It's a small, sticky world out there.
Story by Adam Summers - Illustrations by Sally J. Bensusan
Mantis shrimp, or stomatopods, are creatures out of a sci-fi flick. They have a pair of raptorial appendages armed with (depending on the species) either wicked spikes or a crushing mallet. Their elongate, jointed carapace sports stalked eyes, the most complex in the animal kingdom. Each eye is stereoscopic, and in some species the eyes have sixteen visual pigments (humans have three), allowing them to see ultraviolet and polarized light. As good as the eyes of stomatopods are, however, their ability to see clearly is limited by the way light is diffracted, absorbed, and diffused in water. In addition, many stomatopods feed at dawn and dusk, when light levels are low. Therefore, like many other crustaceans, they also rely on an excellent sense of smell to find mates and food and to avoid predators.
Smelling can be tricky. If you think of odor particles as cars passing you on a street corner, then sensing an odor would be like recognizing the make and model of individual cars. Where traffic is light and moving slowly, you can identify the cars easily, but in fast-moving, heavy traffic the task is harder. A stomatopod is faced with a similar problem: it needs to sample a lot of water quickly yet still pick up low concentrations of particular odors. To do this, it samples the environment as frequently as seven times a second with flicks of its "nose"-two whiplike appendages, or antennules, in front of its eyes. Each antennule ends in three filaments, the middle one of which is lined with bundles of tiny, hairlike odor-sensing organs. These organs are only part of the story, though; much also depends on the way water flows past them. For the past several years, Mimi Koehl and Kristina Mead, of the University of California, Berkeley, have been sniffing out the remarkable fluid mechanics behind the stomatopod's olfactory acuity.
The story begins with an 1883 paper by Osborne Reynolds that addresses a common problem in engineering: why the flow of a fluid (out of a pipe, over the wing of an airplane, past the hull of a ship) is sometimes smooth and sometimes turbulent. Through a series of elegant experiments using streams of dye in transparent pipes, Reynolds discovered that smoothness of flow is related to a fluid's viscosity (essentially, its stickiness), whereas turbulence is largely determined by its mass and velocity. By mathematically mashing together such factors as the viscosity, density, and velocity of various fluids and the lengths and widths of various pipes, he came up with what became known as the Reynolds number (Re), which makes it possible to predict whether the flow of a fluid around or along an object will be smooth or turbulent. For example, the seawater surrounding a blue whale swimming at ten knots would have a very high Reynolds number, indicating turbulent flow, while the viscous ejaculate in which a batch of microscopic sperm are swimming would have a low Re, characteristic of smooth flow.
Using high-speed videos and dye streams, Mead and Koehl learned a great deal about how stomatopods sample flowing water for smells, but the researchers wanted to get a close-up look at what was going on. One solution was to build giant (more than 200 times life-size) replicas of the antennules, complete with sensory hairs. What's so useful about the Reynolds number is that engineers and biomechanists don't have to work with life-size models. They can increase or decrease the size of, say, an airplane wing being studied in a wind tunnel or a model of a stomatopod's immersed antennule as long as they change the flow speed and/or viscosity of the fluid so as to produce the same Reynolds number that is produced at normal scale.
Suppose, for instance, that the screenwriters for the movie Honey, We Shrunk Ourselves wanted to know what it would be like for the diminutive parents to sip their morning coffee from a minuscule cup. To see how coffee would behave at that scale, they could experiment with a full-size mug-as long as they replaced the coffee with a sufficiently viscous fluid, such as honey. Doing so would retain the same Reynolds number, so the overall system would behave the same. The experiment would show that, counterintuitive as it may seem, real coffee in a teeny tiny cup would act like honey, and the poor parents in the movie would find that their morning cup poured badly, stirred worse, and was hard to drink.
Mead and Koehl immersed their replicas in a large tank of viscous Karo syrup and moved the oversize appendages around at the speed needed to produce the same Reynolds number in the syrup as in the water flowing past the antennules of live stomatopods. Suspended in the syrup were tiny beads, some flowing between the hairs and others bypassing them. Observing the beads enabled the researchers to measure the hairs' "leakiness"-that is, how much fluid passed between the hairs (potentially coming in contact with the odor-sensing organs) rather than flowing uselessly past them. Leakiness is thus a good thing. But what if all that odor information goes by too quickly for organs on the hairs to "read" it?
The speed of the antennules' flick and the distance between the hairs turn out to be the key to keen smelling. During the rapid outward stroke, the array of hairs becomes very leaky, allowing water with lots of interesting smells to contact the sensors. During the much slower return stroke, leakiness drops, not much new water flows through the hairs, and the water from the outward stroke is trapped there until the next flick clears it out and brings in a fresh batch of water and aromas. Each flick is a separate sniff.
Amazingly, as a stomatopod grows, the size of its antennules and the speed with which they flick stay in tune, so that there is always both a leaky and a nonleaky stroke. For these little creatures, a nose has to be more than sensitive; it also has to have all the right moves.