Physicists are pondering all manner of difficult problems and daunting challenges here at the 2007 APS April Meeting in Denver, but among the most pressing questions at the end of each day is, where do we go for dinner? There's always the hotel bevy of eateries to assuage hunger pains in a pinch, but savvy meeting attendees invariably whip out their pocket Zagat's Guide for the area to locate recommended restaurants in the nearby vicinity. (I'm more inclined to rely on the concierge, or Jen-Luc Piquant, who's a whiz at ferreting out online restaurant guides for the city du jour.)

But what do you do if you're not a hungry physicist or science writer, and are, instead, a single-celled organism? Where's the handy Zagat's Guide for amoebae? Apparently, they don't need one, because they've got a built-in mechanism for an optimal food-foraging strategy. Amoeba are smarter than we think (there goes that classic schoolyard taunt). For instance, scientists have always assumed that microbes move in random patterns unless they are specifically hot on the scent of tasty nibbles, but recent research has shown otherwise. A species of amoeba called Dictyostelium seems to remember its previous "steps" and uses that remembered information to explore new ground, thereby increasing its chances of finding food.

What remains unclear is how, exactly, this simple organism manages to have any kind of memory at all, even the recollection of something as short term as where it was located just a minute ago. It's not like it has a fully functioning, complex brain. Liang Li of Princeton University thinks there may be a clue in the mechanism by which the creature moves. The amoeba moves by rearranging its squishy body into a protruding shape known as a pseudopod.

It's simple four-step sequence involving an initial protrusion, followed by adhesion to the surface on which the creature is moving, then a contraction/tension of the pseudopod, followed by "deadhesion" so the creature can move instead of sticking in one place. Using phase contrast microscopy, she tracked a bunch of Dictyostelium over 100 hours, charting the "runs" and "turns" they made, which formed a zigzag pattern of motion. She specifically looked at how often the creatures made a left turn followed by a right turn, and found they showed a clear bias for that kind of variation.

How could this possibly have anything to do with an amoeba's hunting strategy? Li speculates that the formation of these pseudopods leaves temporary "scars" in the cell's cytoskeleton -- a bit like short-term memory, since the scars fade over time -- and this makes it far more likely that the next pseudopod the creature forms will point in a new direction. Because it changes direction and doesn't retrace its steps, it covers more ground and therefore increases its chances of successfully finding food.

What about more complex, higher organisms, like zooplankton? Ricardo Garcia has something to say on the subject. He's with the Center for Neurodynamics at the University of Missouri in St. Louis, and was on hand yesterday to talk about his research on the role of specific swimming characteristics in achieving optimal food foraging strategies for zooplankton. The work was billed in the press release as "the first observation in a living animal of an inherent swimming characteristic -- the turning angle -- that optimizes the food obtained in a patch of fixed size for an organism foraging for a fixed time."

Garcia and his University of Missouri colleague, Frank Moss, studied the zooplankton Daphnia, more commonly known as water fleas. They looked at the swimming movements of five different Daphnia species of varying sizes, both adults and juveniles, all of which exhibit a distinct hop-pause-turn-hop again sequence while swimming. Specifically Garcia and Moss analyzed the turning angles the creatures made after each hop in th sequence, plotting the number of times a given angle was observed in a type of chart known as a histogram. These turning angles were almost, but not quite, completely random -- they found evidence of a preferred turning angle value.

What kind of evidence? Well, it all sounded very complicated as Garcia described their work, but the gist, as far as I can tell is this: they did a mathematical analysis of the underlying random processes, or "noise intensities" in the water fleas. "Noise" in this context refers to neural (brain-related) noise, or the "random electrical 'static' in the neural systems of the zooplankton." Most of us find any kind of outside noise distracting or irritating if we're trying to focus on a particular task -- even more so if dinner is involved -- never mind if it's our own brain activity wreaking havoc with our concentration. But in the case of the humble water flea, it's helpful while foraging for food. The neural noise influences the turning angle in such a way as to enable the water flea to explore the most amount of space and gather the most food within a given time frame.

While Garcia and Moss were watching water fleas, their theoretical collaborators at the Laboratory for Applied Stochastic Processes at Humboldt University in Berlin, Germany, were developing a theory of the foraging process, complete with computer simulations. Scientists have known for many years that biological systems frequently rely on stochastic resonance (or noise) as a stimulus to the sensory systems, which in turn can affect the behavior of creatures both great and small -- usually in positive, optimizing ways that improve said creatures' chances or survival.

It's a pretty counter-intuitive notion at first glance: stochastic resonance involves adding random noise to a weak signal bearing "information" -- say, the signal from a country-western radio station in the Denver area that is just beginning to be out of range. Imagine trying to get that radio station to come in clearly on your car radio while driving through downtown Denver, but instead of fine-tuning to decrease the static, you decide to increase the level of static instead. What would prove disastrous to your enjoyment of the latest Tim McGraw opus, turns out to be beneficial for living organisms. In this case, adding noise actually enhances the detectability and/or effective transmission of the information-bearing signal.

Researchers like Garcia suspect that natural stochastic resonance may have a played a significant role in the evolution of sensory systems, although he is careful to emphasize that his results don't outright prove this hypothesis; they merely offer strong supporting evidence in favor of that notion. In the case of Daphnia, Garcia suspects that the water flea's distinctive swimming patterns evolved over tens to hundreds of millions of years via Darwinian natural selection.

It all comes down to the intensity of the neural noise signals. Those noise intensities correlate with the width of distribution of the turning angles favored by the water fleas, and it turns out that the creatures gather the most amount of food in a single foraging session at a very specific noise intensity. Per Garcia: "A small noise intensity means that the animal obtains less than the maximum possible amount of food within its patch during its fixed feeding time. Likewise, less fod is ingested if the distribution is too broad." (Jen-Luc thinks it sounds an awful lot like Goldilocks and the Three Bears: "This noise intensity is too small." "Well, this noise intensity is too large." "But THIS noise intensity is juuuust riiiight....") The findings were consistent across all five species of Daphnia studied, regardless of size or age of the organisms. So Garcia and Moss' experimental data fits the Berlin collaborators' theory just fine, and a technical paper on their work is pending publication in Mathematical Bioscience even as I type.

Which brings us to dessert. It's something I generally skip, but when I do choose to indulge, I'm partial to cookies or more elaborate concoctions like bread pudding and tiramisu. My pals MondoBob and El Finster have simpler tastes: they swear by their nightly dish of ice cream. They're not alone: according to this 2004 article from Discover on the physics of ice cream, US sales of ice cream top $20 billion, with Americans consuming as much as 20 quarts per capita. Only New Zealanders eat more ice cream. The article also contains a brief sidebar noting that some ice cream manufacturers combat the formation of unwanted ice crystals in their products by adding antifreeze or ice structuring proteins (ISPs) found in certain fish, insects, plants, fungi and bacteria. (It's all about achieving an optimally appealing mouthfeel. Nobody likes a crunchy dish of ice cream, unless it's from the addition of nuts or candies.)

Yes, ISPs can prevent recrystallization. The high fat content of most ice creams serves the same purpose, but if you're watching your cholesterol, like El Finster and Mondo Bob, you're probably buying sherbets or lower-fat varieties, which do contain antifreeze proteins to step in for the missing fat. From a medical standpoint, recrystallization is undesirable because it can also cause damage to the structure of biological tissue, limiting the "shelf life" or vital organs awaiting transplant, for example. In fact, that's one of the greatest challenges in cryogenics. Cells contain a lot of water, you see, and water expands when it freezes, bursting the cell walls in the process. That's why the Pittsburgh scientists who revived the infamous "zombie dogs" a few years ago had to replace the canine's blood with a kind of "antifreeze" to prevent tissue and brain damage during the thawing out process.

The ISPs accomplish this by attaching to the surface of ice crystals in such a way as to inhibit their growth. But some creatures have more effective ("hyperactive") ISPs, such as the spruce budworm. According to Ohio University's Ido Braslavsky, who reported on his latest research in this area yesterday, the hyperactive ISPs of the spruce budworm stop ice crystals from growing in particular directions. He and his collaborators fluorescently tagged the ISPs from fish and the budworms and studied them under a fluorescent microscope to see how the proteins interacted with the surfaces of ice crystals, on order to reach that conclusion. Future applications could include prevention of frostbite in crops; reducing highway damage from de-icing procedures in the winter; and the aforementioned food preservation and improved organ preservation.

So clearly, when it comes to science, much depends on dinner... or dessert. We shall try to bear that in mind when we undertake yet another food-foraging mission this evening in beautiful downtown Denver.