Density of wild honey bee, Apis mellifera, colonies worldwide.

The western honey bee, Apis mellifera, lives worldwide in approximately 102 million managed hives but also wild throughout much of its native and introduced range. Despite the global importance of A. mellifera as a crop pollinator, wild colonies have received comparatively little attention in the scientific literature and basic information regarding their density and abundance is scattered. Here, we review 40 studies that have quantified wild colony density directly (n = 33) or indirectly using genetic markers (n = 7) and analyse data from 41 locations worldwide to identify factors that influence wild colony density. We also compare the density of wild and managed colonies at a regional scale using data on managed colonies from the Food and Agriculture Organization (FAO). Wild colony densities varied from 0.1 to 24.2/km2 and were significantly lower in Europe (average of 0.26/km2) than in Northern America (1.4/km2), Oceania (4.4/km2), Latin America (6.7/km2) and Africa (6.8/km2). Regional differences were not significant after controlling for both temperature and survey area, suggesting that cooler climates and larger survey areas may be responsible for the low densities reported in Europe. Managed colony densities were 2.2/km2 in Asia, 1.2/km2 in Europe, 0.2/km2, in Northern America, 0.2/km2 in Oceania, 0.5/km2 in Latin America and 1/km2 in Africa. Wild colony densities exceeded those of managed colonies in all regions except Europe and Asia. Overall, there were estimated to be between two and three times as many wild colonies as managed worldwide. More wild colony surveys, particularly in Asia and South America, are needed to assess the relative density of wild and managed colonies at smaller spatial scales.

Why Petals? Naïve, but Not Experienced Bees, Preferentially Visit Flowers with Larger Visual Signals.

Flower evolution includes a range of questions concerning the function of showy morphological features such as petals. Despite extensive research on the role of petals in attracting pollinators, there has been little experimental testing of their importance in attracting naïve versus experienced flower-visitors. In an exploratory field study, we manipulated the ray petals of inflorescences of two garden flowers, Rudbeckia hirta and Helenium autumnale, to test the hypothesis that these showy structures primarily function to attract first-time, naïve, visitors. On their first inflorescence visit to both species, naïve honey bees and bumble bees were more likely to visit intact inflorescences, than those with ray petals removed. However, by the tenth consecutive inflorescence on the same visit to the flower patch, test insects showed no preference. A positive correlation was observed between the visitation of inflorescences with zero petals and inflorescence number on both study plants, for both bees. These results suggest that a key function of showy petals is to attract naïve, first-time visitors. Similar to how a restaurant attracts diners with a large sign, showy signals may be vital to enticing first-time visitors when competing with other establishments or plants for customers or pollinators. We hope the findings of this exploratory study will stimulate further work in this area.

Plants and pollinators: Will natural selection cause an imbalance between nectar supply and demand?

Pollination is an important ecological process. However, plant and pollinator needs are not always met. Commonly, pollen limitation reduces seed set or bees experience nectar dearth. Using a cost-benefit approach, we show that natural selection will lead to lower nectar production when pollinators are abundant, and vice-versa. At the community level, competition among plants for pollinators causes positive feedback that exacerbates pre-existing seasonal imbalances between nectar supply and demand. When pollinators are scarce, plants will be selected to produce more nectar to outcompete other plants in attracting pollinators, and when pollinators are abundant, plants will be selected to produce less nectar. We suggest ways to test this positive feedback hypothesis and note that evidence for seasonal variation in nectar availability provides preliminary empirical support. If correct, our hypothesis indicates that pollination faces a particular challenge in balancing nectar supply with pollinator demand and is a further example of the underappreciated role of positive feedback in ecology and evolution.

Energetic efficiency of foraging mediates bee niche partitioning.

Revitalizing our understanding of species distributions and assembly in community ecology requires greater use of functional (physiological) approaches based on quantifiable factors such as energetics. Here, we explore niche partitioning between bumble and honey bees by comparing a measure of within-patch foraging efficiency, the ratio of flower visitation rate (proportional to energy gain) to body mass (energy cost). This explained a remarkable 74% of the variation in the proportions of bumble to honey bees across 22 plant species and was confirmed using detailed energy calculations. Bumble bees visited flowers at a greater rate (realizing greater energy benefits) than honey bees, but were heavier (incurring greater energy costs) and predominated only on plant species where their benefit : cost ratio was higher than for honey bees. Importantly, the competition between honey bees and bumble bees had no consistent winner, thus highlighting the importance of plant diversity to the coexistence of competing bees. By contrast, tongue : corolla-tube-length ratio explained only 7% of the variation (non-significant). Our results confirm the importance of energetics in understanding community ecology and bee foraging niche and highlight the energetic tightrope navigated by foraging bees, since approximately half the nectar energy gained was expended in its collection.

Seasonal variation in exploitative competition between honeybees and bumblebees.

Honeybees (Apis mellifera) and bumblebees (Bombus spp.) often undergo exploitative competition for shared floral resources, which can alter their foraging behaviour and flower choice, even causing competitive exclusion. This may be strongest in summer, when foraging conditions are most challenging for bees, compared to other times of the year. However, the seasonal dynamics of competition between these major pollinator groups are not well understood. Here, we investigate whether the strength of exploitative competition for nectar between honeybees and bumblebees varies seasonally, and whether competitive pressure is greatest in summer months. We carried out experimental bee exclusion trials from May to late September, using experimental patches of lavender, variety Grosso, in full bloom. In each trial, we compared the numbers of honeybees (HB) foraging on patches from which bumblebees had been manually excluded (bumblebee excluded, BBE) versus control (CON) patches, HB(BBE-CON). This measure of exploitative competition varied significantly with season. As expected, mean HB(BBE-CON) was significantly greater in summer trials than in spring or autumn trials. This was despite high nectar standing crop volumes in BBE patch flowers in spring and autumn trials. Mean HB(BBE-CON) was not different between spring and autumn trials. Our results show that nectar competition between honeybees and bumblebees varies seasonally and is stronger in summer than spring or autumn, adding to current understanding of the seasonality of resource demand and competition between bee species. This information may also help to inform conservation programs aiming to increase floral resources for bees by showing when these resources are most needed.

Queen Execution, Diploid Males, and Selection For and Against Polyandry in the Brazilian Stingless Bee Scaptotrigona depilis.

Female mating frequency varies. Determining the causes of this variation is an active research area. We tested the hypothesis that in stingless bees, Meliponini, single mating is due to the execution of queens that make a matched mating at the complementary sex determination locus and have diploid male offspring. We studied the Brazilian species Scaptotrigona depilis. We made up 70 test colonies so that 50% (single matched mating), 25% (double mating), 12.5% (quadruple mating), or 0% (single nonmatched mating) of the emerging brood were diploid males. Queen execution following diploid male emergence was equal and high in colonies producing 50% (77% executed) and 25% (75%) diploid males versus equal and low in colonies producing 12.5% (7%) and 0% (0%) diploid males. These results show that queens that mate with two males with similar paternity suffer an increased chance of being executed, which selects against double mating. However, double mating with unequal paternity (e.g., 25∶75), which occasionally occurs in S. depilis, is selectively neutral. Single mating and double mating with unequal paternity form one adaptive peak. The results show a second adaptive peak at quadruple mating. However, this is inaccessible via gradual evolutionary change in a selective landscape with reduced fitness at double mating.

Garden centre customer attitudes to pollinators and pollinator-friendly planting.

Growing nectar- and pollen-rich flowering plant varieties in domestic gardens and other greenspace is an important pro-environmental behaviour that supports pollinating insects. Wildlife gardening is popular in the UK; however, public attitudes and behaviour relating to planting for pollinators are currently not well understood. We investigated these through questionnaires and interviews with customers in five garden centres in Sussex, southeast England, a relevant and useful consumer group representing horticulturally-engaged members of the public. Garden centre customers had strongly positive attitudes and were motivated to plant for bees and other pollinators: most (77%) grew pollinator-friendly varieties, while 64% would be more likely to buy a plant with a pollinator-friendly logo. Personal motivation to support pollinators was linked to a recent increase in personal and public awareness of their declines through (often negativistic) information from mass media sources. Practical implications of these findings in relation to the horticultural retail industry are discussed.

Organization enhances collective vigilance in the hovering guards of Tetragonisca angustula bees.

One benefit of group living is vigilance against predators. Previous studies have investigated the group size effect, where individual vigilance decreases as group size increases without reducing the overall ability of the group to detect predators. However, there has been comparatively little research on whether the positioning of individuals can improve the collective vigilance of the group. We studied the coordination of vigilance and its effect on predator detection in the eusocial bee Tetragonisca angustula. Nests are defended by hovering guards that detect and intercept intruders before they reach the nest entrance, in addition to those that stand upon it. We show that hovering guards are positioned nonrandomly, with a strong tendency for equal numbers on both sides of the entrance. This organization increases the collective vigilance of the guard group, as groups distributed in an even ratio, either side of the entrance, have a greater collective field of view than groups that deviate from an even ratio. Finally, we use a bioassay to show that when guards are on both sides of the entrance, their ability to detect intruders before they reach the entrance increases. Overall, our results provide strong evidence that vigilance is coordinated and that this improves nest defense. Although other group-living animals are often selfish in their individual vigilance behaviors and face competing time constraints such as foraging, the altruistic nature of eusocial insect workers has probably facilitated the evolution of coordinated vigilance, as documented here in T. angustula.

Both hygienic and non-hygienic honeybee, Apis mellifera, colonies remove dead and diseased larvae from open brood cells.

Hygienic behaviour is a group defence in which dead or diseased individuals are excluded. In the honeybee, Apis mellifera, hygienic behaviour refers to uncapping and removing dead and diseased larvae and pupae from sealed brood cells. We quantified removal of freeze-killed and chalkbrood-infected larvae from open cells in 20 colonies. We also measured removal of freeze-killed brood from sealed cells. Study colonies ranged from non-hygienic to fully hygienic (52-100% removal within 2 days). All larvae killed in open cells were removed. This shows that all colonies, including those with low hygienic behaviour against dead brood in sealed cells, are highly hygienic against dead brood in open cells and suggests that low hygienic behaviour against dead brood in sealed cells is a trait in its own right. This may also contribute to understanding why hygienic behaviour is uncommon in A. mellifera, which is puzzling as it reduces several diseases without detrimental effects. In particular, the result provides indirect support for the hypothesis that there are two adaptive peaks conferring disease resistance: (i) high hygienic behaviour: diseased brood are removed quickly, in some cases before becoming infective; (ii) low hygienic behaviour: diseased brood remain isolated within sealed cells.This article is part of the Theo Murphy meeting issue 'Evolution of pathogen and parasite avoidance behaviours'.

Landscape Scale Study of the Net Effect of Proximity to a Neonicotinoid-Treated Crop on Bee Colony Health.

Since 2013, the European Commission has restricted the use of three neonicotinoid insecticides as seed dressings on bee-attractive crops. Such crops represent an important source of forage for bees, which is often scarce in agro-ecosystems. However, this benefit has often been overlooked in the design of previous field studies, leaving the net impact of neonicotinoid treated crops on bees relatively unknown. Here, we determine the combined benefit (forage) and cost (insecticide) of oilseed rape grown from thiamethoxam-treated seeds on Bombus terrestris and Apis mellifera colonies. In April 2014, 36 colonies per species were located adjacent to three large oilseed rape fields (12 colonies per field). Another 36 were in three nearby locations in the same agro-ecosystem, but several kilometers distant from any oilseed rape fields. We found that Bombus colony growth and reproduction were unaffected by location (distant versus adjacent) following the two month flowering period. Apis colony and queen survival were unaffected. However, there was a small, but significant, negative relationship between honey and pollen neonicotinoid contamination and Apis colony weight gain. We hypothesize that any sublethal effects of neonicotinoid seed dressings on Bombus colonies are potentially offset by the additional foraging resources provided. A better understanding of the ecological and agronomic factors underlying neonicotinoid residues is needed to inform evidence-based policy.

Diploid Male Production Results in Queen Death in the Stingless Bee Scaptotrigona depilis.

As in most Hymenoptera, the eusocial stingless bees (Meliponini) have a complementary sex determination (CSD) system. When a queen makes a "matched mating" with a male that shares a CSD allele with her, half of their diploid offspring are diploid males rather than females. Matched mating imposes a cost, since diploid male production reduces the colony workforce. Hence, adaptations preventing the occurrence or attenuating its effects are likely to arise. Here we provide clear evidence that in the stingless bee Scaptotrigona depilis, the emergence of diploid males induces queen death, and this usually occurs within 10-20 days of the emergence of diploid male offspring from their pupae. Queens that have not made a matched mating die when introduced into a colony in which diploid males are emerging. This shows that the adult diploid males, and not the queen that has made a matched mating herself, are the proximate cause of queen death. Analysis of the cuticular hydrocarbon profiles of adult haploid and diploid males shows six compounds with significant differences. Moreover, the diploid and haploid males only acquire distinct cuticular hydrocarbon profiles 10 days after emergence. Our data shows that the timing of queen death occurs when the cuticular hydrocarbons of haploid and diploid males differ significantly, suggesting that these chemical differences could be used as cues or signals to trigger queen death.

Most ornamental plants on sale in garden centres are unattractive to flower-visiting insects.

BACKGROUND: Gardeners and park managers seeking to support biodiversity in urban areas often plant ornamentals attractive to flower-visiting insects. There is a huge diversity of garden plant varieties, and some recommendations are available as to which are attractive to insects. However, these are largely not based on rigorous empirical data. An important factor in consumer choice is the range of varieties available for purchase. In the UK, garden centres are a key link in the supply chain between growers and private gardens. This study is the first to determine the proportions of flowering ornamentals being sold that are attractive to flower-visiting insects. METHODS: We surveyed six garden centres in Sussex, UK, each over two days in 2015, by making 12 counts of insects visiting patches of each ornamental plant on display for sale that was in bloom. To provide a consistent baseline among different locations, we brought with us and surveyed marjoram (Origanum vulgare) plants in pots, which are known to be attractive to a wide range of flower-visiting insects. The attractiveness of plant varieties to insects was then expressed in two ways: the absolute number and relative to that on marjoram ('marjoram score'), both per unit area of plant cover. In addition, we noted whether each variety was recommended as pollinator-friendly either via a symbol on the label, or by being included in the Royal Horticultural Society's 'Perfect for Pollinators' list. Furthermore, we compared the attractiveness of plants that are typically grown for more than one year versus only one year. RESULTS: We surveyed 59-74 plant varieties in bloom across the six garden centres. In each garden centre, the distributions of variety attractiveness were highly skewed to the right, with most varieties being relatively unattractive, and few varieties highly attractive to flower-visiting insects. The median attractiveness of varieties with a recommendation was 4.2× higher than that of varieties without. But, due to the large variation there was a substantial number of both poor varieties that had a recommendation and good varieties that did not. Median attractiveness of multi-year plants was 1.6× that of single-year plants, with a similar overlap in distributions. DISCUSSION: Our study demonstrates the practicality of carrying out plant surveys in garden centres. Garden centres display large numbers of varieties for sale, most of which are in bloom. Furthermore, data gathered in garden centres appear to correlate well with data gathered in two previous studies in Sussex for plants established in gardens. Although it is unclear whether the varieties being sold in garden centres are a fair representation of varieties that are actually grown by gardeners, our results suggest that there might be considerable scope for making parks and gardens considerably more insect-friendly through judicious variety choices.

Hygienic behaviour in Brazilian stingless bees.

Social insects have many defence mechanisms against pests and pathogens. One of these is hygienic behaviour, which has been studied in detail in the honey bee, Apis mellifera Hygienic honey bee workers remove dead and diseased larvae and pupae from sealed brood cells, thereby reducing disease transfer within the colony. Stingless bees, Meliponini, also rear broods in sealed cells. We investigated hygienic behaviour in three species of Brazilian stingless bees (Melipona scutellaris, Scaptotrigona depilis, Tetragonisca angustula) in response to freeze-killed brood. All three species had high mean levels of freeze-killed brood removal after 48 h ∼99% in M. scutellaris, 80% in S. depilis and 62% in T. angustula (N=8 colonies per species; three trials per colony). These levels are greater than in unselected honey bee populations, ∼46%. In S. depilis there was also considerable intercolony variation, ranging from 27% to 100% removal after 2 days. Interestingly, in the S. depilis colony with the slowest removal of freeze-killed brood, 15% of the adult bees emerging from their cells had shrivelled wings indicating a disease or disorder, which is as yet unidentified. Although the gross symptoms resembled the effects of deformed wing virus in the honey bee, this virus was not detected in the samples. When brood comb from the diseased colony was introduced to the other S. depilis colonies, there was a significant negative correlation between freeze-killed brood removal and the emergence of deformed worker bees (P=0.001), and a positive correlation with the cleaning out of brood cells (P=0.0008). This shows that the more hygienic colonies were detecting and removing unhealthy brood prior to adult emergence. Our results indicate that hygienic behaviour may play an important role in colony health in stingless bees. The low levels of disease normally seen in stingless bees may be because they have effective mechanisms of disease management, not because they lack diseases.

Dancing to her own beat: honey bee foragers communicate via individually calibrated waggle dances.

Communication signals often vary between individuals, even when one expects selection to favour accuracy and precision, such as the honey bee waggle dance, where foragers communicate to nestmates the direction and distance to a resource. Although many studies have examined intra-dance variation, or the variation within a dance, less is known about inter-dance variation, or the variation between dances. This is particularly true for distance communication. Here, we trained individually marked bees from three colonies to forage at feeders of known distances and monitored their dances to determine individual communication variation. We found that each honey bee possesses her own calibration: individual duration-distance calibrations varied significantly in both slopes and intercepts. The variation may incur a cost for communication, such that a dancer and recruit may misunderstand the communicated distance by as much as 50%. Future work is needed to understand better the mechanisms and consequences of individual variation in communication.

Quality versus quantity: Foraging decisions in the honeybee (Apis mellifera scutellata) feeding on wildflower nectar and fruit juice.

Foraging animals must often decide among resources which vary in quality and quantity. Nectar is a resource that exists along a continuum of quality in terms of sugar concentration and is the primary energy source for bees. Alternative sugar sources exist, including fruit juice, which generally has lower energetic value than nectar. We observed many honeybees (Apis mellifera scutellata) foraging on juice from fallen guava (Psidium guajava) fruit near others foraging on nectar. To investigate whether fruit and nectar offered contrasting benefits of quality and quantity, we compared honeybee foraging performance on P. guajava fruit versus two wildflowers growing within 50 m, Richardia brasiliensis and Tridax procumbens. Bees gained weight significantly faster on fruit, 2.72 mg/min, than on either flower (0.17 and 0.12 mg/min, respectively). However, the crop sugar concentration of fruit foragers was significantly lower than for either flower (12.4% vs. 37.0% and 22.7%, respectively). Fruit foragers also spent the most time handling and the least time flying, suggesting that fruit juice was energetically inexpensive to collect. We interpret honeybee foraging decisions in the context of existing foraging models and consider how nest-patch distance may be a key factor for central place foragers choosing between resources of contrasting quality and quantity. We also discuss how dilute solutions, such as fruit juice, can help maintain colony sugar-water balance. These results show the benefits of feeding on resources with contrasting quality and quantity and that even low-quality resources have value.

Caffeinated forage tricks honeybees into increasing foraging and recruitment behaviors.

In pollination, plants provide food reward to pollinators who in turn enhance plant reproduction by transferring pollen, making the relationship largely cooperative; however, because the interests of plants and pollinators do not always align, there exists the potential for conflict, where it may benefit both to cheat the other [1, 2]. Plants may even resort to chemistry: caffeine, a naturally occurring, bitter-tasting, pharmacologically active secondary compound whose main purpose is to detract herbivores, is also found in lower concentrations in the nectar of some plants, even though nectar, unlike leaves, is made to be consumed by pollinators. [corrected]. A recent laboratory study showed that caffeine may lead to efficient and effective foraging by aiding honeybee memory of a learned olfactory association [4], suggesting that caffeine may enhance bee reward perception. However, without field data, the wider ecological significance of caffeinated nectar remains difficult to interpret. Here we demonstrate in the field that caffeine generates significant individual- and colony-level effects in free-flying worker honeybees. Compared to a control, a sucrose solution with field-realistic doses of caffeine caused honeybees to significantly increase their foraging frequency, waggle dancing probability and frequency, and persistency and specificity to the forage location, resulting in a quadrupling of colony-level recruitment. An agent-based model also demonstrates how caffeine-enhanced foraging may reduce honey storage. Overall, caffeine causes bees to overestimate forage quality, tempting the colony into sub-optimal foraging strategies, which makes the relationship between pollinator and plant less mutualistic and more exploitative. VIDEO ABSTRACT.

Using the British National Collection of Asters to Compare the Attractiveness of 228 Varieties to Flower-Visiting Insects.

Wildlife-friendly gardening practices can help conserve biodiversity in urban areas. These include growing ornamental plant varieties attractive to flower-visiting insects. Because varieties vary greatly in attractiveness, there is a need to quantify it in order to give objective advice to gardeners. Here, we used the British national collection of asters to compare the attractiveness of varieties to flower-visiting insects. We counted and identified insects as they foraged on flowers in 228 varieties growing in discrete patches that flowered during the survey period, 14 September-20 October 2012. In each variety, we also determined the overall capitulum size, the central disc floret area, and the ray floret color (blue, red, purple, or white). We also scored attributes relevant to gardening: attractiveness to humans, ease of cultivation, and availability in the United Kingdom. There was great variation among varieties in their attractiveness to insects, ranging from 0.0 to 15.2 per count per square meter, and highly skewed, with most being unattractive. A similar skew held for the two main insect categories, honey bees and hover flies, which comprised 28 and 64% of all insects, respectively. None of the floral traits or attributes relevant to gardening correlated significantly with attractiveness to insects. Our study shows the practicality of using a national collection for quantifying and comparing the attractiveness of ornamental varieties to flower-visiting insects. These results imply that choosing varieties carefully is likely to be of conservation benefit to flower-visiting insects, and that doing so is a no-cost option in terms of garden beauty and workload.

Appetite for self-destruction: suicidal biting as a nest defense strategy in Trigona stingless bees.

Self-sacrificial behavior represents an extreme and relatively uncommon form of altruism in worker insects. It can occur, however, when inclusive fitness benefits are high, such as when defending the nest. We studied nest defense behaviors in stingless bees, which live in eusocial colonies subject to predation. We introduced a target flag to nest entrances to elicit defensive responses and quantified four measures of defensivity in 12 stingless bee species in São Paulo State, Brazil. These included three Trigona species, which are locally known for their aggression. Species varied significantly in their attack probability (cross species range = 0-1, P < 0.001), attack latency (7.0-23.5 s, P = 0.002), biting duration of individual bees (3.5-508.7 s, P < 0.001), and number of attackers (1.0-10.8, P < 0.001). A "suicide" bioassay on the six most aggressive species determined the proportion of workers willing to suffer fatal damage rather than disengage from an intruder. All six species had at least some suicidal individuals (7-83 %, P < 0.001), reaching 83 % in Trigona hyalinata. Biting pain was positively correlated with an index of overall aggression (P = 0.002). Microscopic examination revealed that all three Trigona species had five sharp teeth per mandible, a possible defensive adaptation and cause of increased pain. Suicidal defense via biting is a new example of self-sacrificial altruism and has both parallels and differences with other self-sacrificial worker insects, such as the honey bee. Our results indicate that suicidal biting may be a widespread defense strategy in stingless bees, but it is not universal.

Trail pheromones: an integrative view of their role in social insect colony organization.

Trail pheromones do more than simply guide social insect workers from point A to point B. Recent research has revealed additional ways in which they help to regulate colony foraging, often via positive and negative feedback processes that influence the exploitation of the different resources that a colony has knowledge of. Trail pheromones are often complementary or synergistic with other information sources, such as individual memory. Pheromone trails can be composed of two or more pheromones with different functions, and information may be embedded in the trail network geometry. These findings indicate remarkable sophistication in how trail pheromones are used to regulate colony-level behavior, and how trail pheromones are used and deployed at the individual level.

Survey of insect visitation of ornamental flowers in Southover Grange garden, Lewes, UK.

Ornamental flowers commonly grown in urban gardens and parks can be of value to flower-visiting insects. However, there is huge variation in the number of insects attracted among plant varieties. In this study, we quantified the insect attractiveness of 79 varieties in full bloom being grown in a public urban garden that is popular due to its beautiful flowers and other attractions. The results showed very clearly that most varieties (77%, n = 61) were either poorly attractive or completely unattractive to insect flower visitors. Several varieties (19%, n = 15) were moderately attractive, but very few (4%, n = 3) were highly attractive. Closer examination of Dahlia varieties showed that "open" flowered forms were approximately 20 times more attractive than "closed" flowered forms. These results strongly suggest that there is a great potential for making urban parks and gardens considerably more bee- and insect-friendly by selecting appropriate varieties.