Science1 publisher3 min readPublished
Laser tracking catches a Medinilla anther tip swinging against its own stalk
Buzz-pollination biomechanics has mostly measured how hard a flower shakes. A Vienna group measured when each part of a stamen moves, and found the anther able to work against its own stalk.
The Scientist · Science desk

What happened
- A team at TU Wien, working with botanists from Uni Wien in an FWF-funded project, tested how the complex stamens of the tropical plant Medinilla magnifica respond to the vibrations bees generate.
- Individual stamens were mounted on a custom-built electromechanical shaker, driven with sinusoidal motion, and tracked by a laser line sensor reading the motion at different points along the organ.
- Zhou reports three characteristic resonances below about 1,000 hertz, with the second inside the 100 to 400 hertz range bees typically use, the first just below it and the third well above.
- At some frequencies the filament and the anther move in opposite directions, and above the second resonance the anther tip can move opposite to the rest of the stamen.
- Earlier buzz-pollination biomechanics concentrated on Solanum species, including the tomato, whose straighter and stiffer stamens behave roughly like a rigid body moving up and down.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- constraint The work measures motion in one ornamental species and stops there, so anyone arguing from it to crop yield, cultivar choice or hive placement is arguing past the data.
- decision Labs that vibrate flowers with a conventional single-point laser have to decide whether their rig resolves the upper part of the bee range, given the team's simulated 200 hertz limit.
- capability Measuring phase as well as amplitude yields a testable prediction about which frequencies bend the anther against its stalk. A pollen-count experiment can be designed to check it.
- precedent The rigid-body approximation inherited from Solanum work now has to be tested species by species, rather than assumed for flowers with curved, multi-part stamens.
When the filament and the anther move in the same direction, the stamen travels as one piece and the pollen inside is shaken along one axis. When they move against each other, the anther is bent relative to its stalk, and the grains still have to leave through the single small opening at the anther tip [8].
The second resonance sits inside the 100 to 400 hertz band, and the opposed motion of the anther tip appears above the second resonance [15][16]. Part of that anti-phase range therefore falls inside the band bees use, between the second resonance and the top of the band [25]. A bee shifting its buzz by a few tens of hertz changes the shape of the motion it delivers.
The phys.org account compares this to a washing machine, which at one spin speed becomes very noisy or starts moving across the floor while at a slightly higher or lower speed the effect disappears [24]. "This can happen, for example, because different parts of an object move together at certain frequencies, but move in opposite directions at others," said Siyang Zhou of the Institute of Mechatronics and Power Electronics at TU Wien [17][23]. "Exactly this also happens in the stamens of the tropical plant Medinilla magnifica. When we excite them at different frequencies, they respond in very different ways." [18]
Seeing phase at all required changing the sensor. "The laser point sensor conventionally used for such experiments would not be enough for Medinilla magnifica stamens," Zhou said [10]. "Our simulations showed that, with this approach, the stamen response above 200 hertz would become difficult to resolve reliably." [11] Bees vibrate flowers at 100 to 400 hertz [15]. By width, two thirds of that band lies above 200 hertz: (400 - 200) divided by (400 - 100) [20]. The line sensor reads several points along the stamen instead of one. That is how the team got usable data higher up [9].
The framing in the announcement is that what matters is not only how fast and how strongly a flower vibrates but also how its parts move relative to one another [6]. The measurements establish the first half of that: the parts do move relative to one another, inside the frequency range bees use. The report covers motion, resonances and two models, and does not include a count of pollen grains released [21]. Grains per unit of input at each frequency is what would connect opposed motion to release, and what a breeder would need in order to select on it.
Medinilla magnifica is a tropical ornamental [18]. The crop species sit in Solanum, the genus earlier work had already treated as approximately a rigid body moving up and down [7]. The team compared its measurements against a detailed finite-element model and a simplified mechanical multibody model [13]. If the simple one reproduces measured phase in other species, the next group can predict where a stamen's resonances fall without shaking every specimen.
What to watch
- Whether a follow-up pairs the same shaker with pollen counts per frequency, so opposed motion can be tied to grains released.
- Whether the simplified multibody model reproduces measured phase in Solanum stamens, where earlier work assumed rigid-body motion.
- Whether anyone measures which frequencies bees actually choose on flowers of this type, instead of citing the 100-400 hertz band as a whole.