srep30401.pdf (1.69 MB)
Signatures of a globally optimal searching strategy in the three-dimensional foraging flights of bumblebees
journal contribution
posted on 2023-07-26, 13:53 authored by Mathieu Lihoreau, Thomas C. Ings, Lars Chittka, Andy M. ReynoldsSimulated annealing is a powerful stochastic search algorithm for locating a global maximum that is hidden among many poorer local maxima in a search space. It is frequently implemented in computers working on complex optimization problems but until now has not been directly observed in nature as a searching strategy adopted by foraging animals. We analysed high-speed video recordings of the three-dimensional searching flights of bumblebees (Bombus terrestris) made in the presence of large or small artificial flowers within a 0.5 m3 enclosed arena. Analyses of the three-dimensional flight patterns in both conditions reveal signatures of simulated annealing searches. After leaving a flower, bees tend to scan back-and forth past that flower before making prospecting flights (loops), whose length increases over time. The search pattern becomes gradually more expansive and culminates when another rewarding flower is found. Bees then scan back and forth in the vicinity of the newly discovered flower and the process repeats. This looping search pattern, in which flight step lengths are typically power-law distributed, provides a relatively simple yet highly efficient strategy for pollinators such as bees to find best quality resources in complex environments made of multiple ephemeral feeding sites with nutritionally variable rewards.
History
Refereed
- Yes
Volume
6Issue number
30401Publication title
Scientific ReportsISSN
2045-2322External DOI
Publisher
Nature ResearchFile version
- Published version
Language
- eng
Official URL
Legacy posted date
2016-07-27Legacy creation date
2016-07-27Legacy Faculty/School/Department
ARCHIVED Faculty of Science & Technology (until September 2018)Usage metrics
Categories
No categories selectedLicence
Exports
RefWorks
BibTeX
Ref. manager
Endnote
DataCite
NLM
DC