The Journey of a Seed on the Wind

Picture a dandelion beside a path just after the pavement has dried. The round head is full of pale tufts. A passer-by might call each flying tuft a seed. For this journey, the distinction matters. The small brown body below the tuft is a dry fruit containing a seed. The tuft of hairs is the pappus, joined to the fruit by a slender stalk. Together they form the unit that travels. We can call it a fruit with a parachute when we need a picture, and a dispersal unit when we need to include every part.

Our traveller is imagined, but the questions come from experiments. We will follow one unit from the head into moving air and then down to a surface. The story is not a prediction of where any actual fruit will land. It is a way to keep separate the things that have to happen. First the fruit has to detach. Then its fall interacts with the motion of the air. Finally it reaches a place where movement might end, or might begin again.

Begin just before release. A tuft still on the head is exposed to the weather without being free to travel. A breeze can push it without carrying it away. It is tempting to start the story when the white speck crosses the path, but that would leave out a filter. Only some of the fruits that are ready to travel will leave in any one gust. The conditions at that instant will then be the starting conditions of their flight.

Madeleine Seale and colleagues studied how moisture changes the dandelion pappus. Its hairs can draw together under moist conditions. The researchers examined the change in shape, measured aspects of flight and tested detachment from heads in a wind tunnel. These were separate tests of connected events. In one, the object was already free of the plant. In another, the question was whether it would leave at all.

The distinction makes a puzzle easier to see. If you compare an open and a partly closed pappus in the air, a change in shape can alter the way the fruit falls. But the shape can also change the force the wind exerts on a fruit that is still attached. The authors found that closure reduced detachment under their test conditions and reduced the ability of the detached unit to stay aloft. Their modelling then explored how those effects combine. A fruit that is less likely to leave in weak winds is drawn from a different set of departures than a fruit that leaves readily whenever air moves.

There is no need to imagine a plant watching the forecast. The change in shape responds to moisture. The experiment lets us test what consequences that response can have under specified conditions. The weather outside is not controlled in the same way. An individual head beside a path can be in a pocket of air unlike the air measured elsewhere, and wind can change before a fruit travels far.

For our imagined traveller, suppose the tuft is open and a gust loosens the fruit. We do not need to give the gust a precise speed. What matters is that the first filter has been passed. The little assembly is no longer attached to the head. The question now changes from what it takes to leave to what happens when the fruit and the air begin to move relative to one another.

Look at the pappus in outline. It seems almost solid from a distance, but a closer view shows spaces between its hairs. Calling it a parachute conveys the way it slows descent. A conventional parachute and a tuft of fine hairs need not interact with the air in the same way. The spaces are part of the structure, and air can pass through them.

Cathal Cummins and colleagues used a vertical wind tunnel to study that interaction. Upward airflow could balance the descent of a fruit so that it hovered in the observation area. The researchers made the airflow visible and found a ring-shaped region of circulating air separated from the pappus. It formed downstream of the tuft, above it in the hovering arrangement. This was evidence about the airflow, in addition to the readily visible motion of the fruit.

To understand why a visible swirl counts as evidence, imagine trying to explain a falling scrap of paper from a photograph of the paper alone. Its shape gives clues. Its position tells you where it was at one instant. Neither directly shows the motion of the air around it. Flow visualisation adds a way to observe that motion. A steadier setup gives the camera a chance to see a structure that would otherwise move through the frame with the fruit.

The team also tested artificial porous disks. Changing the degree of openness helped them investigate the conditions in which a separated vortex ring formed. A disk is not a plant part. Its usefulness is that it lets the experimenter vary a feature of the shape without changing every property of the fruit at once. The study links the arrangement of the pappus and its porosity to the way air moves. It does not follow an individual fruit across a field or measure whether it grows at the far end.

Suppose our fruit has settled into a steady descent in still air. Gravity pulls it down and the air resists its motion. At a steady falling speed the upward resistance balances the downward weight. This is a useful reference case. It tells us how the assembly falls relative to the air around it when conditions are steady. It does not mean that a fruit outdoors moves straight down at a fixed speed relative to the ground.

Now imagine the surrounding air moving sideways. In the simplest picture, once the fruit is moving along with that sideways flow, it travels across the ground while falling through the air. The longer it remains aloft, the longer the sideways motion can carry it. To make that connection visible, use invented numbers. If a fruit descends 0.25 metres per second relative to still air and begins one metre above an unobstructed surface, it reaches the surface in four seconds in this model. A steady sideways motion of one metre per second would carry it four metres in those four seconds. These are explanatory numbers, not measurements from the cited studies.

Alter only one condition. Let the air rise at 0.15 metres per second while the fruit continues to descend through it at 0.25 metres per second. In this simplified picture the fruit now descends towards the ground at 0.10 metres per second. It has ten seconds rather than four before reaching the surface. At the same sideways speed it could travel ten metres. If the updraft exceeded the assumed falling speed, the model fruit would temporarily rise relative to the ground. The speed through the air and the speed past the path are different quantities.

An outdoor journey seldom offers such steady conditions. Gusts, vegetation and changes in height can alter which air the fruit encounters. The value of the arithmetic is to show why a horizontal wind measurement alone does not specify the distance. A modest change in time aloft can matter even when the sideways flow is held fixed. Our imagined fruit can cross the path without having the same journey as the next one released.

Across the path, picture a low strip of grass beside a bare patch of soil. A tuft may catch on a stem before the fruit reaches the ground. It may settle directly on the soil. Later movement can shift it again. The point at which we stop calling it in flight is not necessarily its last position. For a study that maps landing, it matters whether a researcher records first contact, the place where the fruit rests later, or where a seedling is eventually found.

This difference changes how a map should be read. A map of where seedlings occur includes several events after release and travel. A seed in a fruit must arrive in a place where it can germinate, and a seedling must survive long enough to be counted. Some fruits may travel far without being represented on that map. A map of intercepted fruits asks a different question and can record arrivals that never produce a plant.

You can see the logic with an imaginary pair. One fruit travels a short distance and stops in a crack with a little soil. Another travels much farther and comes to rest on a smooth roof. The second is the longer journey in a study of movement. If the first later produces a seedling and the second does not, the first is the one visible in a survey of establishment. The example specifies outcomes to separate the questions. It does not predict how often either outcome occurs.

Our traveller catches briefly on a grass stem and drops into the strip beside the path. The descent stops for now. The book stops short of assigning it a future plant. We have followed the mechanism and reached a place where other evidence would be needed. Germination and survival cannot be read from the shape of a pappus in flight.

If you compare two journeys, keep the starting point and the end point explicit. A fruit released from a taller stem has farther to fall even in still air. One that meets grass above the soil can make its first contact sooner than one over bare ground. The distance across the map is easier to interpret when you also know the vertical distance available for descent.

A record of the release would begin at the head. It might note whether the pappus was open, what the nearby air was doing and whether other fruits left in the same gust. A record of the flight would ask about the falling unit and the air around it. A record of arrival would specify what counted as contact and where it was seen. The three records can be linked, but none is a substitute for the others.

Think of a film clip that begins after release. It may show a beautiful track across a dark background. If the camera did not record the head, the clip cannot reveal why the fruit detached at that moment. If there is no scale or reference to the motion of the air, the clip cannot separate descent through the air from movement of the air itself. If the fruit leaves the frame, the clip cannot tell you where it landed. Each limit suggests what another observation could add.

There is a similar lesson in the laboratory studies. Showing a separated ring of circulating air helps explain the performance of an open tuft. Showing how moisture changes the shape and detachment helps explain which fruits begin to travel under different conditions. Predicting a spread of possible distances requires further assumptions about weather and the surroundings. Checking what grows at the destination would ask yet another question. The power of the studies lies in making particular parts of the journey observable.

Return to the original head. There are still pale tufts on it. Some might leave in the next gust, while others may wait through a change in the weather. The small traveller we followed lies beyond the path. For a moment we can hold two pictures at once, the quiet head and the air moving through a released tuft. The journey becomes intelligible when we can say where one picture ends and what the next one needs to show.

Cummins and colleagues’ [original wind-tunnel study](https://www.nature.com/articles/s41586-018-0604-2) and its [author manuscript](https://pure.hw.ac.uk/ws/portalfiles/portal/25369594/A_separated_vortex_AAM_Cummins_et_al_2018_AAM.pdf) document the pappus, hovering arrangement, flow visualisation and porous disks. Seale and colleagues’ [experimental and modelling study](https://elifesciences.org/articles/81962) documents moisture-related shape change, flight and detachment experiments, and their combination in a dispersal model. The path, individual traveller, alternative landings and numerical flight examples in this book are illustrative. Sources consulted 4 November 2026.