Data from: Heathland pollinator communities are more strongly influenced by land use management than by honey bee competition
Data files
Aug 06, 2026 version files 1.04 MB
-
Brett_et_al_2026b_bee_pollen_data.csv
59.64 KB
-
Brett_et_al_2026b_flowering_phenology_data.csv
376.22 KB
-
Brett_et_al_2026b_plant_pollinator_network_data.csv
596.87 KB
-
README.md
6.02 KB
Abstract
Managed honey bees have the potential to compete with wild pollinators. Pollinator populations are also affected by land management such as grazing, but how this mediates competition with honey bees is not well understood. Many competition studies are short-term, lack appropriate controls, or do not consider the impacts of hives on the interactions of whole plant-pollinator communities. Our objective was to assess the effects of apiaries and two land management strategies (grazed/burned vs. regenerating) on wild plant and pollinator populations in upland heathlands. We used a landscape-scale before-after control-impact experiment where apiaries were introduced to half of 20 sites, with the remainder left as controls. From May to September, we recorded at each site: a) habitat structure and nectar phenology, b) plant-pollinator network structure, and c) mass and protein: lipid ratio of pollen collected by bumblebees (which have high dietary overlap with honey bees). We found that nectar phenology and pollinator community composition were strongly driven by land management. For example, grazed sites had 63% less nectar available to bumblebees in spring due to sheep browsing of Vaccinium, but supported more abundant and diverse fly communities in the summer, associated with increased grass cover and habitat heterogeneity. Whilst plant-pollinator network metrics were unaffected by apiary introductions, nestedness and pollinator robustness were lower at regenerating sites. The mass and protein: lipid ratio of bumblebee-collected pollen were consistent across treatments. Synthesis and applications: Apiaries had modest impacts on heathland plant-pollinator communities, likely buffered by the abundance of Calluna vulgaris, which provides 91% of landscape nectar during hive introductions. We recommend avoiding hive introductions outside of peak flowering periods and managing grazing in the spring to ensure nectar resources for emerging pollinators. Heathland management strategies that encourage habitat heterogeneity, such as conservation grazing, could support abundant and diverse pollinator communities that are more robust to disturbance such as hive introductions.
Dataset DOI: 10.5061/dryad.v6wwpzhc5
Description of the data and file structure
These data were collected between May and September 2022 across 21 one-hectare sites in Scottish upland heathlands. The sites were subject to two treatments: grazed v. not grazed, and apiaries v. no apiaries. Any NAs represent the missing data.
First, we quantified landscape habitat structure and nectar sugar provision at the sites, through fortnightly habitat surveys and analysis of satellite imagery. These data are provided in Brett_et_al_2026b_flowering_phenology_data.csv.
Secondly, to assess changes in pollinator foraging behaviour in response to land management and apiary introductions, we compared plant-pollinator interactions among the sites. Plant-insect interaction data were collected at the sites during fortnightly active sampling surveys. These data are provided in Brett_et_al_2026b_plant_pollinator_network_data.csv.
Thirdly, we compared the mass, protein and lipid content of the pollen baskets (corbiculae) collected from bumblebees and honey bees, providing a more direct measure of the impacts of land management and apiaries on pollinator foraging success. These data are provided in Brett_et_al_2026b_bee_pollen_data.csv.
Files and variables
File: Brett_et_al_2026b_plant_pollinator_network_data.csv
Description: A dataset containing all plant-insect visits at the sites between May and September 2022. These were used to construct interaction networks for the sites.
Variables:
- ref_no: a numeric identifier for each row, organised in chronological order of survey dates and times from May to September.
- site_name: the name of the field site
- land_type: The land management treatment, 'managed' or 'regenerating'.
- bee_type: The apiary treatment, 'apiary' or 'apiary-free'.
- site_block: The 21 sites were grouped into 3 blocks based on elevation and proximity, included in some models as a random effect during analysis.
- plot: Each site was made up of two 1-hectare plots, 'a' and 'b'.
- period: Whether the survey was prior to hive introductions ('before'), or during hive introductions ('during').
- date: the date on which the survey was carried out
- survey time: the 30-minute period during which the survey was carried out
- temp_celsius: the ambient temperature in degrees Celsius during the survey; reading taken at the beginning of the survey.
- cloud_%: the percentage cloud cover during the survey
- survey_repeat: the survey repeat e.g. the second time surveying this site
- plant_sp: the plant species observed, or to the highest possible taxonomic resolution
- insect_sp: the insect species observed, or to the highest possible taxonomic resolution
- insect_order: the order of the insect observed
- insect_family: the family of the insect observed
- freq_observed: The number of insects observed on that plant during the survey
File: Brett_et_al_2026b_flowering_phenology_data.csv
Description: A dataset containing all observations of flowering plants from May-September 2022 at the sites, used to assess nectar provision across the sites through time.
Variables:
- site_name: the name of the field site
- land_type: The land management treatment, 'managed' or 'regenerating'.
- bee_type: The apiary treatment, 'apiary' or 'apiary-free'.
- period: Whether the survey was prior to hive introductions ('before'), or during hive introductions ('during').
- date: The date on which the floral survey was carried out.
- survey_repeat: The survey repeat, e.g. the second time surveying this site.
- plot: Each site was made up of two 1-hectare plots, 'a' and 'b'.
- transect: The transect surveyed; each survey contained 4 transects.
- quadrat: The quadrat surveyed; each transect was made up of 5 quadrats.
- species: The flowering plant species.
- freq: The number of inflorescences observed in the quadrat
File: Brett_et_al_2026b_bee_pollen_data.csv
Description: A dataset containing body size and the protein and lipid content of pollen collected by honey bee and bumblebee individuals collected during the surveys.
Variables:
- site_name: the name of the field site
- plot: Each site was made up of two 1-hectare plots, 'a' and 'b'.
- bee_type: The apiary treatment, 'apiary' or 'apiary-free'.
- land_type: The land management treatment, 'managed' or 'regenerating'.
- date_collected: the date on which the individual bee was collected
- corbiculae_ref: numeric reference column for each sample
- survey_repeat: The survey repeat, e.g,. the second time surveying this site
- period: Whether the survey was prior to hive introductions ('before'), or during hive introductions ('during').
- plant_sp: the plant species on which the individual bee was observed immediately before collection
- insect_genus: the genus of bee
- insect_sp: the species of bee collected
- insect_ITD_mm: The intertegular distance measurement in millimeters for the individual bee. Used as a proxy for body size.
- protein_experimental_weight_mg: The weighed mass of the corbicula used to assess protein content (mg)
- lipid_experimental_weight_mg: The weighed mass of the corbicula used to assess lipid content (mg)
- protein_spec_mean: The spectrometer reading for protein content (%)
- lipid_spec_mean: The spectrometer reading for lipid content (%)
- predicted_protein_concentration_ug: The final calculated protein content in ug
- predicted_lipid_concentration_ug: The final calculated lipid content in ug
- protein_percent: Final percentage protein of the corbicula
- lipid_percent: Final percentage lipid of the corbicula
- pl_ratio: The protein: lipid ratio in the corbicula collected for the individual bee
Access information
These data are affiliated with the article to be published in the Journal of Applied Ecology.
