Related Experiment Video
Updated: Aug 1, 2026

Tactile Conditioning And Movement Analysis Of Antennal Sampling Strategies In Honey Bees (Apis mellifera L.)
Published on: December 12, 2012
The role of mechanosensory input in flower handling efficiency and learning by Manduca sexta
Joaquín Goyret1, Robert A Raguso
1Department of Biological Sciences, Coker Life Sciences Building, 700 Sumter Street, University of South Carolina, Columbia, 29208, USA. jgoyret@biol.sc.edu
Abstract:
Nectar-foraging animals are known to utilize nectar guides-- patterns of visual contrast in flowers-- to find hidden nectar. However, few studies have explored the potential for mechanosensory cues to function as nectar guides, particularly for nocturnal pollinators such as the tobacco hornworm moth, Manduca sexta. We used arrays of artificial flowers to investigate the flower handling behavior (the ability to locate and drink from floral nectaries) of naïve moths, looking specifically at: (1) how the shape and size of flat (two-dimensional) artificial corollas affect nectar discovery and (2) whether three-dimensional features of the corolla can be used to facilitate nectar discovery. In these experiments, we decoupled visual from tactile flower features to explore the role of mechanosensory input, putatively attained via the extended proboscides of hovering moths. In addition, we examined changes in nectar discovery times within single foraging bouts to test whether moths can learn to handle different kinds of artificial flowers. We found that corolla surface area negatively affects flower handling efficiency, and that reliable mechanosensory input is crucial for the moths' performance. We also found that three-dimensional features of the corolla, such as grooves, can significantly affect the foraging behavior, both positively (when grooves converge to the nectary) and negatively (when grooves are unnaturally oriented). Lastly, we observed that moths can decrease nectar discovery time during a single foraging bout. This apparent learning ability seems to be possible only when reliable mechanosensory input is available.

