Dance Decoding Research Articles

Foraging distance distributions reveal how honeybee waggle dance recruitment varies with landscape

Honeybee (Apis mellifera) colonies use a unique collective foraging system, the waggle dance, to communicate and process the location of resources. Here, we present a means to quantify the effect of recruitment on colony forager allocation across the landscape by simply observing the waggle dance on the dancefloor. We show first, through a theoretical model, that recruitment leaves a characteristic imprint on the distance distribution of foraging sites that a colony visits, which varies according to the proportion of trips driven by individual search. Next, we fit this model to the real-world empirical distance distribution of forage sites visited by 20 honeybee colonies in urban and rural landscapes across South East England, obtained via dance decoding. We show that there is considerable variation in the use of dancing information in colony foraging, particularly in agri-rural landscapes. In our dataset, reliance on dancing increases as arable land gives way to built-up areas, suggesting that dancing may have the greatest impact on colony foraging in the complex and heterogeneous landscapes of forage-rich urban areas. Our model provides a tool to assess the relevance of this extraordinary behaviour across modern anthropogenic landscape types.

Agricultural grasslands provide forage for honey bees but only when nearby

Knowledge of foraging currencies and costs is important for understanding honeybee food collection economics and to parameterize their foraging behaviors as indicators of habitat quality, which is important in the identification of management targets in human-altered landscapes. Previous research has yielded inconsistent results regarding the relationship between honey bees and important agroecosystems, such as agricultural grasslands. Waggle dance decoding provides a method for resolving these inconsistencies by mapping and quantifying bee recruitment to agricultural grasslands using statistical methods that appropriately account for foraging distance, or cost. Here we decoded 3881 dances across two foraging years to investigate when and where honey bees forage in a mixed-use landscape in Virginia, with a particular interest in honey bee use of agricultural grasslands (pastures and haylands). We initially observe that bees recruited heavily to agricultural grasslands compared to croplands, developed lands and forests, where the percent foraging to that land type was at 30.7% (CI: 29.4–31.8%), and thus significantly higher than its representation in the landscape (c. 23%). Honey bees also recruited heavily to agricultural grasslands across months, with percent foraging ranging from 26.9% (23.5–30.1%) in August to 38.8% (31.3–46.9%) in October. However, when we examined distance-corrected foraging rates, which allowed us to compare land type attractiveness when flight cost is removed, we found that the agricultural grasslands were not more attractive than the broader landscape and were significantly less attractive than, for example, croplands. We additionally identify potential forage gaps in agricultural grasslands during June and August, while also distinguishing them as a possible source of forage in October before colony overwintering in this landscape. Furthermore, we qualitatively observe a hot spot, demonstrating high foraging interest that is composed of agricultural grasslands, developed lands, and croplands and is itself a mixed-use area. Together, these results demonstrate that honey bees utilize heterogeneous land areas and underscore the importance of statistical analyses that incorporate biological knowledge. Lastly, these data will be important in informing future management aimed at pollinators in agricultural grasslands.

Foraging of honey bees in agricultural landscapes with changing patterns of flower resources

The demand for crop pollination is increasing and honey bees are frequently used, in particular as wild pollinators are in decline. Temporal and spatial variation of flower resources affects foraging decisions of wild and honey bees. To optimise crop pollination management a better understanding of potential competition for pollinators in mass- and minor-flowering crops is needed.We combined waggle dance decoding, pollen load analysis and field surveys to identify the habitat preferences and pollen use of honey bees in response to spatio-temporal changes in resource availability. Observation hives were placed on the edge of eleven fields of blooming strawberries (mean 2.24 ha) located in landscapes with different amounts of oilseed rape (OSR), semi-natural habitats (SNH) and apple trees in Germany. In addition, we surveyed honey bees and wild bees in strawberry fields.Honey bee dances more often indicated strawberry, OSR fields and SNH than expected given their landscape-wide areas. Honey bees collected on average 7.9 % strawberry, 49.0 % OSR, 30.2 % Pyrus type (e.g. apple) and 12.9 % other pollen types. The mean honey bee foraging distance was 740 m, and decreased with OSR availability. In the observation hives, dances for strawberry fields were not directly affected by OSR availability or SNH land cover. But large amounts of OSR reduced overall honey bee and bumble bee abundance in strawberry fields, while solitary bees were unaffected. Bumble bees were most abundant in strawberry fields (54.1%) and together with solitary bees (19.7%) they represented about 75.0% of the observed bees.Minor-flowering strawberry fields represent a preferred resource for honey bees, especially for small colonies as indicated by decoding of waggle dances. However, the availability of more attractive OSR and local strawberry flower cover moderates the abundance of social bees (honey bees and bumble bees) in strawberry fields while other wild bees were less affected. Hence, we conclude that wild bee conservation plays a major role for strawberry pollination. If pollination services by solitary bees are limited, small honey bee hives can be used scrupulously to supplement pollination services in strawberries.

Dismantling Babel: creation of a universal calibration for honey bee waggle dance decoding

Noisy animal communication presents a unique challenge for biologists interpreting a produced signal precisely and accurately. One past approach has been to reduce noise. Lately, with the advance of quantitative tools, an alternative is to use the noise to improve model predictions. Here we take the latter approach to improve our ability to recover honey bee ( Apis mellifera spp.) foraging locations, in particular the distance at which they forage, from observed waggle dance communications, which encode the vector from the hive to the forage. This vector gives both the distance from the hive to the forage, which is encoded in dance duration, and the direction from the hive to the forage, which is encoded in the angle that the bee makes while dancing. We analysed the waggle dances from individually marked honey bees ( N = 859 dances from 85 bees from 3 hives) that foraged at multiple known locations to determine a distance-to-duration calibration. Next we compared this calibration to a previously published calibration, which demonstrated that while their slopes were similar ( P = 0.82), their intercepts differed ( P N = 84) (performance comparison: P = 0.36) and second with a linear discriminant analysis, which failed to assign dances to their originating population. Both results confirm that the universal calibration may be used, irrespective of landscape or subspecies. Therefore, by allowing noise to be part of our analysis, the usability of our calibration is broadened for use in future settings.

Following the dance: Ground survey of flowers and flower-visiting insects in a summer foraging hotspot identified via honey bee waggle dance decoding

Decoding of honey bee waggle dances has previously shown that average foraging distances are longest during July and August in Sussex, United Kingdom, indicating a scarcity of summer floral resources. However, it also identified a summer foraging ‘hotspot’ in agricultural land at 2–3km distance. Unfortunately, dance decoding does not yield precise foraging locations or information on the flower species visited. Therefore, we surveyed this hotspot during July and August 2012 and 2013 in order to identify the habitats and flower species used by honey bees and other flower-visiting insects (FVI).The hotspot area consisted predominantly of four habitat types potentially attractive to FVI: pasture fields, field margins/hedgerows of arable fields, set-aside and a National Nature Reserve. We surveyed three fields within each habitat type. The abundance of flowers was found to be a key determent of FVI abundance per field (p=0.002). Field margins/hedgerows were the most flower abundant habitat type (p=0.002) and had more than twice (235%) the FVI abundance (p=0.001) and species richness (p=0.035) per unit area than did pasture fields. Areas with long grass had greater flower abundance (p<0.001) and FVI species richness (p=0.009) than those with short grass (≤30cm). The five plants on which we recorded the greatest number of FVI were species considered to be agricultural weeds.Honey bees represented 19% of all FVI, showing that dance decoding had located a hotspot that was an important foraging location not just for honey bees but also for other types of FVI. Honey bee abundance, per transect, was strongly correlated with that of other FVI (p=0.001), particularly bumble bees (p <0.001). However, FVI groups were not found uniformly across our study site and honey bee abundance was only weakly linked to overall species richness (p=0.069).

Determining the foraging potential of oilseed rape to honey bees using aerial surveys and simulations

The modern agricultural landscape is characterised by large areas of monoculture, including fields of mass-flowering crops. These may benefit bees by providing abundant forage, but may cause harm if the fields are treated with insecticides. Here, we investigated the potential exposure of honeybee colonies to pesticides in flowering oilseed rape (OSR) in England using a combination of aerial photographs and geographical modelling. Aerial photographs were used to detect and map fields of blooming OSR in eight areas across England where OSR is grown. We determined the distance to the nearest OSR field for randomly chosen positions (i.e., potential hive locations) using a digital map made of each area. The proportion of colonies within foraging range to the nearest OSR field in each area was then calculated using published data on honeybee foraging distances obtained by waggle dance decoding. We found that the distribution of distances to the nearest OSR field varied greatly among study areas, ranging from almost all colonies being within range to almost all being out of range (6–79% were within median foraging distance). In one study area, we examined changes over a 4-year period and found little year-to-year variation due to crop rotation (23–36% within median foraging distance). Our specific results show that exposure in England to OSR and to any chemicals used in treatment varies greatly among areas. Overall, our methodology provides a novel way of quantifying the exposure of bee colonies or other central place foragers to a particular crop or land-type at a landscape level.

Environmental consultancy: dancing bee bioindicators to evaluate landscape “health”

Here we explore how waggle dance decoding may be applied as a tool for ecology by evaluating the benefits and limitations of the methodology compared to other existing ways to evaluate the honey bees’ use of the landscape. The honey bee foragers sample and “report” back on large areas (c. 100km2). Because honey bees perform dances only for the most profitable resources, these data provide spatial information about the availability of good quality forage for any given time. We argue that dance decoding provides information for a wide range of ecological, conservation, and land management issues. In this way, one species and methodology gives us a novel measure of a landscape’s profitability and “health” that may be widely relevant, not just for honey bees, but for other flower-visiting insects as well.

Honey bee dance decoding and pollen-load analysis show limited foraging on spring-flowering oilseed rape, a potential source of neonicotinoid contamination

Neonicotinoid insecticides used to treat the seeds of bee-attractive crops occur in trace amounts in nectar and pollen. Possible harm to bees has resulted in the European Commission imposing a precautionary two-year moratorium on the use of neonicotinoids on bee-attractive crops from 2013. Recent laboratory and semi-field studies on colony-level effects of neonicotinoids assumed exclusive or near-exclusive levels of colony foraging on a treated crop. But is this a realistic assumption? Six honey bee (Apis mellifera) colonies were monitored over two springs (April–May 2011/12) in two neighbouring locations (urban and rural) in and near Brighton, UK, to quantify foraging on oilseed rape, the most widespread bee-attractive crop in the UK, by decoding waggle dances and trapping pollen. The study area was representative of the UK agricultural landscape in that the percentage area cover of the blooming oilseed rape fields around the rural location was similar to the national average (3.3–3.9% vs 3.1%). The amount of foraging on oilseed rape fields, as indicated by dance decoding, was variable, but low, 0–0.02% for the urban and 2–26% for the rural location. Almost all foraging, 91–99%, was within 2km, even though honey bees can forage at distances of over 10km. Pollen trapping in 2012 supported the dance decoding results, with oilseed rape pollen comprising 14% of pollen pellets collected by foragers from rural and 4% from urban hives. The results of this study have implications for policy as they cast doubt on the generality of some previous studies on colony-level effects on social bees conducted in laboratory and semi-field settings.

Eating locally: dance decoding demonstrates that urban honey bees in Brighton, UK, forage mainly in the surrounding urban area

Urbanization is increasing worldwide. Urban habitats often support considerable biodiversity and so are of conservation value, even though they are highly modified ecosystems. Urban parks and gardens are rich in flowers that provide food for pollinators, including bees. Here, we use waggle dance decoding to investigate foraging by 3 honey bee hives located in the city of Brighton, UK, over almost an entire foraging season, April to October. Waggle dances were recorded using video cameras and decoded during framewise playback on a computer by measuring the angle and duration of the waggle phase. Foraging was mostly local (mean monthly distances 0.5–1.2 km) and mostly within the surrounding urban area (monthly means 78–92 %) versus the countryside (closest distance 2.2 km) even though this was well within the honey bee maximum foraging range (c. 12 km). These distances were lower than those from a previous study for hives located in a rural area 4.5 km away. Honey bees are very sensitive to foraging economics and foragers make waggle dances only after visiting high-quality feeding locations. Low distances advertised by dances, therefore, indicate sufficient forage nearby and show that urban areas can support honey bees year round. As a corollary, however, urban bees may provide little pollination service to agriculture especially in spring, which had the lowest foraging distances and is when the most economically important animal-pollinated UK crops, apple and oilseed rape, are in bloom.

Incorporating variability in honey bee waggle dance decoding improves the mapping of communicated resource locations

Honey bees communicate to nestmates locations of resources, including food, water, tree resin and nest sites, by making waggle dances. Dances are composed of repeated waggle runs, which encode the distance and direction vector from the hive or swarm to the resource. Distance is encoded in the duration of the waggle run, and direction is encoded in the angle of the dancer's body relative to vertical. Glass-walled observation hives enable researchers to observe or video, and decode waggle runs. However, variation in these signals makes it impossible to determine exact locations advertised. We present a Bayesian duration to distance calibration curve using Markov Chain Monte Carlo simulations that allows us to quantify how accurately distance to a food resource can be predicted from waggle run durations within a single dance. An angular calibration shows that angular precision does not change over distance, resulting in spatial scatter proportional to distance. We demonstrate how to combine distance and direction to produce a spatial probability distribution of the resource location advertised by the dance. Finally, we show how to map honey bee foraging and discuss how our approach can be integrated with Geographic Information Systems to better understand honey bee foraging ecology.

Intra-dance variation among waggle runs and the design of efficient protocols for honey bee dance decoding

SummaryNoise is universal in information transfer. In animal communication, this presents a challenge not only for intended signal receivers, but also to biologists studying the system. In honey bees, a forager communicates to nestmates the location of an important resource via the waggle dance. This vibrational signal is composed of repeating units (waggle runs) that are then averaged by nestmates to derive a single vector. Manual dance decoding is a powerful tool for studying bee foraging ecology, although the process is time-consuming: a forager may repeat the waggle run 1- >100 times within a dance. It is impractical to decode all of these to obtain the vector; however, intra-dance waggle runs vary, so it is important to decode enough to obtain a good average. Here we examine the variation among waggle runs made by foraging bees to devise a method of dance decoding. The first and last waggle runs within a dance are significantly more variable than the middle run. There was no trend in variation for the middle waggle runs. We recommend that any four consecutive waggle runs, not including the first and last runs, may be decoded, and we show that this methodology is suitable by demonstrating the goodness-of-fit between the decoded vectors from our subsamples with the vectors from the entire dances.