Data from: Honeybees directly outcompete other flower visitors through interference competition on shared resource

Buckle, Victoria 1 ; Brown, Peter1; Carter, Toby1; Ings, Thomas 1

Research facility: Anglia Ruskin University

Published Aug 14, 2026 on Dryad. https://doi.org/10.5061/dryad.vhhmgqp96

Data files

Aug 14, 2026 version files 801.10 KB

Abstract

Honeybees are known to have a negative impact on wild pollinators. This study examines the effects of interference competition as a mechanism. We observed the behaviours of pollinators foraging on bramble and ivy. Observations took place across two years at UK sites in close proximity to honeybee hives. We documented the outcome of 300 pairs of interactions, amounting to 93 hours of observation time. The outcomes were recorded as ‘win’ where one individual gains access to the resource resulting in a ‘loss’ of forage for the opposite interacting partner, or ‘share’ where two gain access to the resource in parallel with no interaction. Overall, honeybee abundance strongly negatively influenced the abundance of non-Apis pollinators. Competitive interactions were more frequent and led to greater than expected ‘wins’ for honeybees. Honeybees were avoided in 28.4% of interactions on bramble, and 35.6% on ivy. The honeybee displaced or gave chase when interacting with another pollinator in at least 20% of winning outcomes on both plant species. We conclude that honeybees dominate shared resources by restricting the access of other pollinators to resource, and that the importance of interference competition in pollinator communities is being underestimated.

Dataset DOI: 10.5061/dryad.vhhmgqp96

Description of the data and file structure

This dataset contains data files associated with the article titled: Honeybees directly outcompete other flower visitors through interference competition on shared resource. They include a file containing the characteristics of sites and locations. Two data files for use in models which describe the variables collected during transect surveys. The transect survey model data include associated R code with the models which were run for the analysis. The pollinator interactions data includes encounters with (win, lose, share) on two plant species. In addition an ethogram describing the behaviours is associated with the pollinator interactions file.

Description of data and file structure

File: S1_Honeybees_directly_outcompete_other_flower_visitors_through_interference_competition_on_shared_resource_sites_table.pdf

Description: Table of site locations and characteristics

File: S2_Honeybees_directly_outcompete_other_flower_visitors_on_shared_resource_Bramble_modeldata.csv.csvBramble_modeldata

Description: Bramble_modeldata. Totals of pollinator abundances, densities and visit numbers for each samples and site at bramble patches.

S3_Honeybees_directly_outcompete_other_flower_visitors_through_interference_competition_on_shared_resource_Ivy_modeldata.csv.csv

Description: Ivy_modeldata. Totals of pollinator abundances, densities and visit numbers for each samples and site at Ivy patches.

Variables for S2 & S3

A (site_ID): site identifiers
B (date): date of the record
C (year): year of record
D (sampling_run): the sampling visit number (run), from 1-4.
E:(Patch_no): the patch visiting during the record
F: (timeblock): time of day of each record. Detailed as 1-3 (1= 10:00-12:0, 2 = 12:00-14:00, 3 = 14:00-16:00)
G: (fl_units): the number of floral units recorded
H: (temp): temperature at each sample record, taken with kestrel monitor (specified in method)
I: (hum): humidity at each sample record, taken with kestrel monitor (specified in method)
J: (windsp): windspeed at each sample record, taken with kestrel monitor (specified in method)
K: (maxwind): maximum windspeed measured at each sample record, taken with kestrel monitor (specified in method)
L: (hb_ab): Honeybee absolute abundance
M: (hb_relab): Calculated honeybee relative abundance
N: (hb_visits): The total number of honeybee visits
O: (hb_visitrate): Honeybee visit rate. The total number of honeybee visits/time of sample.
P: (hb_density): Honeybee density
Q: (bb_ab): Bumblebee absolute abundance
R: (bb_relab): Calculated Bumblebee relative abundance
S: (bb_visits): The total number of Bumblebee visits
T: (bb_visitrate): Bumblebee visit rate
U: (sb_ab): Solitary bee absolute abundance
V: (sb_relab): Calculated Solitary bee relative abundance
W: (sb_visits): The total number of Solitary bee visits
X: (hf_ab): Hoverfly absolute abundance
Y: (hf_relab): Calculated Hoverfly relative abundance
Z: (hf_visits): The total number of Hoverfly visits
AA: (other_ab): Other pollinators absolute abundance
AB: (other_visits): The total number of Other pollinators visits
AC: (other_relab):Calculated Other pollinators relative abundance
AD: (nonapis_ab): Absolute abundance of all non-apis pollinators in the sample line
AE: (nonapis_relab): relative abundance of all non-apis pollinators in the sample line
AF: (nonapis_visits): all non-apis pollinators in the sample line visits
AG: (nonapis_visitrate: visitation rate of all non-apis pollinators
AH: (nonapis_relab): relative abundance of all non-apis pollinators in the sample line
AH: (tot_poll): absolute abundance of all pollinators counted in this sample line
AI: (tot_visits): visit number of all pollinators counted in this sample line
AJ: (tot_visits): visit rate of all pollinators counted in this sample line
AK: (sp_no): number of species observed
AL: (int_pairs): number of pairs of pollinator encounters (interactions)
AM: (Flower): plant species

File: S4a_Honeybees_directly_outcompete_other_flower_visitors_through_interference_competition_on_shared_resource_BrambleInteractionsdata.csv

and

File: S4b_Honeybees_directly_outcompete_other_flower_visitors_through_interference_competition_on_shared_resource_IvyInteractionsdata.csv 14.99 KB

Description: List of interactions observed for bramble and ivy, shows pollinator group, taxa, interaction type, year, outcome.

Variables for S4

A (Plant sp): Plant species on which interaction was observed.
B (intno): The identifier number given to the record
C (group): Pollinator group involved in the interactions (Honeybees, Bumblebee, Solitary bee, Hoverfly or Other)
D (pollinator): Pollinator taxa, to genus or species (where possible).
E (category): the category detailing the type of behaviour observed (Attack/Attacked, Avoided/Avoids, Both leave, Chase/Chased, Collides, Displaces/Displaced, Seesaw, Seesaw (res), Shares) see ethogram (S7: table 1) for definitions of behaviours.
F (cat) abbreviation of category, where they are the recipient to a behaviour with losing outcome a '2' was added to abbreviation
G (outcome): Overall outcome of an encounter (Win, Lose, Share, Both leave)

File: S5_Honeybees_directly_outcompete_other_flower_visitors_through_interference_competition_on_shared_resource_IvymodelRcode.R.

Description: R code for models of the relative abundance of pollinator groups, species richness and number of interactions on bramble patches.

File: S6_Honeybees_directly_outcompete_other_flower_visitors_through_interference_competition_on_shared_resource_BramblemodelRcode.R.

Description: R code for models of the relative abundance of pollinator groups, species richness and number of interactions on ivy patches.

File: S7_Table1_EthogramofBehaviours.pdf

Description: Shows the descriptions of insect behaviours recorded in files S4a and S4b.

Code/software

All statistical analysis were conducted using R 4.3.0 (R core team 2023). The R code has been supplied, it details the models used in the data analysis.

This is a summary of the methods section in the associated article in order to interpret data in the supplementary files. Further details of methodology are found in the associated publication ‘Honeybees directly outcompete other flower visitors through interference comp on shared resource’,

Focused behaviour observations of honeybees and non-Apis pollinators were made on bramble and ivy during their flowering period across two years 2023-2024 in Cambridgeshire, UK. Site visits were made across a month for two plant species, Rubus fruticosus agg (Bramble) and Hedera helix (ivy).
Sampling areas were selected from five site locations where bramble and ivy were abundant (Table S1). These sites were selected as they were within 300m of known honeybee colonies (ranging 8-12 across sites) so there was a consistent abundance of honeybees and they had a similar habitat, mix of public parkland, nature reserves and edges of meadows. At each site three patches were identified and monitored. During 3-4 sampling runs per patch, conducted on separate days, three quadrats (0.5m²) were placed per patch using random coordinates. The quadrats were monitored in three 15-minute time periods during which the following were recorded: the number of flowers (floral units), flower visitor identity and abundance, the number of flowers visited by each individual, and taxonomic richness. For analysis, flower visitors were placed into one of five groups i) honeybees ii) bumblebees iii) solitary bees iv) hoverflies or v) ‘other’ which comprised non-Syrphidae Diptera, butterflies, beetles and wasps. Additional abiotic data, including temperature, humidity, and wind speed (collected with a Kestrel 3000 weather meter (manufactured by Nielsen-Kellerman, USA)) were measured three times at each sampling run, prior to patch observations. Measurements collected were included in the supplemental files (S2 and S3).

In addition to recording the abundance of flower visitors and other variables above, we also observed interactions between flower visitors. During each observation period an interaction was recorded where one individual flower visitor was being aggressive or otherwise disrupting access of another visitor to the same flower head (or cluster). Measurements taken included: the order with which individuals visited a flower (i.e., resident, or incoming), taxa of each visitor, behaviour observed, the outcome of an interaction, which flower visitor initiated the behaviour, and which was the recipient. The outcome of an interaction was allocated as win, lose, share or both leave The data collected were summarised in the interactions supplemental files S4a and S4b, with the ethogram describing behaviours in S7.