Volume 1, Issue 3
The Tale of the Otter and the Abalone
When David Lindberg, chair of UC Berkeley's Department of Integrative Biology, was a graduate student more than twenty years ago, he was faced with a conundrum off the coast of California. Lindberg was assisting UC Santa Cruz marine ecologist Jim Estes in a protective effort to move a population of endangered sea otters from central to southern California. While the relocation was deemed necessary, conservationists also worried that the otters might destroy the rich shellfish population already living there, particularly the large abalone so familiar to seafood connoisseurs in the western United States and Japan. The concern seemed valid, but raised an evolutionary riddle.
David R. Lindberg was formerly the director of UC Berkeley's Museum of Paleontology. He's currently a curator there. (courtesy Howard Hughes Medical Institute)
"If you look at the fossil record, big abalone appear around the same time as otters in California," Lindberg says. "So if otters are a horrible predator that just bash the abalone, why did the abalone evolve to a size so big that they couldn't fit back in the rocks where they'd be protected?"
Now, Lindberg, Estes, and Charles Wray of Mount Desert Island Biological Laboratory, may have solved the mystery. The story they uncovered is a quintessential example of how modern-day integrative biology is a melting pot of molecular genetics, paleontology, and good old-fashioned natural history.
Most of the world's abalone live in tropical coral reefs and reach no more than 50mm in length, far smaller than the Northern Pacific's salad bowl-sized species. This led the researchers to hypothesize that the Northern Pacific abalone get big by eating well, specifically the high-quality algae in the cold waters. To test their hypothesis, Lindberg and his colleagues went back 40 million years.
"When you do historical studies, you can't run controlled experiments," says Lindberg, also a curator at the UC Berkeley Museum of Paleontology. "But you can look at the distant past, make predictions about the more recent past, and test your hypothesis that way. So we set out to map the abalone's size increase through time."
During the early Cenozoic era, when the configuration of earth's continents was very different than today, the forming of the Antarctic Ice Sheet allowed some abalone to dive into Australia and South America. With the cooler temperatures came an abundance of fleshy macroalgae for the abalone to feast upon leading to their first increase in size. Over millennia, the abalone followed the high-energy food, spreading eastward and north along the south Pacific basin.
During the course of their research, the scientists ran genetic tests on tissue samples for dozens of abalone species from around the world. The results showed that the large animals were all related, having completed their radiation from South to North America before the Isthmus of Panama rose from the sea three million years ago.
Placing one abalone shell inside another reveals how the 40 mm tropical species Haliotis glabra is dwarfed by the cold water Pacific species Haliotis rufescens. (courtesy the researchers)
"That's when their size absolutely explodes," Lindberg says. "But these north Pacific abalone are not grazers. They feed on the drift of the high-quality kelp."
And therein lies the secret to the large abalone's survival among the otters. The abalone hide in the cracks and crevices of the shallow reefs and essentially order in their dinner. Or they make their homes in the surf zone, hiding underneath the sides of boulders while waiting for the kelp to accumulate.
"When you piece together the natural history with the fossil record and the molecular phylogeny, the entire pattern makes perfect evolutionary sense," Lindberg says.
Still, there was one thing that continued to trouble Lindberg. There was no evidence that any abalone had ever lived in South America along the Chilean coast.
"That was a problem because if you look at the phylogeny and track the position of the continents, the abalone should have been there during their radiation toward the eastern North Pacific Ocean," he says.
Fortunately, Lindberg stumbled across scientific papers describing how El Niño periodically wipes out the algae and kelp beds on the Chilean coast. Armed with that information, Lindberg became convinced that abalone had in fact established populations in the area but that El Niño wiped out their food source, leading to their extinction after perhaps a few thousand years. That limited lifetime combined with the constant uplift of the tectonic plates in the region, Lindberg says, would mean that intact fossilized abalone would be few and far between.
Still, Lindberg says, "I have no doubt that at some point someone will pull an abalone out of the Chilean fossil record."
And when that day comes, Lindberg's evolutionary puzzle of the Abalone will be complete. For the time being, anyway.
Related Web Sites
Boundaries Unbounded
James Sethian is the author of Level Set Methods and Fast Marching Methods: Evolving Interfaces in Computational Geometry, Fluid Mechanics, Computer Vision, and Materials Science (Cambridge University Press, 1999).
What do inkjet printing, MRI brain scans, and microchip manufacturing have in common? They've all been improved by UC Berkeley professor James Sethian's pioneering work on the mathematics of boundaries.
"The world is filled with dynamic boundaries or interfaces where one thing is separated from another," says Sethian, who is also the head of the Applied and Computational Mathematics Department at Lawrence Berkeley National Laboratory (LBNL). "Oil and water is the most obvious example. The etching process in semiconductor manufacturing is less obvious. I'm in the business of creating mathematical tools that can help formulate and solve all sorts of problems involving dynamic interfaces."
Indeed, Sethian has built a career on looking at the subtle and striking motion of myriad interfaces, from flames dancing in a fireplace to a robot weaving around obstacles. Earlier this year, he was awarded the prestigious Norbert Wiener Prize in Applied Mathematics in honor of his "seminal work on the computer representation of the motion of curves, surfaces, interfaces, and wave fronts, and for his brilliant applications of mathematical and computational ideas to problems in science and engineering."
From the petroleum industry to Hollywood's computer animation studios, Sethian is known for his mathematical approach to battling the boundary problem. Algorithms based on his methodologies already aid oil companies in checking layers of rock for pockets of oil and inform engineers how to design more precise inkjet printers.
Time sequence of a three-dimensional axisymmetric inkjet printhead. The ink bath (at bottom) undergoes a rapid pressure change which forces ink out of the nozzle; the moving ink separates into satellite droplets as it progresses. (courtesy the researchers)
Most numerical techniques for solving "evolving interface problems" use a collection of marker points placed at various locations on the shifting boundary line. Imagine that a handful of buoys are connected by ropes on the surface of a lake. Mathematically representing those boundaries becomes incredibly difficult if the buoys cross, a sharp angle forms between two of the buoys, or multiple strings of buoys combine. Tracking this over time is even more complicated.
On the other hand, Sethian's Fast Marching Methods and Level Set Methods, the latter developed with UCLA professor Stanley Osher, work by overlaying a grid, or coordinate system, on top of the area being studied. In this higher-dimensional perspective, the value of each coordinate changes depending on its distance to the shifting boundary.
"If this were basketball, buoy methods would resemble man-to-man coverage and level set methods provide a zone defense," Sethian says.
With the basic algorithmic research already under his belt, Sethian is now focused on a growing number of applications for Level Sets and Fast Marching Methods. For several years, he has collaborated with semiconductor companies to improve the efficiency and accuracy of their manufacturing. Level Set Methods, he explains, are well-suited to model the etching and deposition processes used to carve out channels in integrated circuits and deposit insulators.
"It's art moved down to the nanometer scale," Sethian says. "It's difficult to dig long trenches, make smooth surfaces, and lay down insulators. This is coupled to equations off the surface of the wafer, in the plasma chamber, for instance, where the rate of the etching is controlled. But it's all moving boundaries."
The human body is also filled with moving boundaries. Some, like tumors, must be carefully tracked if they're to be beaten. Sethian is developing ways that his mathematical techniques can be used to automatically extract "news you can use" from the raw data of medical scans such as magnetic resonance imaging (MRI) and computer-aided tomography (CAT).
This reconstruction of the brain's cortical structure was created by applying the edge-detecting capabilities of Level Set Methods to MRI data. (courtesy the researchers)
"Right now, a physician identifies a tumor in a CATscan of the brain by drawing a line around it on the computer," Sethian says. "It's hard though to detect the noise so the doctor can throw it away and find something meaningful, like if a tumor is getting bigger or smaller."
Sethian, in collaboration with postdoctoral fellow Thomas Deschamps and LBL mathematician Ravi Malladi, have devised an automatic approach for edge detection in medical images. Previous techniques extract the outline, say, of the colon by looking for differences in intensity between neighboring pixels. However, determining the intensity value that signifies one side of a boundary versus the other is tricky. Sethian and his colleagues instead start with a small shape placed on the image in a location that's obviously within the boundaries of the colon. The shape then evolves outwards. The Fast Marching and Level Set algorithms cause the shape to locally slow its expansion at the places where individual pixel intensities begin to change. When it comes to a halt, the boundary has been defined and the colon is "filled in from the inside."
"We can then run a robotic algorithm to figure out the best way for an endoscopic probe to go through without puncturing the sides," Sethian says.
So far, the researchers have demonstrated the potential of their technique for virtual colonoscopies, vascular reconstruction, and brain scans. It's not perfect, Sethian points out, but it's fast, automated, and gets around the pockets of noise. As with all of his research, the key to improving the tools' effectiveness is continued collaboration with the people they're intended for.
"I get excited by teaching this bag of tricks to people who know their particular scientific fields much better than I do," Sethian says. "Hopefully they can then benefit from applying these general mathematical ideas to their own specialized problems."
Related Web Sites
The Evolutionary Secret of Body Segmentation
When UC Berkeley biologist Nipam Patel was searching for a new crustacean to study, one of his graduate students paid a visit to a large public aquarium. Rather than select an organism from one of the breeding tanks, the student sifted through the aquarium's filter system. There he found Parhyale hawaiensis, a sea scud. Because of its living conditions, Parhyale was already naturally selected to require minimal care, perfect for life in Patel's bustling laboratory. Patel, a professor of Integrative Biology and Molecular and Cell Biology, studies the development of arthropods to better understand evolutionary differences among a wide range of organisms.
A butterfly collector since he was a child, Nipam Patel also studies the pattern formation in butterfly wings.
"We focus on the development of crustaceans to better understand what characteristics are conserved between organisms and also the evolutionary differences," says Patel, also a researcher with the Howard Hughes Medical Institute.
Specifically, Patel's group studies how various animals segment their bodies during fetal development. Before delving into the evolutionary lineage of the scud, Patel made great strides several years ago studying genes that establish the body plan of several insects, including Drosophila melanogaster, the fruit fly. When the fly is still in an early embyronic state, a set of genes kick in that subdivide the fetus into increasingly smaller domains. Eventually those domains develop into the segments of the head, thorax, and abdomen of the adult fly.
Parhyale hawaiensis (male seen here) reach .5 to 1.0 cm in length with a generation time of seven weeks. (courtesy the researchers)
"One of the hallmarks in biology during the last couple of decades is the knowledge that the mechanisms and genes behind development seem to be quite universal in animals," Patel explains. "But of course, we don't look anything like flies. Why? How does evolution make different organisms with nearly the same set of genes?"
Recently, the researchers extended their comparative studies to include other arthropods such as grasshoppers, beetles, and lobsters. In one major breakthrough, Patel and his colleagues determined that evolutionary changes in the shape and function of certain crustacean appendages are tied to one particular gene family, known as the Hox genes.
Now he's hoping to tease out the evolutionary secrets behind Parhyale's segmentation. Patel's approach is two-pronged: determine how the animal segments and then compare its pattern-forming process to that of other arthropods. While it's clear that these animals' earliest shared ancestor was segmented, the mechanisms vary greatly between species. Patel would like to know which mechanisms and arthropod features are evolutionarily derived and which are common with the shared ancestor.
Already, the research group has made the remarkable determination that even when a Parhyale fetus is just eight cells old, those individual cells are already destined to become precursors of a particular type of tissue--muscle, for example.
At this stage in embryonic process of segmentation in Parhyale hawaiensis, the cells are organizing themselves into a precise pattern of rows and columns. The "engrailed" gene, expressed in cell rows stained red in this image, is instrumental in determining the pattern of each segment. (courtesy the researchers)
"There's no other animal where we've seen that level of restriction so early on," Patel says. "Now we'd like to know the level of that commitment."
The researchers are also analyzing the pattern of cell formation and its potential role in segmentation. The group is able to visually observe the amazingly orderly pattern in which the cells align by injecting a fetus, when it consists of just one cell, with RNA that expresses a protein that's fluorescently red. All of that single cell's progeny then glow red as well.
Another benefit of studying Parhyale as the "model system" for non-insect arthropods is that it appears well-suited to manipulation at the genetic level. By developing methods to alter gene expression during the fetal development, Patel hopes to gain insight into the similarities and, more importantly, the differences between Parhyale and its taxonomical brethren.
"Evolution is not a story of things staying the same, it's about how things change," he says.
Related Web Sites
Berkeley's Scientific Legacy
1933: William F. Giauque and the quest for absolute zero
How cold can cold get? Absolute zero, of course. But how cold is that? Beginning in the mid-18th century with Gabriel Daniel Fahrenheit's temperature experiments, scientists sought to demonstrate the theoretical temperature at which all molecular motion would stand still.
By the early 20th century, the magic number of 459.688 degrees below zero on the Fahrenheit scale (-273.15 C) was agreed upon to be absolute zero. Until 1933 though, the coldest temperature reached was 458 degrees below zero. Many scientists doubted it was possible to go any lower.
William F. Giauque (1895-1982) received his BS in Chemistry in 1920 and his PhD in 1922 from the UC Berkeley College of Chemistry.
UC Berkeley chemist William F. Giauque wasn't convinced though. During his graduate studies at Berkeley, Giauque immersed himself in low-temperature work, a line of research he continued when joining the chemistry faculty in 1922. Five years later, he proposed a new magnetic refrigeration method to achieve extremely low temperature. The process, called adiabatic demagnetization and still used today, involves removing the magnetic field from certain materials. In 1929 through low-temperature studies of oxygen, Giauque and collaborator H.L. Johnston discovered the oxygen isotopes of mass 17 and 18. But Giauque's inquiries into entropy drove him even further.
One day in 1933, after working 21 straight hours on a key experiment, Giauque made his breakthrough observation. His machine had achieved a temperature within one-tenth of a degree of absolute zero. (The current record low, hit last year, is one-half-billionth of a degree Fahrenheit above absolute zero.) Giauque's experiments confirmed the third law of thermodynamics, which states that the entropy of a pure perfect crystal is zero at absolute zero--the atoms are perfectly aligned and don't move. In 1949, he received the Nobel Prize in Chemistry for opening new vistas for chemists and physicists to study the very essence of matter.
Upon presentation of the award, a Nobel Committee member introducing Giauque said, "In order to extend our knowledge of those laws in Nature which determine the properties of matter and its transformations, it has been necessary to penetrate into the field of the lowest temperatures ever reached by man. You have created methods necessary for accurate measurements under these extreme conditions, and you have applied these methods to a precise study of previously unknown phenomena which are of the deepest significance for science. Your results have afforded the final proof of one of the most fundamental laws in Nature, a law which is also of immense practical importance."
Giauque died in 1982 but his legacy lives on at UC Berkeley's Giauque Hall, home to the University's Low Temperature Laboratory, and across all of physical science.