Investigating the impact of habitat structure on queen and worker bumblebees in the field

Lead Research Organisation: University of East Anglia
Department Name: Biological Sciences


This research will unravel fundamental aspects of bumblebee behaviour and ecology to show how habitat structure affects bumblebees at different stages in their colony cycle. Bumblebees are key pollinators of wildflowers and crops but their populations are declining worldwide. Although bumblebees have been well-studied in some respects, fundamental aspects of their ecology remain unknown. In this proposal we plan to fill some of these gaps so that we can implement the most effective measures to reverse declines. We have four specific objectives. First, little is known about the dispersal of nest-founding queens in spring, so we will determine the fine-scale spatial genetic structure and behaviour of nest-founding queens and relate this to the distribution of nesting and foraging habitats across an agricultural landscape. Second, although new genetic methods have permitted estimates of the foraging ranges of worker bees, we do not know how workers and colonies use space in relation to the fine-scale structure of their habitat; we will therefore quantify both workers' spatial distribution and habitat structure to determine this relationship. Third, some bumblebee species have long tongues and so are the most effective pollinators of wildflowers and crops in which the floral tube (corolla) is deep. The UK's longest-tongued bumblebee species, Bombus hortorum, remains common but the closely related B. ruderatus is scarce. We will compare the spatial genetic structure of these two species directly in the same landscape in order to elucidate the biological reasons for this difference. Fourth, because our study will take place at a site where targeted measures to provide forage for bumblebees (agri-environment schemes) have already been established, we will be able to use our results to model the impacts of these measures on bumblebee foraging range and nest density. To meet our objectives, we will study the five species B. lapidarius, B. pascuorum, B. terrestris B. hortorum and B. ruderatus, chosen to represent an appropriate range of ecological and behavioural differences. We will use a novel combination of genetic, ecological and modelling methods in a unique field setting: an agricultural landscape (2100ha) in England in which blocks of standardised agri-environment options targeted at insect pollinators have been established as part of an ongoing landscape-scale experiment. Our methods will consist of systematic nonlethal sampling of queens and workers for DNA, followed by microsatellite genotyping coupled with sibship reconstruction (grouping of individuals into families) to determine the spatial genetic distribution of queens and workers. We will also use advanced marking techniques (e.g. radio-frequency identification (RFID), pit-tags) to supplement our genetic investigations of queens and specifically to relate queen space use to the distribution of small mammals (whose nests queens are believed to exploit for nesting and which are already being monitored as part of ongoing studies). Finally, we have already characterized the study landscape using both high-resolution multi-spectral scanners and field surveys. The resulting detailed information on habitat composition and structure, when integrated with our data on bumblebee space use, will allow us accurately to model the impact of habitat structure on the bees. Overall, therefore, this project matches the requirements of the Insect Pollinators Initiative very well. First, because the study is taking place at a site where we have already characterized habitat structure and in which an experimental agri-environment scheme is established, it will serve to advance our understanding of the effects of agriculture and land-use change on bumblebee behaviour and colony dynamics. Second, it will use new tools and data analysis to discover fundamental aspects of bumblebee ecology relevant to pollinator declines from molecular to population levels.

Technical Summary

We will combine new molecular, technological and modelling approaches to determine how bumblebees use space at a local scale. The research will exploit a unique 2,100-ha agricultural landscape in which agri-environment flower mixes have been planted, habitat features have been aerially mapped in 2-D and 3-D detail, and the distribution and abundance of small mammals (whose disused nests potentially provide bumblebees with nest-sites) are being monitored. (1) To determine the spatial behaviour of nest-founding queens in relation to nesting and foraging habitats, we will sample queens (B. pascuorum, B. terrestris) nonlethally for DNA systematically across the landscape, genotype samples at a panel of microsatellite loci, and use the software COLONY and KINSHIP to determine dispersal distances of sister queens in relation to habitat features and small mammal distributions; we will supplement this approach with a mark-recapture study of founding queens using RFID (radio frequency identification) tags. (2) To quantify the spatial distribution of workers, we will sample foraging workers (B. hortorum, B. lapidarius, B. pascuorum, B. ruderatus, B. terrestris) systematically across the landscape for DNA, genotype them at our microsatellite loci, then estimate foraging ranges from the distribution of separation distances of sisters to determine foraging behaviour in relation to species ecology and habitat features. (3) To compare a declining species (B. ruderatus) with a related widespread species (B. hortorum), we will identify potentially important differences in these species' spatial ecology from Objective 2, confirming species identities via mtDNA differences (Ellis et al. 2006). (4) To provide an evidence base for optimising land management for bumblebee conservation, we will integrate all data to construct and parameterise models (e.g. with GIS) to predict how habitat structure and forage and nest-site availability determine local abundance of queens and workers.

Planned Impact

Who will benefit from this research? Our research is designed to address the requirements of the Initiative directly, therefore should have a substantial impact in terms of implementing actions to reverse declines in UK bumblebee populations. Academic beneficiaries will include those with interests in social insect biology as well as in agro-ecology and impacts of land-use on species and populations. Primary users of the proposed research from a policy perspective will be: i) Natural England (NE) - responsible for agri-environment schemes and promoting recovery of declining species in England; ii) Defra - the government department responsible for policy on the environment, food and rural affairs; iii) Joint Nature Conservation Committee (JNCC) - the statutory advisor to government on UK and international nature conservation; iv) The Scottish Government - responsible for agri-environment schemes in Scotland; v) The National Assembly for Wales - responsible for agri-environment schemes in Wales. Other important stakeholders include a number of NGOs working on plant and invertebrate conservation including: the local Wildlife Trusts; Bees, Wasps and Ants Recording Society (BWARS); Bumblebee Conservation Trust (BBCT), Butterfly Conservation (BC), Buglife and PlantLife. They also include the farming community, e.g. the Farming and Wildlife Advisory Group (FWAG). How will they benefit from this research? The project will allow us to model at new levels of resolution the optimal distribution of habitats for different bumblebee species. For the first time, key aspects of the behaviour of founding queens will be determined and common and rare species will be compared within the same study landscape against known distributions of their forage and nesting habitats. Together, these benefits will have relevance to those interested in the conservation of bumblebee populations worldwide. Our results will also directly inform strategies such as agri-environment schemes, thus increasing the effectiveness of major policies for reversing pollinator declines. Natural England's Environmental Stewardship (ES) scheme funds measures to benefit pollinators, specifically the sowing of a 'nectar flower mixture' which has been targeted for bumblebees. Although about 65% of English farmland is in ES, take-up of pollinator options has been low and only 2% of land actively managed under the scheme is currently under a nectar flower mixture, representing just 0.03% of the farmed landscape. It is therefore vital to both target this limited resource at the correct spatial resolution, and to gather evidence to support an increase in coverage of management options for pollinators across the UK. Our data will also help quantify the potential for pollen transport between patches or populations of plants which will be of relevance to crop production, risk assessment for novel crops and conservation strategies for native plants. What will be done to ensure that they benefit? As outlined in the Impact Plan, we will take targeted action to communicate our findings to each of the primary users named above, including a one-day workshop at the study site during the final project year for policy makers and advisors. This will help to formulate policy-relevant documentation and maximise the impacts of the research. Scientific findings will be published in high-impact scientific journals and via scientific conferences at national and international levels. Both scientific and policy conclusions will be communicated to the public at large through the CEH, ZSL and UEA websites and by press releases and follow-up media interviews. The project team has a strong history of engagement with the press and policymakers at both local and national levels. CEH maintains a Knowledge Exchange section that contains considerable expertise for activities relating to knowledge transfer. We will liaise with this department and use their expertise to achieve effective dissemination.
Description Effective pollination is essential for both wildflowers and crops, but insect pollinators are increasingly threatened. Bumble bees represent major insect pollinators in temperate ecosystems. Some species are declining, and much remains to be established regarding the ecology of even the common and widespread species. A key aspect of foraging ecology is bees' use of space, since this determines how pollination services are distributed across landscapes. In this project, we used a combination of field ecology, remote sensing, microsatellite genotyping and landscape modelling to understand the relationship between space use by bumble bee queens and workers and fine-scale features of their habitat, especially the quantity and quality of local foraging resources. The study species were Bombus terrestris, B. lapidarius, B. pascuorum, B. hortorum and B. ruderatus, and the study site was the Hillesden Estate, Buckinghamshire, UK. This is an area of arable farmland experimentally planted with sown field margins providing floral resources (from wildflower mixes) for pollinators as specified by the UK government's agri-environment scheme for England. We had three key findings:

1. Queens tend not to nest near where their mothers nested: It was previously unknown, even for common bumble bee species, whether queens tend to found nests near their natal nest or not. We captured workers of the five study species and sampled them non-lethally for DNA (before releasing them again in the study landscape). We genotyped the DNA samples at a panel of marker genes (microsatellite loci). Aided by sib-reconstruction software (COLONY), we used the genetic data to group workers into full sisterhoods, which represent workers all produced by one queen. Such workers must have come from the same nest. We estimated the position of each nest from the positions of the workers within the sisterhood, which we had mapped using GPS devices when the workers were collected. Using COLONY, we also reconstructed the genotype of each colony queen from the workers' genotypes.

We were then able to calculate, across all pairs of colony queens, the association between the genetic relatedness of queens and the physical distance between their nests in the landscape. We found that, for all five species, there was either no association between queen-queen relatedness and physical distance, or only a very weak one. This showed that queens do not tend to found nests near their natal nest (which would have led to a strong negative association of queen-queen relatedness and physical distance). Instead, on average, young queens from the same nest must disperse quite far, in different directions, before founding their nests, leading to populations that are genetically well mixed at fine spatial scales. These findings have been published in Dreier et al. (2014).

2. Workers fly less far to forage when there are more floral resources in their immediate surroundings: We next sought to understand the movements of foraging workers. Using the data on workers' positions in the landscape, we estimated workers' colony-specific foraging distances, i.e. the average distance flown by workers from a given colony to forage at flowers. We then estimated the availability of floral resources around each nest from remote sensing data and field surveys. Combining these datasets showed that workers forage more closely to their nests in areas with greater coverage of semi-natural vegetation, including agri-environmental field margins. This suggests that sowing these margins will potentially help bee populations by reducing the energy workers spend on foraging longer distances (Redhead et al. 2016).

3. Colonies are more likely to produce queens surviving to the following year in areas with more floral resources: Focusing on Bombus terrestris, B. lapidarius and B. pascuorum, we used our genetic data to estimate, for a given colony, the chances of its producing a queen that survived to emerge the following year. We did this by looking for genetic matches indicating a sister relationship between workers sampled in the summer of one year and queens caught after emerging from hibernation in spring in the following year. We then looked for associations between such family lineage survival and the nature of the floral resources around each nest. This way, we showed that bumble bee colonies nesting within 250 - 1,000 metres of high-quality floral resources within the landscape are significantly more likely to have daughter queens surviving to the following spring. These findings both confirm the value of agri-environment schemes for pollinators and indicate the spatial and temporal scale of the floral resources required for such interventions to be effective. We also showed that, on average, queens emerging in the spring have dispersed 1.0 to 1.5 kilometres from their natal nest, allowing ample opportunity for the fine-scale spatial mixing of queens from different colonies detected earlier in the project. Our paper reporting this work is Carvell et al. (2017).
Exploitation Route The findings should be relevant to conservation managers, farmers and policy-makers because of their relevance to the design of agri-environment schemes for bumble bees.
Sectors Agriculture, Food and Drink,Environment

Description In a Policy and Practice Note published by the Living with Environmental Change initiative: Carvell C, Heard M, Vanbergen A, Bourke A, Dicks L (2016) Managing farmed landscapes for pollinating insects. Policy and Practice Note No. 27, LWEC, Polaris House, Swindon. Our results have also fed into the development of the Wild Pollinator and Farm Wildlife Package within Countryside Stewardship, the Government's agri-environment scheme for England.
First Year Of Impact 2016
Sector Agriculture, Food and Drink,Environment
Impact Types Societal,Economic,Policy & public services

Description All the Buzz 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Schools
Results and Impact I and my research group assisted in the production of an animated film ('All the Buzz') and lesson plan about bumble bee conservation for use in primary schools, led by the Science, Art and Writing (SAW) Trust and Hedley Griffin Films. This was delivered to two primary schools in East Anglia and there are plans to roll it out to more schools across the UK.
Year(s) Of Engagement Activity 2017
Description Pollinators in peril: bees, genes and conservation (E England) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Public/other audiences
Results and Impact This is a talk including material on the RCUK project results given by AB to bee-keeping groups and local societies in eastern England (Cambridgeshire, Norfolk and Suffolk) on five occasions.

Public dissemination of research.
Year(s) Of Engagement Activity 2012,2013,2014
Description Poster display at UEA Open Day 
Form Of Engagement Activity Participation in an open day or visit at my research institution
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Public/other audiences
Results and Impact Many members of the public viewed a poster displaying our IPI project work and expressed interest in the methods and conclusions.

Year(s) Of Engagement Activity 2014,2015