Linking agriculture and land use change to pollinator populations

Lead Research Organisation: NERC CEH (Up to 30.11.2019)


There is growing evidence that both domestic honeybees and wild pollinators are in decline in Britain, as are the wildflowers that they both use for food. We expect these declines are linked, but there is little evidence to demonstrate what the links are, and what can be done to halt or reverse them. Among the most likely explanations for these declines are changes in the way the British landscape is managed: loss of natural habitats and increasingly intensive agricultural practices. We will test for links between such land use changes and the condition of British pollinator populations. One aspect of our work involves examining how pollinator populations and land use have changed over time, to test whether pollinator losses are correlated with landscape changes. This is made difficult because there is no national monitoring programme for pollinators, and so we only have scattered information about pollinator changes. To overcome this problem, we will take two contrasting approaches: (1) comparing how land use has changed in areas where we know pollinator populations either have declined markedly or have not done so, and (2) repeating historical surveys of pollinators in sites chosen to have undergone different amounts of subsequent land use change. A second focus of our work will be on how current land uses link to current honeybee, wild pollinator and wildflower populations. To do so, we need to survey pollinators and wildflowers in sites with very different sets of conditions in a carefully controlled manner. First we will use existing datasets to estimate likely flower densities in different habitats and regions, as well as the distribution of habitat types (and changes in them), various aspects agricultural management (including pesticide usage), and the density of domestic honeybee colonies. We will then choose a set of 96 sites in six clusters across Britain, with sites chosen to represent a wide range of flower resources, different natural and agricultural habitats, different levels of pesticide use, and different numbers of domestic honeybee hives. We will survey wildflower and pollinator populations at these sites, observe how well flowers are pollinated, and test how well honeybees and one wild bee species perform when placed at the different sites. This will be the largest scale survey of flower resources and pollinator communities ever performed. We will then examine how well pollinator populations can be predicted from flower densities, and how both are affected by various aspects of land use and agricultural management. The final aspect of our work will be to apply these findings to make recommendations as to how both domestic honeybees and wild pollinators can best be protected. This will build off of the results of the previous sections, which will provide links between specific land management options and pollinator stocks. Such findings could be of use to farmers, beekeepers, conservationists and policy makers, and so we will involve all four of these communities in the project, involving representatives of each on a project Steering Committee. We will hold workshops both at the beginning of the project (to get ideas for additional issues to study) and at the end of the project (to discuss our findings and their policy implications). We will edit fact-sheets and briefing papers for these different communities, and distribute them in specialist magazines and through a project web site. We will also make information available on a public website, including teaching materials and other resources about pollinator declines.

Technical Summary

There is growing evidence of declines in honeybees and wild pollinators, and parallel declines in animal-pollinated plants. Many likely drivers of these changes involve shifts in land use and agricultural practice. We will examine links between pollinators, floral resources and land management, using both historical analyses and current observations. Historical analyses will involve (1) assessing land-use change in sites with contrasting recent pollinator dynamics, and (2) resurveys of pollinators in sites with contrasting land use changes. We will calculate land cover change for sites identified in a recent analysis as having experienced contrasting shifts in bee and in hoverfly diversity. Land cover change will be assessed relative to 1930's, 1990 and 2000 surveys. Land use change will also be calculated for sites where pollinator surveys were performed pre-1980; matched sets with contrasting changes will be resurveyed using original and standard protocols. Maps of current land management will be derived from landcover, crop, grazing, AES and pesticide datasets. Pollinator and floral resource surveys will be conducted in 96 contrasting landscapes within 6 regional clusters over 2 years. Honeybee and solitary bee colonies will be set out to test landscape effects on foraging and colony growth, and pollination services assessed on test plant arrays and wildflowers. Data will be analysed with Bayesian network methods to assess the importance of different variables in explaining floral resources, honeybee performance, wild pollinator status and pollination services. Potential mitigation options will be explored throughout the research, and workshops will be held at the beginning and end of the project to explore mitigation-related issues and to promulgate results. Project recommendations will be drafted for stakeholder communities (farmers, beekeepers, conservationists, policy-makers) and disseminated through specialist magazines and electronically.

Planned Impact

We expect the proposed research to have high scientific impact, high policy relevance, and to be of high interest to the general public. Scientific impact: Our proposed research will provide an important step forward in pollinator research, providing the strongest test to date of the role of land management and agriculture in bee declines. Our work will be by far the largest scale pollinator habitat and floral resource inventory ever carried out. We anticipate at least a dozen high impact publications will result from the work, ranging from highly applied analyses of drivers of honeybee colony mortality and wild pollinator diversity declines to more theoretical discussions of plant-pollinator interaction networks. Policy relevance: Our proposed research has a high potential for direct application: * We will assess what aspects of land use, agricultural practice (including pesticide usage), floral resources and competing pollinator densities are associated with honey production and colony mortality in domesticated honeybee stocks. * We will assess the impact of land use, agricultural management practices, floral resourses and competing honeybee densities on wild pollinator density and diversity, with special attention to the role of current mitigation practices (agri-environmental schemes) and to designing potential future mitigation methods Concerns about the viability of honeybee stocks and wild pollinator populations have been the subject of growing national, European and global policy concern in recent years. Our work will assess the importance of agricultural practices and land management in driving these losses, and as such, has a high potential impact. We expect our results will be directly relevant to policy formation, in particular to the development and assessment of agri-environmental schemes, of ESA guidelines, and other land-management related policy initiatives. Issues of land use and agricultural management have a broad policy interest beyond the specific issues affecting pollinators (e.g. CAP reform, Habitats initiative), and our work will also contribute to this much broader policy discussion. Public interest: The proposed research is of high potential public interest: both pollinating insects and the flowers upon which they feed are widely valued by the public, and there is great interest in the maintenance of the British rural environment more generally. Thus we expect the results of our work to be of interest to the press, and through them to the public at large. As a team, we have substantial experience dealing with journalists, and all five participating institutions maintain active press offices, with whom we have good relations. We will also make available information about the project and about pollinator losses more broadly, including primary and secondary school teaching materials, through a user-friendly public website. Impact actions: Our work will be of direct relevance to farmers (and other land managers), beekeepers, conservationists and policy makers. To enhance our engagement with these groups, we will establish a project Steering Committee, involving representatives from relevant stakeholder organisations. A broader set of stakeholders will be involved in two project workshops, at the start and close of the project. Project findings will be disseminated through leaflets in specialist magazines and though a public website.


10 25 50
Description Land cover change detection techniques from satellites.
Tools and protocols for national scale experiments.
Links between land cover change and pollinator decline.
Exploitation Route Management of natural resources.
Sectors Agriculture, Food and Drink,Digital/Communication/Information Technologies (including Software),Environment

Description Research results are likely to contribute to future grassland management to help conserve pollinators.
First Year Of Impact 2017
Sector Agriculture, Food and Drink,Environment
Impact Types Societal,Economic

Title Land cover based national stratification technique 
Description The method uses national (or regional) scale land cover maps. These are analysed using a regular grid. The proportion of land cover types in each grid square is calculated. The combined proportion of n (n is a positive integer) grid squares is compared against a target proportion (for example national land cover proportions) using Euclidian distance in feature space. The process is repeated for all combinations of n in order to minimize the Euclidian distance. This gives a set of locations that maximize 'representitiveness' of the target and in doing so provides the spatial structure in which to embed field experiments or make observations. 
Type Of Material Improvements to research infrastructure 
Year Produced 2016 
Provided To Others? Yes  
Impact The technique is being used to design natural experiment to monitor the effects of sustainable intensification on ecosystems and vice versa. This is part of a NERC funded multi-site national capability project, ASSIST. 
Title Land cover change detection 
Description Satellite based land cover change analysis techniques. 
Type Of Material Improvements to research infrastructure 
Provided To Others? No  
Impact It will enable us to develop satellite based methods for land cover monitoring. The ability to track changes in land cover will help us understand ecological dynamics, including changes to pollination services.