Smelly footprints left by bumblebees can help them find good sources of food, scientists have found.
The insects secrete invisible markers when they touch their feet on a surface, which can be detected by themselves and other bumblebees.
Researchers from the University of Bristol found that bees can distinguish between their own scent, the scent of a relative and that of a stranger.
This ability can be used to improve their success at finding good sources of food and avoiding flowers that have already been visited and mined of nutrients.
Richard Pearce, from the school of biological sciences at the university, said: "Bumblebees secrete a substance whenever they touch their feet to a surface, much like us leaving fingerprints on whatever we touch.
"Marks of this invisible substance can be detected by themselves and other bumblebees, and are referred to as scent-marks.
"We performed three separate experiments with bumblebees, where they were repeatedly exposed to rewarding and unrewarding flowers simultaneously that had footprints from different bees attached to them."
In the study, each flower type carried scent-marks from bumblebees - either from them, sisters from their nest or strangers from another nest - or were unmarked.
The bees were able to distinguish between these four flower types, showing that bees can tell marks of their own nests from strangers but also can discriminate between the smell of their own footprints and those of their nest mate sisters.
"This is the first time it has been shown that bumblebees can tell the difference between their scent and the scent of their family members," Mr Pearce added.
"This ability could help them to remember which flowers they have visited recently.
"Bumblebees are flexible learners and, as we have discovered, can detect whether or not it is they or a different bumblebee that has visited a flower recently.
"These impressive abilities allow them to be cleverer in their search for food, which will help them to be more successful."
The work, published in Scientific Reports, was funded by the EPSRC, through the Bristol Centre for Complexity Science.