20 files

Cultural cognition in bumblebees: puzzle box open diffusion opening incidence dataset

Version 3 2023-02-01, 14:30
Version 2 2023-02-01, 10:08
Version 1 2022-10-18, 11:21
posted on 2023-02-01, 14:30 authored by Alice BridgesAlice Bridges

Raw data included in the manuscript "Bumblebees acquire alternative puzzle-box solutions via social learning" (previously published as the preprint "The establishment and transmission of novel foraging techniques indicates a capacity for culture in bumblebees (Bombus terrestris)" doi:

Dataset includes the raw data in two seperate csv files, and further 18 csv files that contain the data underlying the figures presented in the manuscript. Details of both can be found below.


Contained in two csv files: one compiling all box-opening incidence data in "experiment 1 and 2" (single-demonstrator diffusion experiments run for 6 or 12 consecutive days, respectively) and "experiment 3" (multiple-demonstrator diffusion experiments run for 12 days). 

Experiment 1 included colonies B1, B2, R1, R2, C1 and C2.

Experiment 2 included colonies B3, R3, C3 and C4.

Experiment 3 included populations 1R2B2 and 2R2B2.

Full methodological details are available in the manuscript.

Point events are listed indicating whenever a puzzle-box was opened. 

For each point event, the colony/population ID, day, camera number and time stamp are included, as is the behavioural variant used ("TabCol", with b = the blue-pushing behavioural variant and r = the red-pushing behavioural variant) and the ID of the bee responsible.

"FullOpening" refers to whether the the bee responsible for opening the box pushed the tab ≥50% of the required distance in the process of doing so. In some cases, tabs would be partially pushed by bees that left before opening the box fully, meaning that bees that would then land on the box and gain the reward would not be pushing the tab all the way. This ensured that all openings assigned to observers involved directed, sustained pushing at the tab, and so were unlikely to be by chance.

In the analysis included in the manuscript, whole colony data included demonstrator data and incomplete openings by learners.

Data that specifically referred to learners only included only full openings by observers.

Untrained bees were considered to have made the transition to learners when they had performed full box-opening twice, irrespective of behavioural variant.


A total of 18 csv files show the data underlying the following figure panels: Fig. 1C, 1D, 2A, 2B, 3A, 3B, 4B, 4C, 4D, 5A and 5B, and the following supplementary figure panels Fig. S1A, S1B, S2A, S2B, S3A, S3B and S3C. Figure legends in the main paper detail their contents. 


Brains on Board: Neuromorphic Control of Flying Robots

Engineering and Physical Sciences Research Council

Find out more...


Usage metrics



    Ref. manager