A study in the Friday, Nov. 30, issue of Science by researchers at the University of California, Berkeley, and Stanford University details the sophisticated way in which spiny lobsters sniff their way around a watery world, and may provide strategies for robot builders looking for efficient ways to create odor sensors.

"If you want to build unmanned vehicles or robots to go into toxic sites where you do not want to send a scuba diver, and if you want those robots to locate something by smell, you need to design noses or olfactory antennae for them," said lead author Mimi A. R. Koehl, professor of integrative biology in the College of Letters & Science at UC Berkeley. "We are learning how animal antennae capture odor molecules from the water around them. We want to understand which designs of odor-catching antennae work successfully in nature so that they could provide inspiration for man-made antennae."

Lobsters and other crustaceans sniff by flicking a pair of antennules, dragging them through the water to bring chemosensory hairs on the ends of the antennules into contact with odor molecules. On some lobsters, the antennules can be rather short, though in the foot-long Caribbean spiny lobster Panulirus argus, they are between 3 and 4 inches long, with split ends. On the outer edge of one of the split ends of each antennule is a brush of hairs sensitive to chemicals.

The question the researchers asked is whether the incessant flicking of antennules can pick up fine details of the swirling odors, and how odor molecules penetrate into the brush of chemosensory hairs.

The UC Berkeley researchers first made high-speed videos of a lobster flicking its antennules in order to determine how fast, how far and how often they flick, and the angles of the down and return strokes.

Once they digitized the images and measured these details, they created a mechanical lobster that flicked in the same way. The mechanical lobster, which they dubbed Rasta Lobsta, was simply the molted shell of a spiny lobster filled with epoxy. Fresh antennules from lobsters could be mounted on this mechanical lobster and moved by a computerized motor to reproduce the motion of a flicking antennule.

They placed the mechanical lobster downstream of an "odor" source in a large water flow tank in the Environmental Fluid Mechanics Laboratory at Stanford University. Since odors are invisible, instead of the aroma of a tasty item, such as a rotting fish, the researchers substituted a fluorescent dye.

The tank, operated by Jeffrey R. Koseff, professor of civil and environmental engineering at Stanford, and his colleagues, simulated the degree of turbulence a lobster might encounter while strolling along the ocean bottom.

Because they needed to see only the narrow slice of the odor plume hitting the antennule, which is only one millimeter wide, they shone a thin sheet of laser light through the plume. While flicking the antennule, they made high-speed, close-up videos of the eddies and filaments in the dye plume to determine if and how the dye penetrated the array of chemosensory hairs at the antennule's tip.

What they found is that, during the downstroke, the lobster pushes the antennule through the water just fast enough for the water and dye to penetrate into the brush of sensory hairs, maintaining much of the detail in the swirls of dye.

On the return stroke, however, it sweeps more slowly, and the water is unable to move between the hairs. The fine filaments of dye that penetrated between the hairs during the downstroke are trapped within the brush of hairs until the next rapid downstroke. The lobsters sniff when they flick, and with each flick their antennules capture a detailed map of the swirling odors in the water, Koehl said.

She and her colleagues had predicted this after building a large model of the tiny hairs on a lobster nose and swishing it through Karo syrup, a set-up that mimics the physics of swishing real antennules through water. These experiments showed that water does not flow between the sensory hairs unless the antennule moves very rapidly - at the speed of the flick downstroke.

What this means is that, in the lobster's real world, small differences in odor concentration in a plume are preserved and captured by the array of hairs, though it is unclear whether the lobster can take advantage of this detailed information.

"It's clear that very detailed information does get into the receptor area when the lobster sniffs," Koehl said. "The next step is to figure out if it is using that information."

This will involve working with neuroscientists who can help relate odor concentration in the hairs to electrical signals triggered by the hairs. Much work has already been done on the nervous system of spiny lobsters, one reason Koehl chose to study them.

Koehl's lobster work is one of her many projects on the boundary between biology and engineering, where she seeks to discover the physical principles embodied in biological design.

"When you look at the animal kingdom, you see lots of creatures that capture odor from water or air using antennae that are feathery or hairy," Koehl said. "We want to know how these feathery structures interact with water or air when the creatures fly or sit in a current to catch molecules, and which aspects of their design affect how they perform at catching odors."

Earlier this year, she and Catherine Loudon, a former postdoctoral student now at the University of Kansas in Lawrence, described how the silkworm moth uses its wings to fan odors efficiently through its feathery antennae. Koehl and her UC Berkeley colleagues also study the hairy noses of crabs and mantis shrimp.

These feathery or hairy structures are used for more than smelling. Many animals have feathery gills to extract oxygen from water. Copepods, the most abundant animal in the oceans, use them to filter and eat single-celled algae. Many sea creatures use them for swimming, and tiny insects use them to fly.

"While engineers need to design structures to perform specific functions, we are studying organisms that already have a structure, and our job is to figure out how they function," she said.

Coauthors on the Science paper with Koehl and Koseff are graduate student Michael G. McCay and laboratory technician Tim Cooper of UC Berkeley; graduate student Megan B. Wiley and former postdoctoral fellow John P. Crimaldi of Stanford; and Paul A. Moore of the J. P. Scott Center for Neuroscience, Mind, and Behavior at Bowling Green State University, Ohio. Crimaldi now is an assistant professor of engineering at the University of Colorado.