures within a narrow range between 32 and 36°C. Yet this small variation in brood temperature is sufficient to cause significant differences in the behavior of adult bees. To study the consequences of variation in pupal developmental temperature we raised honeybee brood under controlled temperature conditions (32, 34.5, 36°C) and individually marked more than 4,400 bees, after emergence. We analyzed dancing, undertaking behavior, the age of first foraging flight, and forager task specialization of these workers. Animals raised under higher temperatures showed an increased probability to dance, foraged earlier in life, and were more often engaged in undertaking. Since the temperature profile in the brood nest may be an emergent property of the whole colony, we discuss how pupal developmental temperature can affect the overall organization of division of labor among the individuals in a self-organized process.     

Reproductive biology of Pachira aquatica Aubl. (Malvaceae: Bombacoideae): a tropical tree pollinated by bats,sphingid moths and honey bees

We investigated the reproductive biology, including the floral biology, pollination biology, breeding system and reproductive success, of Pachira aquatica, a native and dominant tropical tree of fresh water wetlands, throughout the coastal plain of the Gulf of Mexico. The flowers present nocturnal anthesis, copious nectar production and sugar concentration (range 18–23%) suitable for nocturnal visitors such as bats and sphingid moths. The main nocturnal visitors were bats and sphingid moths while bees were the main diurnal visitors. There were no differences in legitimate visitation rates among bats, moths and honey bees. Bats and honey bees fed mainly on pollen while moths fed on nectar, suggesting resource partitioning. Eight species of bats carried pollen but Leptonycteris yerbabuenae is probably the most effective pollinator due to its higher pollen loads. The sphingid moths Manduca rustica, Cocytius duponchel and Eumorpha satellitia were recorded visiting flowers. Hand pollination experiments indicated a predominant outcrossing breeding system. Open pollination experiments resulted in a null fruit set, indicating pollen limitation; however, mean reproductive success, according to a seasonal census, was 17 ± 3%; these contrasting results could be explained by the seasonal availability of pollinators. We conclude that P. aquatica is an outcrossing species with a pollination system originally specialized for bats and sphingid moths, which could be driven to a multimodal pollination system due to the introduction of honey bees to tropical America.     

Corpse Management in Social Insects

Undertaking behavior is an essential adaptation to social life that is critical for colony hygiene in enclosed nests. Social insects dispose of dead individuals in various fashions to prevent further contact between corpses and living members in a colony. Focusing on three groups of eusocial insects (bees, ants, and termites) in two phylogenetically distant orders (Hymenoptera and Isoptera), we review mechanisms of death recognition, convergent and divergent behavioral responses toward dead individuals, and undertaking task allocation from the perspective of division of labor. Distinctly different solutions (e.g., corpse removal, burial and cannibalism) have evolved, independently, in the holometabolous hymenopterans and hemimetabolous isopterans toward the same problem of corpse management. In addition, issues which can lead to a better understanding of the roles that undertaking behavior has played in the evolution of eusociality are discussed.     

Visual associative learning in restrained honey bees with intact antennae

A restrained honey bee can be trained to extend its proboscis in response to the pairing of an odor with a sucrose reward, a form of olfactory associative learning referred to as the proboscis extension response (PER). Although the ability of flying honey bees to respond to visual cues is well-established, associative visual learning in restrained honey bees has been challenging to demonstrate. Those few groups that have documented vision-based PER have reported that removing the antennae prior to training is a prerequisite for learning. Here we report, for a simple visual learning task, the first successful performance by restrained honey bees with intact antennae. Honey bee foragers were trained on a differential visual association task by pairing the presentation of a blue light with a sucrose reward and leaving the presentation of a green light unrewarded. A negative correlation was found between age of foragers and their performance in the visual PER task. Using the adaptations to the traditional PER task outlined here, future studies can exploit pharmacological and physiological techniques to explore the neural circuit basis of visual learning in the honey bee.     

Vibration Signal Modulates the Behavior of House-hunting Honey Bees (Apis mellifera)

Honey bees ( Apis mellifera ) in house-hunting swarms perform vibration signals (dorsoventral abdominal vibration (DVAV)) of 18.05 ± 0.45 Hz for 1.36 ± 0.23 s throughout the house selection process. These signals are performed by a specialized subset of bees, most of whom never perform recruitment dances to nest sites. Individuals repeatedly vibrate others. The patterns of vibration signal performance are consistent with the hypothesis that it serves to activate bees for take-off, but may also activate bees to scout for nest sites.     

Floral bagging differentially affects handling behaviours and single-visit pollen deposition by honey bees and native bees

1. Measurements of pollinator performance are crucial to pollination studies, enabling researchers to quantify the relative value of different pollinator species to plant reproduction. One of the most widely employed measures of pollinator performance is single-visit pollen deposition, the number of conspecific pollen grains deposited to a stigma after one pollinator visit. To ensure a pollen-free stigma, experimenters must first bag flowers before exposing them to a pollinator. 2. Bagging flowers, however, may unintentionally manipulate floral characteristics to which pollinators respond. In this study, we quantified the effect of bagging on nectar volume in watermelon (Citrullus lanatus) flowers, and how this affects pollinator performance and behaviour. 3. Experimental bagging resulted in roughly 30-fold increases in nectar volume relative to unmanipulated, open-pollinated field flowers after only a few hours. Honey bees, but not native bees, consistently displayed elevated handling times and single-visit pollen deposition on unmanipulated bagged flowers relative to those from which we removed nectar to mimic volumes in open-pollinated flowers. 4. Furthermore, we identify specific bee foraging behaviours during a floral visit that account for differences in pollen deposition, and how these differ between honey bees and native bees. 5. Our findings suggest that experimental bagging of flowers, without accounting for artificially accumulated nectar, can lead to biased estimates of pollinator performance in pollinator taxa that respond strongly to nectar volume. We advise that pollination studies be attentive to nectar secretion dynamics in their focal plant species to ensure unbiased estimates of pollinator performance across multiple pollinator species.     

Emergence and Consequences of Division of Labor in Associations of Normally Solitary Sweat Bees

Division of labor is a pervasive feature of animal societies, but little is known about the causes or consequences of division of labor in non-eusocial cooperative groups. We tested whether division of labor self-organizes in an incipient social system: artificially induced nesting associations of the normally solitary sweat bee Lasioglossum ( Ctenonomia ) NDA-1 (Hymenoptera: Halictidae). We quantified task performance and construction output by females nesting either alone or with a conspecific. Within pairs, a division of labor repeatedly arose in which one individual specialized on excavation and pushing/tamping while her nestmate guarded the nest entrance. Task specialization could not be attributed to variation in overall activity, and the degree of behavioral differentiation was greater than would be expected due to random variation, indicating that division of labor was an emergent phenomenon generated in part by social dynamics. Excavation specialists did not incur a survival cost, in contrast to previous findings for ant foundress associations. Paired individuals performed more per capita guarding, and pairs collectively excavated deeper nests than single bees – potential early advantages of social nesting in halictine bees.     

Influence of brood, vent screening, and time of year on honey bee (Hymenoptera: Apidae) pollination and fruit quality of greenhouse tomatoes

Greenhouse tomatoes, Lycopersicon esculentum Miller (Solanaceae), are autogamous, but facilitated pollination results in increased fruit size and set. Previous research examining honey bee pollination in greenhouse tomato crops established that fruit quality resulting from honey bee visitation is often comparable to bumble bees (Bombus spp.) and significantly better than in flowers that receive no facilitated pollination. However, management alternatives have not been studied to improve tomato fruit quality when honey bees are the only pollination option available for the high-value greenhouse industry. We investigated whether the quantity of brood (eggs, larvae, and pupae) in a honey bee colony in the winter and screening on greenhouse vents in the summer would encourage honey bee foraging on tomato flowers. We also established the influence of time of year on the potential for honey bees to be effective pollinating agents. We constructed small honey bee colonies full of naive forager bees with either two frames of brood ("brood colonies") or two empty frames ("no-brood") and compared total fruit set and the number of tomato seeds resulting from fruit potentially visited by honey bees in each of these treatments to bagged flowers that received no facilitated pollination. There was no significant difference in the quality of fruit resulting from honey bees from "brood" and "no-brood" colonies. However, these fruits produced significantly more seeds than bagged flowers restricted from facilitated pollination. Honey bees from brood and no-brood colonies also resulted in 98% fruit set compared with 80% fruit set in bagged flowers that received no facilitated pollination. During the summer, the number of seeds per fruit did not differ significantly between unbagged flowers potentially visited by honey bees in screened greenhouses and unscreened greenhouses and bagged flowers that received no facilitated pollination. However, time of year did have a significant influence on the quality of fruit produced by honey bees compared with flowers that received no facilitated pollination, because no difference in seed number was observed between the treatments after mid-April. The results from this study demonstrate that the management of brood levels and vent screening cannot be used to improve the quality of fruit resulting from honey bee pollination and that honey bees can be a feasible greenhouse pollination alternative only during the winter.     

Age and Behavior of Honey Bees,Apis mellifera (Hymenoptera: Apidae), that Perform Vibration Signals on Queens and Queen Cells

The vibration signal of the honey bee (Apis mellifera) may play a central role in the regulation of queen behavior during reproductive swarming and supersedure. We examined honey bee workers that performed vibration signals on queens and developing queen cells in three observation hives, each containing a population of marked bees of known age. In all three colonies, workers of all ages greater than 2 d old could perform vibration signals on queens and queen cells. However, most signals were performed by a small proportion of the bees of greater than 10 d of age. Relatively few workers less than 10 d old vibrated queens and queen cells, even though this age-group is typically associated with queen care. Thus, the regulation of queen behavior by the vibration signal may occur primarily through a relatively small subset of older workers that, under most circumstances, have only limited involvement with queens. It is unclear what triggers the vibrating of queens. Workers producing vibration signals did not differ from same-age non-vibrating controls in rate of locomotion in the hive or in task performance, and they rarely engaged in foraging, even though the majority of observed bees were of foraging age; vibrators also did not spend more time with queens and queen cells compared with controls. Vibration signals performed on queens and queen cells therefore do not appear to be influenced by task performance or increased contact with queens.     

Co-dependency between a specialist <Emphasis Type="Italic">Andrena</Emphasis> bee and its death camas host, <Emphasis Type="Italic">Toxicoscordion paniculatum</Emphasis>

Among associations of plants and their pollinating bees, mutually specialized pairings are rare. Typically, either pollen specialist (oligolectic) bees are joined by polylectic bees in a flowering species’ pollinator guild, or specialized flowers are pollinated by one or more polylectic bees. The bee Andrena astragali is a narrow oligolege, collecting pollen solely from two nearly identical species of death camas (Toxicoscordion, formerly Zigadenus). Neurotoxic alkaloids of these plants are implicated in sheep and honey bee poisoning. In this study, T. paniculatum, T. venenosum and co-flowering forbs were sampled for bees at 15 sites along a 900-km-long east–west transect across the northern Great Basin plus an altitudinal gradient in northern Utah’s Bear River Range. Only A. astragali bees were regularly seen visiting flowering panicles of these Toxicoscordion. In turn, this bee was never among the 170 bee species caught at 17 species of other prevalent co-occurring wildflowers in the same five state region (38,000 plants surveyed). Our field pollination experiments show that T. paniculatum is primarily an outcrosser dependent on pollinator visitation for most capsule and seed set. Thus, both A. astragali and two sister species of Toxicoscordion are narrowly specialized and co-dependent on each other for reproduction, illustrating a rare case of obligate mutual specialization in bee–plant interactions.     

The pollination efficiency of a pollinator depends on its foraging strategy,flowering phenology,and the flower characteristics of a plant species

Honey bees and stingless bees are generalist visitors of several wild and cultivated plants. They forage with a high degree of floral fidelity and thereby help in the pollination services of those plants. We hypothesized that pollination efficiency might be influenced by flowering phenology, floral characteristics, and resource collection modes of the worker bees. In this paper, we surveyed the foraging strategies of honey bees (Apis cerana, Apis dorsata, and Apis florea) and stingless bees (Tetragonula iridipennis) concerning their pollination efficiencies. Bees showed different resource gathering strategies, including legitimate (helping in pollination as mixed foragers and specialized foragers) and illegitimate (serving as nectar robbers and pollen thieves) types of flower visitation patterns. Foraging strategies are influenced by the shape of flowers, the timing of the visitation, floral richness, and bee species. Honey bees and stingless bees mainly acted as legitimate visitors in most plants studied. Sometimes honey bees served as nectar robbers in tubular flowers and stingless bees as pollen thieves in large-sized flowers. Among the legitimate categories, mixed foragers have a comparatively lower flower visitation rate than the specialized nectar and pollen foragers. However, mixed foragers have greater abundance and higher values of the single-visit pollination efficiency index (PEi) than nectar and pollen foragers. The value of the combined parameter ‘importance in pollination (PI)’ was thus higher in mixed foragers than in nectar and pollen foragers.     

Visual marks learned by the solitary bee Megachile rotundata for localizing its nest

We studied homing behaviour of leaf-cutter bees, Megachile rotundata, by using artificial landmarks. We evaluated their nest-searching behaviour in different test situations to elucidate the nature of the visual marks they used in this task. When we modified or removed geometrical figures surrounding the nest, the bees searched for longer, showing that they noticed the introduced changes. However, these manipulations never prevented bees from finding their nest, suggesting that other visual cues were crucial in the task. Manipulations of the edges provided by the boundaries of the device (nest block, metal sheet on which the block was mounted) strongly impaired the homing performance. The further away the edges that were left intact, the stronger was the impairment of the homing behaviour. These results suggest that bees learn the distances of the various edges from the goal and that edges have a hierarchical significance according to their distance from the nest. The most distant edges provide vague information, which suffices to guide the insect towards the next edge in the sequence, until it recognizes the final, precise location of the nest. The results support the conclusion that information on distances is acquired using cues derived from motion parallax generated by the insect's self-motion. Recognition of edge parameters such as position and orientation might be achieved by an image-matching mechanism based on dynamic processes. Thus, in the homing task, there is no clear discrepancy between the eidetic and the parametric hypotheses of spatial representation.     

The Potential Influence of Bumble Bee Visitation on Foraging Behaviors and Assemblages of Honey Bees on Squash Flowers in Highland Agricultural Ecosystems

Bee species interactions can benefit plant pollination through synergistic effects and complementary effects, or can be of detriment to plant pollination through competition effects by reducing visitation by effective pollinators. Since specific bee interactions influence the foraging performance of bees on flowers, they also act as drivers to regulate the assemblage of flower visitors. We selected squash (Cucurbita pepo L.) and its pollinators as a model system to study the foraging response of honey bees to the occurrence of bumble bees at two types of sites surrounded by a high amount of natural habitats (≥ 58% of land cover) and a low amount of natural habitats (≤ 12% of land cover) in a highland agricultural ecosystem in China. At the individual level, we measured the elapsed time from the departure of prior pollinator(s) to the arrival of another pollinator, the selection of honey bees for flowers occupied by bumble bees, and the length of time used by honey bees to explore floral resources at the two types of sites. At the community level, we explored the effect of bumble bee visitation on the distribution patterns of honey bees on squash flowers. Conclusively, bumble bee visitation caused an increase in elapsed time before flowers were visited again by a honey bee, a behavioral avoidance by a newly-arriving honey bee to select flowers occupied by bumble bees, and a shortened length of time the honey bee takes to examine and collect floral resources. The number of overall bumble bees on squash flowers was the most important factor explaining the difference in the distribution patterns of honey bees at the community level. Furthermore, decline in the number of overall bumble bees on the squash flowers resulted in an increase in the number of overall honey bees. Therefore, our study suggests that bee interactions provide an opportunity to enhance the resilience of ecosystem pollination services against the decline in pollinator diversity.