Amazing Ant Rafts

It has long been known that when some ants are faced with a flood, they build “living rafts” that float on the water in search of land. The earliest reference I know of to this phenomenon comes from the book Insect Architecture, which was published in 1838:1

The ants consisting of the basis of this group, lay hold of some shrub for security, while their companions hold on by them; and thus the whole colony, forming an animated raft, floats on the surface of the water until the inundation (which seldom continues for longer than a day or two) subsides.

So scientists have known about this for a long time, but how such rafts manage to float has been a mystery. You see, a single ant can float on water for two reasons: First, water has surface tension, and that tension must be broken in order for something to sink. If laid carefully on the surface of water, for example, a metal needle will float. Even though the density of the needle indicates it should sink, the water’s surface tension will keep it afloat. Second, the water-repelling nature of an ant’s outer covering (its exoskeleton) causes tiny bubbles of air to cling to it. The combination of the air bubbles’ buoyancy and the water’s surface tension keeps the ant afloat, but not by much.

Now…if I start stacking ants on top of each other, only a few of them will be in contact with the water. The water’s surface tension combined with the buoyancy of the ants that are actually in contact with the water just isn’t enough to keep the whole stack of ants afloat. Thus, something else must be going on when ants get together to form rafts.

A recent report on fire ant rafts was just published online in the Proceedings of the National Academy of Sciences, USA. It tells us what is actually going on in ant rafts, which is really quite fascinating.

Nathan J. Mlot, Craig A. Tovey, and David L. Hu used time-lapsed photography to carefully watch the process by which fire ants form rafts. They found that the ants which are in contact with the water grasp each other by the mandibles or the legs, and then they pull each other together with a strength of up to 400 times their own weight! That would be like me pulling with a force of more than 82,000 pounds!2

What does this do? It allows them to trap air bubbles underneath them, and those air bubbles form a water-repellent surface that is significantly more efficient than the sum total of each individual ants’ water-repellent exoskeleton. The authors state it this way:

We find that ants can considerably enhance their water repellency by linking their bodies together, a process analogous to the weaving of a waterproof fabric.

Interestingly enough, the researchers found that when a portion of the raft sinks because of a water disturbance or an unexpected increase in weight, the ants that end up underwater start pulling even harder on each other. This forms an even more water-repellent surface, allowing that part of the raft to float again.

The authors make the point that when scientists look at nature, they generally think of water-repellency as a static thing. Either something is water-repellent or it is not. In addition, the amount of water-repellency is fixed. A leaf, for example, is water repellent because it is covered with a waxy cuticle. The water repellency is based solely on the thickness and makeup of that cuticle. However, this research shows that in some cases, the water-repellency of an organism is dynamic. It can change based on the behavior of the organism.

The authors suggest that engineers can learn from how ant rafts are formed, allowing them to build better lifejackets, boats, raincoats, etc. That’s not surprising, of course. Any engineer can learn a lot by studying the work of the Master Engineer!


1. James Rennie, Insect Architecture (London: Charles Knight and Co, 1838), p. 260
Return to Text

2. Nathan J. Mlot, Craig A. Tovey, and David L. Hu, “Fire ants self-assemble into waterproof rafts to survive floods,” Proceedings of the National Academy of Sciences, USA doi: 10.1073/pnas.1016658108, 2011.
Return to Text


  1. Daniel Gray April 30, 2011 5:37 pm

    Hello Dr. Wile,
    I have commented on a couple of things on your past blog posts… I am an avid young earth and I have stumbled upon a problem… The problem arises from the facts about light, which takes a specific amount of time to travel a certain distance. For example, a galaxy measured to be about 13 billion light years away must have existed about 13 billion years ago. That’s when the light from that galaxy started on its way to Earth’s telescopes. What do you think?

    • jlwile April 30, 2011 5:52 pm

      Hi Daniel, thanks for your question. It turns out that the analysis you present makes a rather critical assumption that we know is incorrect. In order to assume that light from a galaxy 13 billion light years out took 13 billion years to get here, you have to assume that the gravitational fields the light experienced from the time it left the galaxy to the time it reached the earth were all exactly the same. You see, the rate at which time passes depends on the amount of gravity in the area. Time passes more quickly when gravitational fields are low and more slowly when gravitational fields are high. This is an experimentally-confirmed fact. Indeed, the only reason your GPS device works is because the GPS takes this into account. Given that there are many ways to assume how the universe has expanded, there are many different gravitational possibilities. For example, Russell Humphreys has shown that if you assume the universe expanded in a spherical manner with earth near the center, light could reach the earth from the very edge of the universe in significantly less than 6,000 EARTH years. You might be interested in learning more about that by reading Russell Humphreys’s book, Starlight and Time. A variation of that view can be found here.

  2. Daniel Gray May 1, 2011 5:45 pm

    Thank you for responding. So the explanation in a nutshell is that because of gravity, time goes by quicker in most parts of the universe. Does time go by that much quicker? And we know that the gravity is that much weaker?

    • jlwile May 2, 2011 7:54 am

      The explanation in a nutshell is that the passage of time depends on the gravitational field. However, we have no idea exactly what the gravitational fields in different parts of the universe have been. As a result, we have to develop models to attempt to figure that out. Under the Big Bang model, it is assumed that the universe has been homogeneous at fairly large scales, so it is assumed that the gravitational fields are essentially uniform over large distances. Thus, you can assume that time behaves fairly evenly throughout the universe, at least on long scales. In addition, if you assume the universe has been expanding without any geometry at all (no edge, no center), then time behaves fairly evenly throughout the history of the universe as well. That leads to the idea that it takes 13 billion earth years for light to travel 13 billion light years across the universe.

      However, that is just an assumption, and based on observation, that assumption is pretty poor, since the observed universe does not seem homogeneous on any scale. If we make other assumptions about the nature of the universe and its expansion, we get other results. If we assume, for example, that the universe expands in sphere and we assume that the earth is near the center of that sphere, then we definitely do come up with a scenario by which the gravitational fields are so uneven throughout the universe’s expansion that light from billions of light years away can, indeed, reach earth in only a few thousand years on earth.

  3. Shawn Huston May 2, 2011 5:01 pm

    Huh, no kidding. Thanks for enlightening me to the effects of different assumptions on GR. Had no idea.

    • jlwile May 3, 2011 7:29 am

      Shawn, as they say, the devil really is in the details. Whenever we are dealing with something as unimaginably huge as the universe, there are bound to be assumptions at play. It is truly amazing how much the assumptions of universal homogeneity and universal expansion without geometry affect the conclusions of modern astrophysics!

      When it comes to dark energy, we have a lot more assumptions. For example, the reason we have to believe in dark energy to begin with is that the Big Bang model won’t work without it. We actually have to believe that more than 70% of the universe is made of something we currently cannot detect and have no understanding of whatsoever. This is one of the reasons many astrophysicists are saying that we have to abandon the Big Bang. In fact, there is at least one astrophysicist who says we can avoid the need for dark energy altogether if we just abandon the idea that the universe is homogeneous.

      In terms of dark matter, once again, we have to assume it exists in order to explain how galaxies could have held themselves together for the supposed duration of the universe. I am more favorably inclined to believe in dark matter than in dark energy. After all, it is tough to detect things even here on earth, much less in space. So I can believe we are missing a lot of the mass in the universe. However, dark matter has been invoked before to explain “unknown” phenomena, and it has been wrong. For example, dark matter was invoked to explain the odd motion of Mercury’s perihelion, but we now understand that it is due to the effects of GR.

  4. Shawn Huston May 3, 2011 5:02 am

    Speaking of GR, any insight into dark mater/energy from a young earth perspective that I should know about?

  5. Shawn Huston May 4, 2011 12:28 pm

    Wow. I guess I should have paid more attention in that GR class I took. ;) And maybe I should have taken a class in astrophysics at some point also….

    Thanks for the continuing education.

  6. Dan May 5, 2011 10:22 pm

    Dr. Jay, I definitely like Humphreys’s white hole cosmology. There is one thing I wonder about it, however. If you assume the Milky Way is the center of the universe (so that Earth could be near the bottom of the gravitational well), that would mean Earth left the white hole before the regions of the Milky Way that are closest to the center. Thus, those regions of the galaxy would have experienced more time dilation than Earth, and would be younger than Earth. Wouldn’t that defeat the whole purpose? How would that allow enough time (in Earth’s reference frame) for light from those stars to reach Earth?

    Or, could one suggest that some other part of the Milky Way was the center of the cosmic white hole at the beginning? That would seem to require that the gravitational center of the universe orbit around the local gravitational center of the galaxy. Is that even possible?

    • jlwile May 6, 2011 8:44 am

      Dan, you are correct that if the universe was centered on the Milky Way during the white hole expansion, there would be parts of the universe whose clocks have ticked for fewer years than earth’s clocks. However, if the universe was centered on the Milky Way during the expansion, the Milky Way would not have the nice spiral shape it seems to have. After all, as the event horizon shrank, parts of the galaxy would be rotating faster than other parts, which would really mess up its structure. Thus, I am not sure the model is developed enough to really answer the question of exactly where the center is. I think the model has to answer some more basic questions first. For example, how exactly did the laws of physics apply while God’s creative power was being exercised? Do we need to worry about things like rotational centers while things are still being made supernaturally?

      I think the strength of the Humphreys model is not necessarily that it tells us how the creation of the universe occurred. It is still too “young” a model for that. Its main strength is that it shows the whole idea of extrapolating time from light-travel distance is highly dependent on ad-hoc assumptions like the geometry of the universe’s expansion.

  7. Dan May 6, 2011 1:09 pm

    Thanks, Dr. Jay.

  8. Henry May 14, 2011 1:09 pm

    Fascinating. Thanks for sharing.