BERKELEY

science
review
Fall 2003 Issue 5
BERKELEY
science F ROM THE EDITOR
review
Dear Readers,
EDITOR IN CHIEF
Colin McCormick Welcome to the fifth issue of the Berkeley Science Review. Like wine and cheese, we just keep getting
better with age. We’ve expanded our print run from 5,000 to 6,000 copies. We’ve begun a high
school outreach program, and science students in local high schools will soon be reading the BSR in
MANAGING EDITOR class. We’re distributing copies to several area science museums and, as always, across the Berkeley
Jessica Palmer campus and at the Lawrence Berkeley lab. We have some great stories for this issue, from Letty
Brown’s walk through the UC Botanical Garden, to Nathan Bramall’s letter from his physics
research in Greenland, to Rupa Datta’s article on new techniques in mathematical genomics.
COPY EDITOR
Kira O’Day But all good things must come to an end, and my time at the BSR is no exception. I’m writing this
letter from Washington DC, where I just began my tenure as an AAAS Congressional Science Fellow.
Although best known for publishing Science magazine, AAAS (the American Association for the
Advancement of Science) plays an extremely important role in American science policy. Their
EDITORS
Science Fellows program each year places scientists in congressional offices and federal agencies to
Carol Hunter provide technical expertise to legislators and administrators. This is a fascinating opportunity for me
Heidi Ledford to watch law and policy being made, and perhaps to influence its direction just a little. Among other
Kira O’Day things, we’re learning how the federal government spends its $2.3 trillion budget, and where all
Christopher Weber that money comes from.…

While nobody’s accused me of lacking confidence, I know I wouldn’t have gotten this job without
ART DIRECTOR the BSR. My interviewers asked how I would feel about giving science advice in fields outside of
Una Ren physics. In response, I showed them copies of the BSR and explained that I’d been writing and
editing for a multidisciplinary popular science journal. They were impressed and I got the job,
although I’m sure Una Ren’s great cover designs also helped. I’m not the first person to parlay
W EBMASTER experience with the BSR into a job: Eran Karmon won a prestigious AAAS Mass Media fellowship,
Tony Le and Sherry Seethaler is now assistant director of science communications at UC San Diego.

If you like what you read here, or even if you don’t, drop us a line at submissions@uclink.berkeley.edu
to tell us what you think.You can read all our issues online at sciencereview.berkeley.edu, which also has
SPECIAL THANKS
information about how to get involved. We are always interested in new writers, editors, and
Keay Davidson
artists/designers. Not only is the BSR a great way to learn about the exciting science going on at
Berkeley, it just might land you somewhere that you never thought you’d be.

PRINTER Cheers,
Sundance Press

Colin McCormick

©2003 Berkeley Science Review. No part of this publication may be reproduced, stored, or transmitted in any form without express permission of the publishers. Published
with financial assistance from the Office of the Vice Chancellor of Research, Department of Nutritional Sciences and Toxicology, Physics Department, the UC Berkeley Graduate
Assembly, the Associated Students of the University of California, and the UC Berkeley College of Natural Resources. Berkeley Science Review is not an official publication of the
University of California, Berkeley, or the ASUC. The content in this publication does not necessarily reflect the views of the University or ASUC. Letters to the editor and
story proposals are encouraged and should be emailed to submissions@uclink.berkeley.edu or posted to Berkeley Science Review, 10 Eshleman Hall, Berkeley, CA 94720.
Advertisers, contact advertise@uclink.berkeley.edu or visit http://sciencereview.berkeley.edu.
 Labscope
And he changes diapers
paltry 5% of mammals
A exhibit male parental
care. But one species of lady
tuco-tucos is in luck, because
their mates share the respon-
sibility of raising a happy and
well-adjusted brood. Colo-
nial tuco-tucos (Ctenomys
sociabilis) are social subterra-
nean rodents native to South
America. After observing
that up to six females and one
Photo: Andrea Caiozzi-Cofre male may share a tunnel sys-
tem and single nest, gradu-
ate student Maria Soares and Professor Eileen Lacey of the Department of Integrative Biology wondered whether the male
was actually making himself useful. To find out, they observed a captive tuco-tuco colony at Cal, and also tracked the
comings and goings of a free-living colony in Argentine Patagonia by capturing wild tucos and fitting them with radio collars.
The researchers discovered that the males do everything for the pups that the females do, except of course for nursing. They
huddle with the pups, retrieve them when they try to leave the nest, and bring food into the tunnels. Soares believes that the
benefits of a housebound tuco male may include bigger and healthier pups and a shelter for males to hide from marauding
predatory birds between breeding seasons. And don’t underestimate the value of a little domestic harmony. Learn more
about the social behavior of tuco-tucos and other vertebrates at http://ib.berkeley.edu/labs/lacey/.

Kira O’Day

Careening cockroaches
obert Full’s laboratory is not for the squeamish. Full, the director of the
R Poly-PEDAL (Performance, Energetics, and Dynamics of Animal Locomo-
tion) laboratory in the Department of Integrative Biology, uses cockroaches to
model the mechanics of movement. In a recent study, Full equipped roaches with
tiny jetpacks that shot them sideways while they ran forward. The roaches were
filmed to determine how they stabilized their bodies and regained the correct
direction.The researchers found that the recovery time following lateral propul-
sion was a miniscule 25 milliseconds. Moreover, the cockroaches did not even
slow down, as we would if knocked off balance while running. Instead, they used
their springy legs to bounce back on track. Full theorizes that in the invertebrate
system the body can be more important than brains for balance.The Poly-PEDAL
lab is currently constructing a robot based on the filmed movements of cock-
roaches. Learn more at http://polypedal.berkeley.edu/.
Photo: Robert Full Adam Schindler
BERKELEY
science 4
review
 Vanishing polar bears
olar bears blend into their snowy
P environment, and their camou-
flage extends beyond the realm of vis-
ible light. Polar bears are also nearly in-
visible in the far infrared, the frequency
range in which bodies radiate heat—a
property that has long vexed scientists
trying to use infrared to locate bears from the air. Researchers from the De-
partment of Mechanical Engineering, led by professors Boris Rubinsky and Ralph
Greif, have now solved the long-standing mystery of the vanishing polar bears.
By measuring infrared properties one hair at a time, they determined the hair’s
emissivity, a measure of how efficiently it radiates heat. They found that the
conventional explanation, that the bears are so well-insulated that their surfaces
are the same temperature as the snow,
is correct but incomplete: the hair’s
emissivity in the infrared is also nearly
equal to that of snow, and lower than
that of other arctic animals, such as cari-
bou.They also determined that this low
emissivity could help to insulate the
bears by lowering the amount of infra-
red heat that the bears radiate away.
Photo: Dr.Ralph Nelson, Carle Foundation ChrisWeber

Giants in the mist

he fog drifts eerily through the towering redwood forests during the

T dry, hot California summer. A group of Berkeley researchers led by

Integrative Biology professor Todd Dawson has set out to “demistify”
this breathtaking sight. Like so many hikers and nature-lovers, biologists
have long wondered how these giants attain their incredible size in the face
of California’s long dry summers. While studying the impact of fog and cli-
matic variation on the giant redwood Sequoia sempivirens, the Dawson group
uncovered the redwood’s unique solution to the lack of summer rain, and
potentially their secret to attaining great heights. Following the distinctive
stable isotope fingerprints of rainwater and fogwater, the Dawson group
showed that the trees acquire water from rain in the winter and from fog in
the summer. The fog is “stripped” from the air by the redwood foliage. Fur-
thermore, the trees are so good at pulling in fogwater that the excess sum-
mer redwood “fogdrip” accounts for a third of the total annual water input
into the coastal forests—a massive natural sprinkler system! Learn more
about redwoods and plant ecophysiology at http://
ib.berkeley.edu/labs/dawson.
Delphine Farmer Photo: ©Vladimir Dinets v0v04ka@hotmail.com
BERKELEY
science 5
review
Current Briefs
GENE HUNTERS
Sifting through
evolution’s shadows
team of Lawrence Berkeley
A Laboratory and UC Berkeley
scientists, originally formed to
research heart disease, has created a new
tool to probe the human genome by
comparing it to the genomes of its siblings
and cousins—apes, Old World monkeys,
and New World monkeys. This new
technique, called phylogenetic shadowing,
will help scientists find genes in humans
that are primate-specific and therefore
overlooked by previous methods. The
scientists—Eddie Rubin, director of LBL’s
Genomics Division, Dario Boffelli, a staff
scientist at LBL with a traditional back-
ground in biochemistry, and Lior Pachter,
a UC Berkeley professor of mathematics—
typify the increasingly close collaboration
between lab-bench biochemistry and
mathematical analysis. The research team is
hoping that phylogenetic shadowing will
extend this kind of collaboration from gene
hunting to evolutionary biology.
Scientists have been comparing small
sections of genetic sequences from
related species to each other almost
since Watson and Crick first postulated
DNA’s double-helix structure. But
many dream of comparing entire
genomes, searching through all three
billion letters quickly and efficiently for
functional elements. Gene hunting on
such a large and automated scale
requires sophisticated mathematics and
complex computational tricks.
The DNA of most complex organisms is
full of “junk”—millions of base pairs (the
A’s, C’s, G’s, and T’s that are DNA’s
simply because it hasn’t had
time to mutate, so similarity
won’t necessarily mean
functionality. “You’ll never
figure out what’s important—
it’s an impractical way to find
genes,” says Boffelli. “There’s
too much background noise.”
However, when you compare
sequences between extremely
distant relatives—like humans
and fugu fish—Boffelli says,
“anything that is conserved is
functional, but few such things
are found, a lot of interesting
You can imagine the set of related genetic sequences as strings
lined up one above the other, with beads marking the spots stuff is not.” It turns out that a
where the genetic code differs from the sequence in human. A mouse is a close enough
light shining down from above through the beads would cast a
shadow on the ground below them, indicating the “junk” se- relative to humans that most of
quences. Where there is no shadow, there is very likely a gene the functional genes are still
or significant genetic information. The Berkeley team took this
one step further, comparing the sequences not only to human,
there, but it’s far enough away
but to each other closely related primate. “Mathematically we on the evolutionary tree that
made this robust,” says Pachter. “We came up with a good way
of defining the shadows.” (Image: Jeff Hunter)
junk DNA has had time to
mutate. According to Boffelli,
“alphabet”) that don’t really seem to do that is the reason the mouse genome is so
anything useful. One way to shake the popular for genetic comparisons. “It’s a
useful genetic elements out of the junk is good compromise point.”
to compare a genome to the sequence of a
related species. Bits of the sequence that But there’s still a great deal of difference
match up between the two species are between mouse and man. Comparisons
more likely to be something functional. with the mouse genome will miss any
These could be sequences called exons recently evolved gene that is unique to
that code for specific proteins, sequences primates. Phylogenetic shadowing is a way
that interact with cellular machinery to to take advantage of all the variation in the
regulate the expression of those exons, or primate family tree to find these primate-
some other functional piece of code that specific genes. Instead of comparing the
we don’t yet understand. This is because, human genome and one other species of
as a species evolves, natural selection will primate at a time, it compares the human
tend to preserve genes and other func- genome and several closely related species
tional bits, but won’t exert strong of primate to each other at the same time.
preservative pressure on the junk, Sequence conserved between any two
allowing it to mutate more rapidly. species is likely to be junk, but sequence
conserved within the whole group is
But comparing species that are too closely probably something functional. “If you
related, like humans and monkeys, will sampled enough species, the total accumu-
show too much meaningless similarity. A lative variation is enough to see what’s
lot of the junk DNA will be the same functional and what’s not,” Boffelli says.
BERKELEY
science 6
review
The LBL/Berkeley team chose four
specific sections of DNA, each about
1,500 base pairs long and containing an
exon they knew existed, to test this idea.
The team needed to see if phylogenetic
shadowing could highlight the known exon
blindly in each case. They picked 17
species distributed among the three
branches of primates closely related to
humans for the comparisons. Unlike
humans, none of these primates had been
fully sequenced, so Boffelli and his
colleagues had to sequence the unknown
sections themselves using genetic material
from the Coriell Cell Repositories and the
San Diego Zoo.
Using other primate sequence data, the
team estimated that the noncoding regions
mutate seven times faster than coding
regions. Pachter and his colleagues then
created a tree—a mathematical estimate of
how long ago each species of primate
evolved from their common ancestors.
This tree was based only on the sequence
information, but ended up basically
matching the traditional phylogenetic tree
based on the fossil record and morphology.
Pachter and his colleagues used a com-
puter program to align the sequences for
each of the four test regions. Using the
tree, the team calculated the likelihood
that each little bit of sequence from the
multiple-species alignment mutated at the
fast, noncoding rate versus the likelihood
that it mutated at the slow, coding rate.
Comparing this mutation rate base pair by
base pair clearly revealed small valleys—
sections of DNA that had mutated very
slowly across all the species—amid steep
mountains of heavily diverged sequence.
The exons were found, as expected, in the
slowly mutating valleys.
In all four cases, phylogenetic shadowing gene hunting in humans. Phylogenetic
found the buried exon using the sequence shadowing also combines the interests of
from only eight species. In fact, research- biomedicine and evolutionary biology and
ers could have found three of these exons creates all kinds of opportunities for
simply by comparing mouse and human. collaboration. “My goal,” says Pachter, “is
But they would never have found the to be able to understand the evolutionary
fourth exon that way, because it is inactive history of the genome.” Understanding the
in mice. genetic threads that tie us to our primate
cousins is a big first step.
Once they had proven the technique in the
four test cases, it was time to test phyloge-
Saheli S. R. Datta
netic shadowing’s ability to find new
functional sequences. The LBL scientists
were interested in finding the regulatory
elements for another primate-specific Want to know more?
gene called apo(a), which codes for a
blood protein that, at high levels, can be a Phylogenetic Shadowing of
major risk factor for youthful heart Primate Sequences to Find
disease. The scientists hoped to use their Functional Regions of the Hu-
new technique to discover these unknown man Genome. D Boffelli et al.,
pieces of code. Science (2003); Vol. 299,
pages 1391–1394.
Using phylogenetic shadowing, Boffelli
and his colleagues were able to pinpoint
eight conserved regions of 40–70 base
pairs and show that these bits interacted
with regulatory DNA-binding proteins
more than other segments. This discovery
provides important new targets of study
to cardiologists trying to understand why
some people have higher levels of apo(a)
in their blood—and a correspondingly
higher chance of having a heart attack in
their thirties.
The team is now applying phylogenetic
shadowing to large sections of the human
genome, trying to find as many primate-
specific genes as quickly as possible. Their
study is an important contribution to the
debate over which primates should be
sequenced next, presenting scientists with
a mathematical basis for picking those
primates that will prove most useful for
BERKELEY
science 7
review
 Current Briefs
A LIVING CARBON RESERVOIR
Sequestering carbon in the sea
o matter how much some
N politicians may try to deny it, the
evidence for global warming is
overwhelming.Yet while we humans are
busy pumping heat-trapping carbon into
the atmosphere, our oceans are teeming
with microscopic algae that consume
carbon during photosynthesis. Scientists
and politicians alike have begun to wonder
if it might be possible to harness the
earth’s oceans as a carbon reservoir. James
Bishop of UC Berkeley’s Department of
Earth and Planetary Science and the
Lawrence Berkeley National Laboratory is
deploying remote-controlled robots to
find out if stimulating phytoplankton
growth could reduce the amount of
carbon in the earth’s atmosphere.
Marine phytoplankton are the primary
producers of the ocean. As phytoplankton
photosynthesize, they convert dissolved
inorganic carbon into organic carbon,
thereby sequestering it within their cells.
Some phytoplankton inevitably sink deep
into the ocean, taking their particulate
organic carbon (POC) with them. POC
lost from the ocean surface is replaced by
carbon from the atmosphere, and the
whole system forms a biological carbon
pump that lowers the amount of CO2 in
the atmosphere.
Scientists have long suspected that they
could increase the flux through the pump
by increasing phytoplankton activity. The
question was, how could phytoplankton
productivity be enhanced on a large scale?
The answer resided within the Pacific
Ocean, in regions that are famous for
being high in nutrients such as nitrogen
and phosphorous, but mysteriously low in
Collecting data in the stormy South Pacific is dangerous
work, so Berkeley scientists are deploying remote-controlled
robots that can continuously monitor carbon levels for
months at a time. (Photo: LBNL)
phytoplankton. For years, scientists
debated the nutrients that might be
limiting phytoplankton growth in these
areas. In the late 1990s, a ship carrying
ten tons of iron sulfate sailed to a high
nutrient/low phytoplankton region in the
South Pacific, south of Tasmania. There,
scientists dumped their cargo of iron into
the sea, triggering a massive algal bloom a
few days later. Space satellites that
monitor chlorophyll fluorescence could
detect the bloom for almost two months
following fertilization. The answer to
ocean fertilization, it seemed, was iron.
But this experiment was only the begin-
ning, and much more must be learned
about the ocean’s carbon cycle before we
begin filling our oceans with iron. In the
past, data collection was limited by a
reliance upon manned research voyages
that were prohibitively expensive and
often dangerous. Furthermore, each
voyage could last only a few months at a
time, and could easily miss sporadic
natural events. Bishop and his team
decided it was time to switch tactics and
BERKELEY
science 8
review
designed remote-controlled robots called
Carbon Explorers.
The neutrally buoyant Carbon Explorers can
move from the surface of the ocean to depths
of 1,000 meters, measuring POC concentra-
tions along the way. Active for the greater
part of a year, these robots can monitor the
ocean continuously and quickly relay their
findings to researchers via satellite.They send
frequent emails to Bishop, who admits, “It’s
like being addicted to a computer program.”
The Carbon Explorers are also incredibly
cost effective: building one Carbon Explorer
and operating it for many months costs about
$25,000—that’s roughly equivalent to
operating an oceanographic ship for one day.
For their first deployment, two Carbon
Explorers were released in a high nutri-
ent/low phytoplankton region of the
North Pacific, approximately 1,000 miles
west of Vancouver Island, British Colum-
bia. On April 7, 2001, NASA satellites
detected a large dust storm originating
near the Gobi desert. The powerful storm
churned up debris over land and then
traveled out over the Pacific Ocean,
depositing iron-rich dust in its wake.
When the dust storm passed over the
Carbon Explorers on April 12, high winds
and seas prevented the Carbon Explorers
from communicating with satellites.
A few days later, the storm had died down
and the Carbon Explorers were again
transmitting data. Although communica-
tions were temporarily interrupted during
the storm, data collection had continued
smoothly. Five days after the storm passed
overhead, the robots detected large
increases in POC. Several days later, POC
levels hit their peak, at nearly double the
pre-storm levels.
On their first mission out, Bishop’s
Carbon Explorers had made a tremen-
dous discovery. They had detected a
completely natural iron fertilization
experiment, and provided the first direct
evidence for ocean fertilization by natural
storms—a phenomenon that would have
been very difficult to observe by manned
research voyages.
While these results make iron fertilization
seem like an easy solution to global
warming, Bishop makes it clear that this
field is “still in the realm of science” and
not yet at the level of practical application.
In 2002, the Carbon Explorers were
involved in iron fertilization experiments
in the South Pacific. These experiments
showed that adding iron can increase
phytoplankton productivity, but the
response is not always consistent. “The
results demonstrate that there is a lot to
learn yet from the ocean,” says Bishop.
Many aspects of the ocean’s carbon cycle
are unknown, and whether this method
would be effective on a large scale
remains an open question. Furthermore,
no one knows what consequences global
TWO WAYS TO SIX LEGS
Creepy crawlies will never be the same
t turns out that not every six-legged
I creepy crawly is an insect. Lawrence
Berkeley Laboratory and University of
Siena researchers have made the unex-
pected discovery that collembolans, a type
of primitive arthropod long thought to be
closely related to insects, are actually only
distant relatives.
Collembolans are commonly referred to
as springtails, due to an appendage on
their abdomen that propels them
through the air. They are small, wing-
less, and harmless, although they can
seem alarming when they mob moist
places like floor drains and damp
basements. Collembolans and insects are
traditionally grouped together in the
climate engineering may hold. While
there may be good outcomes, negative
repercussions for ocean ecology are easy
to imagine. Algal blooms have a terrible
reputation, and sometimes trigger
massive fish kills and perturbations in
the food chain. “Ocean fertilization will
be a global decision,” says Bishop, “after
a lot of people think of the conse-
quences and outcomes.”
But the consequences cannot be fully
understood until scientists discover
more about the ocean’s carbon cycle.
Bishop’s Carbon Explorers are still out
there, now exploring the North
Atlantic, the North Pacific, and the
Southern Ocean, and investigating
other aspects of phytoplankton produc-
tivity and the ocean’s carbon cycle. “We
are still trying to understand how the
ocean behaves naturally,” says Bishop.
“Once we have a predictive basis, we
will be able to predict the outcome of
an ocean fertilization experiment.” And
subphylum Hexapoda (“six feet”). Like
crustaceans, spiders, and many-legged
creatures such as centipedes and
millipedes, hexapods are members of
the phylum Arthropoda. Hexapods are
characterized by the presence of three
main body regions (head, thorax, and
abdomen), a pair of antennae, and, of
course, three pairs of legs.
Scientists have found it difficult to
determine how hexapods are evolution-
arily related to other types of arthropods.
For many years, hexapods were thought to
have diverged most recently from centi-
pedes and millipedes. But recent genetic
and developmental analyses show that
hexapods are actually more closely related
BERKELEY
science 9
review
even if ocean fertilization is ruled out,
Bishop’s experiments provide the basis
for understanding how the natural
carbon cycle will be affected as the
oceans respond to climate change. “Each
time the floats go out,” says Bishop,
“they find something different. And
that’s very exciting.”
Jennifer Skene
Want to know more?
Robotic obser vations of dust
storm enhancement of carbon
biomass in the North Pacific. JKB
Bishop et al., Science (2002);
Vol. 298, pages 817–821.
www-esd.lbl.gov/CLIMATE/
to crustaceans such as shrimp and
lobsters. Still, it was believed that all
hexapods developed from a single
common ancestor. However, recent work
by LBL scientist Jeffrey Boore and
University of Siena collaborator Francesco
Nardi challenges the common-ancestor
model, and provides strong evidence that
acquisition of the hexapod body plan
occurred independently in the insects and
the collembolans. In fact, genomic analysis
of mitochondria (cellular organelles that
contain their own DNA) reveals that
collembolans may have split from the
future insect lineage even before insects
split from crustaceans.
Collembolans are an ancient group of
arthropods, dating back 400 million years.
To the untrained eye, they resemble
wingless insects. But collembolans have
only six abdominal segments whereas
“true” insects have 11. Previous work by
 Current Briefs
Left, Tetrodontophora bielanensis, a collembolan. Right, Thermobia domestica, an insect commonly known as a firebrat.
Although both of these arthropods have six legs and a segmented body, they are not as closely related as scientists once thought. demonstrating what we
(Tetrodontophora photo: Francesco Nardi, University of Siena; Thermobia photo: David Maddison, University of Arizona,
www.tolweb.org)
Francesco Nardi and colleagues placed
collembolans far from insects on the
evolutionary tree, but mitochondrial DNA and collembolan mitochondrial genome
sequences from more species of hexapods
were needed to confirm their preliminary
results. So, in collaboration with Jeffrey
Boore, who heads up the evolutionary
genetics laboratory located at LBL, they
sequenced two more mitochondrial
genomes, one from an additional collem-
bolan, the other from a zygentoman (an
insect order that includes silverfish and
firebrats). According to Nardi, the
zygentomans are “one of the most basal
lineages of the ‘true’ insects.”
Mitochondrial genomic analysis offers
Nardi and his colleagues several advan-
tages over traditional nuclear genomic
analysis when teasing apart evolutionary
relationships. Mitochondria, often called
the “powerhouses” of the cell, are thought
to have descended from a bacterium that
entered into a symbiotic relationship with
a primitive eukaryotic cell. Mitochondrial
genomes are relatively easy to isolate, and
they are much smaller than nuclear
genomes. Mitochondrial genomes are
always inherited maternally and are highly
conserved among animals. According to
Boore, “the actual sequencing of a
mitochondrial genome requires less than
three minutes.” However, he adds, “the
work required to prepare products for
sequencing and analyze results typically
takes a few months.”
tuna and sharks, and
mammals such as
dolphins and whales. The
wings of insects, the
wings of birds, and even
the wing-like structures
of mammals such as bats
are another example of
convergent evolution,
already know—wings
are well-designed for
flying. So what makes having six legs so
The researchers used a series of computa- desirable? One way to address this
tional programs to analyze several insect question would be to determine what
ancestral body plans might have predated
sequences. Their analysis compared both the collembolans and the insects.
several specific sequences of conserved
genes that code for proteins. Changes The convergent evolution of hexapods
between specific sequences in the poses some important questions: What
mitochondrial genomes of the different molecular mechanisms are responsible for
hexapods were used to place them relative the similar forms seen in these two
to one another in an evolutionary sense. independent hexapod lineages? Do
collembolans and insects have develop-
The data Boore and his colleagues gleaned mental “tools” in common, or are largely
from mitochondrial genomic analysis separate sets of genes responsible for the
indicates that although “true” insects are six-legged body plan? Perhaps as Boore,
descended from a common ancestor, Nardi, and their colleagues collect more
hexapods as a whole are not. Rather, as data on collembolans and primitive
Nardi explains, “Hexapoda, as commonly insects, answers will emerge.
defined to include all six-legged creatures,
is paraphyletic [descended from more
than one common ancestor], with Russell Fletcher
Collembola originating earlier in the
evolution of arthropods, possibly before
the split of crustaceans and insects. This
implies an independent evolution to life Want to know more?
on land, and the convergent acquisition by
collembolans and insects of many charac- Hexapod origins: monophyl-
ters, with six-leggedness among others.” etic or paraphyletic? F Nardi
et al., Science (2003); Vol. 299,
The independent evolution of six- pages 1887–1889.
leggedness on at least two occasions
demonstrates the remarkable utility of the Comment on “Hexapod
hexapod body plan. The convergent Origins: Monophyletic or
acquisition of useful body plans and Paraphyletic?” F Delsuc et
al., Science (2003); Vol. 301,
specific body parts is certainly not
page 1482d.
uncommon; fins and flippers were
independently acquired in fish such as
BERKELEY
science 10
review

TRACKING BIOTERROR
A swell way to spot anthrax
othing lives forever. Certain
N species of bacteria, however,
make a pretty good attempt at
it. When available nutrients drop to
dangerously low levels, these bacteria
respond by forming spores that can
survive extreme environmental
conditions and remain viable for
hundreds of thousands of years. Recently,
the public saw this robustness graphically
demonstrated when processed spores of
B. anthracis were used as bioterrorism
agents—sent through the mail as a dry
powder fully capable of infecting a
susceptible host. Anthrax was first linked
to the B. anthracis bacterium in the late
1800s and has been a subject of interest
to microbiologists ever since. But the
attacks in 2001 drew the attention of
scientists from a wide range of other
disciplines to the bacterium, both as a
threat to national security and public
health and as an object of scientific
interest. One of these scientists was Dr.
Andrew Westphal, a physicist at UC
Berkeley’s Space Sciences Laboratory,
who became interested in looking for
physical properties that might distinguish
the spores of B. anthracis from those of
other Bacillus strains.
Dr. Westphal’s lab has a lot of experi-
ence measuring tiny things. About 10
years ago, they developed a technique to
make highly precise measurements using
a standard optical microscope attached
to a video camera. The camera takes
hundreds of thousands of snapshots of
the object, each slightly offset from the
others. By averaging these offset images
and accounting for the spatial variation
of the camera shots, Westphal is able to
achieve a precision of measurement
better than 5 nanometers—or 0.5 % of
Electron microscope cross section of a spore of Bacillus subtilis,
showing the cortex and coat layers surrounding the core (dark
central area). The spore is 1.2 microns across, a fraction of the Bacillus strains based on
width of a human hair. (Photo: S. Pankratz)
one micron. Westphal had been using the
technique to look at the tracks of heavy
ion particles from outer space in special
pieces of glass. “It seems like a strange
departure for us, to be using [our
technique] to look at anthrax spores,” Dr.
Westphal acknowledges.
Current testing methods for anthrax in a
suspicious sample rely on biological
characteristics—using antibodies that bind
to proteins in the spore coat or
polymerase chain reaction (PCR)-based
techniques to identify anthrax-specific
DNA sequences. These are fairly reliable,
but they take time—between 12 and 24
hours—and some of the components,
such as the antibodies, can be expensive
and perishable.
Noting the difficulties involved in
positively identifying the tiny anthrax
spores—just two to six microns in
diameter, a fraction of the width of a
human hair—and in separating them from
other, innocuous materials, Westphal’s lab
decided to see if there was anything that
their technique could contribute. What
they hoped to do was to develop a system
that could identify the spores rapidly,
based strictly on physical characteristics.
“We thought it would be interesting to put
some spores [under the microscope] and
BERKELEY
science 11
review
see what we see, if we can
distinguish them based on
size,” he said. So they got
some spores of harmless
relatives to B. anthracis from
Dr. Terry Leighton, a former
Cal professor now at
Children’s Hospital Oakland
Research Institute, and set
about doing just that.
The lab first tried to see if
they could detect different
spore size alone. They tested
four different types of the
bacteria—B. cereus, which can cause food
poisoning; B. thuringiensis, which is used to
kill agricultural insect pests; and two
harmless soil bacteria, B. subtilis and B.
megaterium—and found that, indeed, each
type differed significantly in size. But their
most intriguing discovery came about
when the lab tested several parameters to
identify possible sources of error in their
readings, conditions that they thought
might change the size of the spores during
the experiment.
One of the conditions they tested was
relative humidity, chosen, Dr. Westphal
says, “rather naively,” because they didn’t
know that humidity was not expected to
affect the size of the spores. Bacterial
spores have long been considered to be
essentially dormant, their coats acting as
solid, impenetrable shells to protect the
DNA from damage. Microbiologists
thought the bacterial spores would not
show any major morphological changes
until they were introduced to a nutrient-
rich environment, where they would
begin to germinate.
However, to the surprise of his
collaborators, Westphal saw that increased
humidity caused the spores to swell in two
distinct, time-dependent stages. The data
suggest that the immediate, rapid stage of
 Current Briefs
swelling could be due to the diffusion of
water into the outer coat and cortex of the
spore, while the second, slower stage
(occurring about eight minutes after the
change in humidity) is likely due to the
penetration of water into the spore core.
Altogether, the spores swelled by an
average of approximately 4%—much too
small to be measured by standard optical
microscopes.
The materials that make up the spore
coat should not take up water from the
air, so the data suggest that this change in
size is due to an active process within the
spore. This raises some interesting
biological questions, such as how these
changes are initiated and carried out and
if they have anything to do with the
spore’s eventual emergence from
dormancy. It may also have an application
in the design of a detection system: if
different spores can be shown to swell
with different amplitudes or time or proposed. Dr. Westphal believes that
constants, it might be a way to confirm one could be made that would be no
the identity of a suspicious sample. bigger than a standard cooler, at a cost that
would be competitive with other
Currently, Westphal’s lab is looking at detection systems. His hope is that such a
different types of spores to determine if device could prove to be very useful, both
size and swelling are species-dependent. in alerting authorities to dangerous
Their eventual goal is to create an samples and limiting the expense and
apparatus that will be able to identify disruption of false alarms.
anthrax spores on the order of seconds—
far faster than any other system, existing Marjorie James
Want to know more?
Kinetics of size changes of individual Bacillus thuringiensis spores
in response to changes in relative humidity. AJ Westphal et al.,
Proceedings of the National Academy of Sciences (2003); Vol. 100, pages
3461–3466.
Resistance of Bacillus Endospores to Extreme Terrestrial and Ex-
traterrestrial Environments. WL Nicholson et al., Microbiology and
Molecular Biology Reviews (2000); Vol. 64, pages 548–572.
BERKELEY
science 12
review
Policy
SCIENCE FINDS ITS PLOS
IN THE SUN
Is open-access publishing
coming of age?
Ben Gutman
cientists are often perceived to be at
S the cutting edge of technology, but
in one critical area they are stuck in
the past. In the midst of the Information
Age, when the Internet can rapidly deliver
free information around the globe, scien-
tists are still communicating with each
other through closed, subscription-based
journals. These expensive subscriptions
make scientific research available only to
those who can afford it, often preventing
scientists from seeing the complete body
of published works in their field. Or so say
the founders of the Bay Area-based Public
Library of Science (PLoS).This fall PLoS is
taking an important step in its campaign
to erase obstacles to the free and easy dis-
tribution of science: armed with initial
funds of $9 million, it is launching a new
online scientific journal intended to com-
pete with top-tier scientific journals—and
maybe even start a new trend in open-ac-
cess publishing.
Lawrence Berkeley Lab scientist and Ber-
keley MCB professor Michael Eisen is one
of the cofounders of PLoS. He describes it
as “a group of scientists who believe that
science in general would be better off if
the published results of scientific research
were fully available.” He also believes the
Internet has enormous potential for ad-
vancing scientific research. Most journals
now offer their articles online as well as in
tribute derivative works, or to use them
for any other lawful purpose.” Eisen sees a
day when scientists will be able
to “search the entire text of jour-
nal articles, [electronically] link
to specific subsections, [and use
the literature] in ways we
haven’t even thought of yet.”
Eisen compares his vision of
open-access research to public
DNA-sequence databases that
collate DNA codes discovered
by various researchers, explain-
ing how the ability to systemati-
cally view whole genomes has
revolutionized biology. If se-
For PLoS to succeed, it will have to be embraced
quences were still published
piecemeal in various journals,
new fields such as genomics and
Eisen’s own field of computa-
by publishing scientists. (Image: www.PLoS.org)
tional biology might not exist.
Computational techniques employed daily,
print to paying subscribers, and are rapidly such as BLAST and homology searches,
scanning in back issues (articles in Journal would never have been developed, he says.
of the American Chemical Society are available These tools, which allow biologists to
online back to 1879). Berkeley recently search and compare genes and proteins
canceled print subscriptions to journals of- across species and even kingdoms, eluci-
date everything from the function of the
Realizing that presenting the idea gene and genome to the pattern of evolu-
of open access and even mak- tion. In the scientific literature Eisen sees a
similar trove of information to be mined
ing a threat didn’t reshape the by enterprising scientists in ways as yet
world of scientific publishing, unconceived, if only all the literature were
Eisen and company undertook available in a single, easily manipulated site.
He and his PLoS colleagues have set out to
a new approach.
make that possible.
This task has turned out to be easier said
fered by major scientific publisher Elsevier,
than done. Eisen first came upon the idea
choosing to rely on electronic versions in-
for a universal scientific literature database
stead. But Eisen envisions more than sim-
as a post doc some years ago with his then
ply making research available to be read
boss, now PloS cofounder, Dr. Patrick
online by paying subscribers. According to
Brown of Stanford University. Eisen was
the PLoS website, “open access” to scien-
working on microarrays, a biological tool
tific literature means “its free availability on
used to study the activation of genes under
the public Internet, permitting any users
various conditions. Huge amounts of prior
to read, download, copy, distribute, print,
data were require to understand the results
search, or link to the full texts of these ar-
generated by this technique. Says Eisen, “Pat
ticles, to crawl them for indexing, to pass
Brown has a photographic memory, but it’s
them as data to software, to make and dis-
BERKELEY
science 13
review
Policy
Professor Michael Eisen is one of the cofunders of PloS. (Photo: www.PLoS.org)
just not possible to know all the things you
need to know. It just seemed logical to have
all the information to interpret [the
microarrays] in one place.” Biologists have
been publishing research on the function
of specific genes and proteins for years, but
there was, and still is, no easy way to search
through all of that information for general
patterns.
Recognizing this need and the potential of
the Internet, Eisen, Brown, and the third
PLoS cofounder, Harold Varmus, Nobel
Laureate and then director of the National
Institutes of Health, set to work. The first
attempt was the 1999 creation, by Varmus’s
initiative at NIH, of PubMed Central—a
free online repository for life-science pub-
lications. They hoped to convince publish-
ers to voluntarily submit articles to
PubMed Central some time after publica-
tion. The delay would allow publishers to
continue to profit by charging customers
for newly published material, but once the
research was somewhat dated, it would
become universally accessible. With a few
exceptions, notably The Proceedings of the
National Academy of Sciences (PNAS), this was
not terribly successful.
The next approach was to formally create
PLoS, and draft an open letter to scientific
publishers. The letter, which Eisen terms
“free market research,” outlined the broad
desire among the scientific community for
open access. It was widely termed a boy-
cott, and was relatively unsuccessful. Signed
by over 30,000 scientists from 180 coun-
tries, the letter asserted that the signers
would not subscribe to, edit, or publish in
journals that did not provide their content
to a centralized repository such as PubMed
Central within six months of publication.
To their surprise, the journals disregarded
these threats, and the boycott fell apart be-
cause the signing scientists couldn’t afford
not to publish.
To hear Eisen tell it the journals were be-
ing obtuse, failing to grasp the obviousness
of the new approach. He claims that the
journals would not have lost subscribers
and could have continued in much the
present model. He points to the physics
community, which for years has posted pre-
prints on the website arXiv.org without
destroying the physics journals. Dr. Nicho-
las Cozzarelli, a professor in the Berkeley
Department of Molecular and Cell Biology
and editor in chief of PNAS, generally
agrees. PNAS follows a hybrid model, he
explains, making content “freely available
after six months, and immediately available
to subscribers and in third-world coun-
tries.” However, he adds, “We are a break-
BERKELEY
science 14
review
even journal. We don’t make a profit, nor
incur losses, at least hopefully.” Other jour-
nals concerned with making a profit may
not be willing to take a similar risk.
Realizing that presenting the idea of open
access and even making a threat didn’t re-
shape the world of scientific publishing,
Eisen and company undertook a new ap-
proach. They decided to demonstrate the
feasibility of open access and of a funda-
mentally new business model for scientific
publishing. They would create their own
journals operating in an open-access, au-
thor-funded system, aiming to compete
with top-tier journals such as Science, Na-
ture, and Cell and demonstrate that this
model is economically viable.They hope to
ultimately convince all of scientific publish-
ing to follow their lead.This may be a more
effective approach. “If this is a great idea,
lots of people will consider it,” says Don
Kennedy, editor of Science. “Every time
something clever succeeds you look at it.”
Cozzarelli agrees: “I think this is a great idea.
It’s obvious that this is the way to go. The
question is … how to make the transi-
tion.… It is difficult for existing journals
to try this model … [but] if they succeed
many journals will follow, and many will
not.” The first PLoS journal, premiering
this October, will be called PLoS Biology, and
is backed by a $9 million grant from the
newly created Gordon and Betty Moore
Foundation (Gordon Moore of Intel and
Moore’s Law fame is a Berkeley graduate).
PLoS has also secured the editorial services
of Vivian Siegel, the well-known former
editor of Cell.
What is different about the PLoS journals
is that authors will be charged to publish
an article. Eisen says this will reverse the
current outdated business model of scien-
tific publishing, whereby the costs are borne
by the consumer of scientific information.
If publishing is done on the Internet, costs
are fixed, and extra copies cost no more to
produce.The research can therefore be dis-
tributed for free, reaching the broadest pos-
sible audience, and the fixed cost is borne
by the producer.
But this author fee also poses potential
problems.While journals in some fields al-
ready charge authors to publish (as well as
charging for subscriptions), the proposed
$1,500 fee is two to five times higher than
standard. In biology it is new; few biology
journals currently charge authors to pub-
lish. Cozzarelli, who sits on the editorial
board of PLoS Biology, says the fee could
make things difficult for authors if they
don’t receive additional funding. “If, as
PLoS hopes, the money currently
given to libraries by institutions and
granting agencies, is instead given
to authors, then the model will
likely succeed,” he says. “If not,
then something is lost.”
Will funding agencies agree to
make publication fees a standard
part of scientific grants? In fields
in which the author-funded sys-
tem already exists, grants tend
to include publication costs. A set
of journals published in Britain by
BioMed Central (BMC) is already
based on an author-funded, open-access
model, but the difference, apart from
BMC being for profit and PLoS not,
seems to be in their reach. PLoS, accord-
ing to Eisen, intends to lead from the top.
A lofty goal, admits Cozzarelli: “I think
it would be a great success if PLoS were
as reputable as The Journal of Biological
Chemistry.” (JBC is a high-quality journal,
but not the most prestigious.) BMC is
taking a more grassroots approach, es-
tablishing a broad stable of mostly niche
publications, from BMC Ecology and Jour-
nal of Biology to Filaria Journal and BMC
Palliative Care. Perhaps both approaches
will be necessary to convince other jour-
nals of the feasibility of open access.
Eisen is confident that PLoS Biology will
succeed in attracting top-notch research.
He expects to have “about a dozen out-
standing papers” in October’s premier is- eral qualms raised even by supporters. Rob-
sue. ert Tjian, editorial-board member of Cell
and professor in the MCB department,
Kennedy of Science agrees that PLoS Biol-
highlights an interesting concern: if pub-
ogy “might very well become a serious
lishers are being paid per article there is an
competitor. We welcome them into the
incentive to publish more articles. As he
business.” However, he is not rushing to
says, “The biggest challenge is to make sure
adopt the open-access model, much to
that there are enough filters so that not
Eisen’s dismay. Eisen feels that Science’s
everything gets in. The other way to deal
publisher, the American Association for the
with that is to charge more per paper, but
Advancement of Science (AAAS), should
that gets back to limiting access.” On the
lead the charge toward open-access pub-
whole, however, he is supportive of the
lishing. He points out that AAAS is a not-
experiment, and while he doesn’t know
for-profit publisher
whether Cell will eventually follow suit, he,
along with Cozzarelli, puts the final judg-
ment in the hands of young scientists. “The
real clientele,” says Tjian, “is the students
and post docs. If [PLoS] can convince them
that it’s better to publish in the new elec-
tronic journals than in Science and Nature
then they’ll have succeeded. But right now
I’m not sure what would make a student
do that.”
Will PLoS Biology succeed? With its con-
siderable bankroll and all-star cast, it is
off to a strong start. But PLoS will truly
be a success only when all scientific re-
Image: www.PLoS.org
search is freely available to all scientists,
without the barriers of subscription fees
with a well-established, successful journal, and fragmented, unconnected informa-
and the stated mission “To advance science tion. While that goal is still far off, open
and innovation through the world for the access may one day change the way that
benefit of all people … [and] foster com- science is done.
munication among scientists, engineers
and the public.” For his part, Kennedy says,
“It’s a mystery to me why [Eisen] is so par-
ticularly critical of Science.” He explains
Want to know more?
that Science is very different from its com-
mercial competitors, already giving away
visit www.plos.org
much free content, and denies that Science
www.sciencemag.org
is a profitable magazine, since members
www.nature.com
of AAAS receive subscriptions as part of www.biomedcentral.com
their dues. “Subscriptions and ad revenue
may exceed costs, or may not,” he notes,
since it “depends on how you cost ac-
count every subscriber.”
That open-access publishing is still very Ben Gutman is a graduate student in the
much an experiment is made clear by sev- Department of Plant and Microbial Biology.
BERKELEY
science 15
review
 Profile
GOLDBERG VARIATIONS
At home in two
cultures
Jessica Marshall
lace a rectangular object—your
P wallet, a small book—flat on a
table in any orientation. Put your
hands on either side of it with your palms
turned inward. Now close your hands like
elevator doors on the object, allowing it
to rotate and align with your hands. At this
point you are touching either the short or
the long sides of the object. Next, release
it, rotate your hands 45 degrees and close
them on the object again. No matter what
its initial position, your hands are now on
the object’s long sides. If you were a ro-
bot, you’d be ready to pick it up.
This was Ken Goldberg’s discovery as a
graduate student at Carnegie Mellon,
where he proved that for any polygon,
there is a series of steps like those above
that can orient it into any position in a plane.
His thesis work addressed a common prob-
lem in robotics. “We tend to think that ro-
bots on an assembly line are working in lock
step with perfect precision. In fact, there
are slight variations,” says Goldberg, now a
Cal professor in Industrial Engineering and
Operations Research (IEOR) and Electri-
cal Engineering and Computer Science.
These variations can make a big difference.
Imagining his hand as a robotic arm and a
coffee cup on the table as the object to be
lifted, Goldberg continues, “If I grab this
cup just slightly off, sometimes it can go
flying. In an assembly, it can end up jam-
ming and breaking everything. I was trying
Professor Ken Goldberg lives at the intersection of
science. (Photo: Martin Sundberg)
to understand how I could pick up some-
thing like this cup in a reliable way.”
In his evenings at Carnegie Mellon,
Goldberg extended his work with robots
into the realm of art, putting paintbrushes
in a robot’s grip and programming it to
paint. In an untitled work from 1987, a blue
equilateral triangle composed of large,
sloppy brushstrokes painted by a robot sur-
rounds text that reads, “On a square lattice
at least one coordinate of an equilateral tri-
angle must be irrational. An irrational num-
ber has no finite numerical representation.
An equilateral triangle represented on a
computer is inherently imprecise.” In a
whimsical way, Goldberg addressed the in-
BERKELEY
science 16
review
herent imprecision of robots that
motivated his dissertation. “I made
a whole series of those triangles. I
was interested in the repeatabil-
ity—no two of them were ever
alike. I put them all on a wall, a
grid of twelve of them, and it il-
lustrated a fundamental aspect of
robots—that they’re imprecise.”
Goldberg continued to live a dual
existence as an assistant profes-
sor at the University of Southern
California, proving theorems that
extended his thesis work to new
geometries and new mechanical
situations and creating art exhib-
its involving robotic painting.
“These art installations would
take months of preparation, and
then the exhibit would be up in
the gallery for two or three
weeks, and I’d always be fixing it
and making sure it worked. It was
art and kind of exhausting.” He continues,
“When the World Wide Web came
out in 1993, there was access to data and
images on the Internet, including live cam-
era images. Once I saw that, I thought,
‘Oh, we can have a robot.’Then, I thought,
‘I can make an art installation that will stay
in my lab, and people from around the
world can come and visit it.’”
An engineer’s drive for efficiency led
Goldberg to be the first person to create
an Internet-controlled robot. He and his
students found a robot in the corner of
someone else’s lab and began working on
this unfunded “hobby.” They announced
the result—The Mercury Project—in
September 1994. Users could log in and
direct the robot, equipped with a stream
of pressurized air, to excavate objects that
were buried in sand. “It took off like
crazy. We were getting thousands and
thousands of hits a day on the website.”
Each user was given five minutes to con-
trol the robot, and at times, 20 people
were in line.
With the success of the Mercury Project,
Goldberg raised funds to build a new and
improved Internet robot. The Tele-Gar-
den, launched in June 1995, evolved
Internet robotics “from hunting and gath-
ering into agriculture.” Again, thousands
of users logged in to command the new
robot to plant seeds in a nine-foot-di-
ameter garden. They could monitor the
plants’ growth, and direct the robot to
water the plants. But the garden wasn’t
big enough to sustain the seedlings of
all of the users. “So, we’d go through
these seasons—we’d call them hurri-
canes—and we’d wipe everything out
and start over. When we announced a
hurricane, people would be very upset.
They’d send us email and say, ‘Please
send me my plant; I’ll pay for the post-
age,’ or ‘I’ll come to USC and pick it up.’
People would email each other and say,
‘I’m going on vacation, can somebody wa-
ter my plants?’”
While the Tele-Garden was of scientific in-
terest as an early Internet robot, Goldberg
and others also viewed it as an art project.
“A garden was the most absurd application
I could think of.The Tele-Garden was a cri-
tique of technology, a critique of the
Internet itself, and at the same time an in-
teresting look at group behavior where you
had to work with others.”
Goldberg came to Cal as an assistant pro-
fessor in 1995. “Berkeley was very recep-
tive to making art and doing science—you
could do both.” In 1997, he started the
monthly Art, Technology and Culture col-
loquium in which artists, philosophers,
computer scientists, and engineers speak
on issues relating to digital media, emerg-
ing technologies, and art.
Ouija 2000, which appeared at the Berke-
ley Art Museum and was selected to be part
of the 2000 Biennial Exhibition at the
Whitney Museum of American Art, ex-
tended Goldberg’s critique of the Internet,
and continued to explore collective behav-
ior. “People believed all kinds of things
could happen with the Internet. It was just
like a Ouija board.” Users who logged on
“It’s an ongoing challenge
to take something that you
have a gut interest in and
find a way to legitimize or
formalize it in a way that
it can be taken seriously
by your colleagues.”
to participate in Ouija 2000 were told:
“When you press the red button, the Ouija
board will materialize. Concentrate on the
question that appears. Rest your hands
lightly on your mouse, and move it over
the white planchette to begin playing.
Warning: many people feel that Ouija
boards summon powerful forces … ” The
Ouija program averaged all of the users’
mouse motions to determine where to
move the robotic planchette (the block that
slides over the Ouija board) to answer such
questions as, “Will the human genome be
decoded this year?” or “Should [user] go on
a spontaneous trip to Paris?”
Goldberg’s most recent work has focused
on the idea of “collaborative telerobotics,”
which originated in the Ouija project. “I
BERKELEY
science 17
review
started thinking that it would be much
more fun to move around in a physical en-
vironment,” says Goldberg. Enter the Tele-
actor, a person outfitted with a wireless
video camera and the ability to receive in-
puts by cellular phone. Internet voters cast
their ballot for what the Tele-actor should
do next, and the Tele-actor responds ac-
cordingly. The Tele-actor debuted at the
2001 Webby awards, where she interacted
with Sam Donaldson on stage under the
direction of web voters. High school stu-
dents have steered the Tele-actor through
the Berkeley Microelectronics Laboratory,
a chemistry lab on campus, and Profes-
sor Michael Eisen’s biotechnology lab at
Lawrence Berkeley National Laboratory.
Students catch on very quickly to the
web interface for driving the Tele-actor.
Says Goldberg, “The first time, I bud-
geted an hour to teach the kids how to
use the interface, which was about 59
minutes too long.” But, to keep students’
interest, Goldberg and his group have
added a “leadership score.” Students gain
points toward their leadership tally by
being first to vote for the move that is
the eventual favorite.
In May 2003, Goldberg and IEOR gradu-
ate student Dezhen Song installed the “Co-
opticon” camera on Evans Hall. The Co-
opticon points toward the Stanley Hall con-
struction site and sends its image to a
website. A panoramic view of the entire
field that the camera can reach is displayed
beneath the current camera shot. Users
drag a frame onto the panoramic shot to
indicate where they’d like the camera to
move next.What makes this camera differ-
ent from other “webcams” is that rather than
handling each request in sequence—which
can cause long queues like those in the
Mercury Project—the Co-opticon soft-
ware computes the view that best satisfies
all of the voters who are online at a given
 Profile
time. This isn’t just the average. If three
people vote to view the far left, and two
people vote to view the far right, the aver-
age would be somewhere in the middle,
which wouldn’t make any viewer happy.
With the Co-opticon the three voters on the
left win that camera view.The program cal-
culates a satisfaction score for each voter that
indicates what fraction of their vote was in
the camera view for that round.
Projects such as the Tele-actor and the Co-
opticon use technology in exciting and un-
conventional ways. But some might question
whether they belong in an academic setting.
Says Goldberg, “It’s an ongoing challenge to
take something that you have a gut interest
in and find a way to legitimize or formalize it
in a way that it can be taken seriously by your
colleagues.” In the case of the Co-opticon this
issue was settled by Song. “Dez was able to
pose the Co-opticon problem in a very nice,
formal way,” says Goldberg. “That completely
strikes a chord with me because it becomes a
geometric algorithmic problem.” Song points
out that “Ken was at first hesitant to start a
major research effort in this area because it’s
hard to do something theoretical, on paper—
something with equations. Then we started
thinking about it, and there are lots of theo-
retical things to do: how to share the camera,
how to make it fast enough, and they make it
theoretically valuable.” He adds, “So, I’m not
just some boy that plays with toys.That makes
me happy.”
Nor is Goldberg just playing with toys. “I
think the jury was out for a long time about
whether we could find a way to really make
a solid contribution out of [this work], he
says, “and now I feel like we can.” Indeed,
his work with Internet robotics was recog-
nized in 2001 with a Major Educational In-
novation Award from the Institute of Elec-
trical and Electronics Engineers.
Goldberg’s unique combination of science
and art breaks new academic ground and
raises questions about the role such projects
play in academia. Song notes, “[Ken’s]
projects are double-sided. They are digital
art, but they are also very serious research
projects. Ken himself is a hybrid. He changes
between artist and engineer.” Jane
McGonigal, a PhD student in performance
studies who has been the Tele-actor on sev-
eral occasions, has a similar view. “The fact
that he’s created space in his lab where an
artist like me can work as a real member of
the lab is an amazing opportunity. Ken is re-
BERKELEY
science 18
review
ally the only person I know who’s reaching
out to both artists and engineers.”
In the end, Goldberg himself doesn’t
draw such a distinction. “For me, doing
research feels very similar to making art.
Both remind me of fishing, in that they
require observation, experience, guess-
work, stubbornness, luck, and willing-
ness to throw back the little ones.” For-
tunately for all of us, Goldberg catches
the big ones pretty often.
Want to know more?
To plant in the Tele-garden, ask
the Ouija a question, or check
out the construction of Stanley
Hall,visit:
w w w. i e o r. b e r k e l e y. e d u /
~goldberg/
Jessica H. Marshall is a graduate student in
the Department of Chemical Engineering
Baytech Beat
BERKELEY BUSINESS PLANS RISE TO THE TOP
Theresa Ho
n a world where
I the boundary
between industry
and academia is
blurring, everybody
could use a little extra
business savvy. To An expanding T.rex illustrates the principle of the gastric retention device (kids don’t try this at home). (Photo: Kira O’Day)
provide that opportu-
nity, each year students from the Haas
School of Business hold the Berkeley
Business Plan Competition. The compe-
tition draws together scientists and
entrepreneurs from the Berkeley
community to turn innovative product
ideas into real-world businesses. Hosted
by Haas’ Lester Center for Entrepreneur-
ship and Innovation, this year’s fifth
annual competition saw 58 teams vie for
$70,000 in cash prizes. Over the course
of three grueling months, six teams made
it to last April’s finals, at which judges
from leading venture capital firms chose
three winners from a field one described
as “among the best I’ve seen.”
Bay Area company Medifuel won both
third place and the People’s Choice
Award, a $5,000 bonus awarded by
audience members at the final public
presentation. Using technology devel-
oped in Professor Liwei Lin’s lab in the
UC Berkeley Department of Mechanical
Engineering, Medifuel builds batteries
for small implantable medical devices,
such as pace makers, spinal cord stimula-
tors, and regulated drug-delivery
systems. Other batteries currently on the
market are bulky, have a limited lifespan,
and need to be surgically replaced.
Medifuel’s battery is roughly the size of a
dime and harnesses baker’s yeast to turn
the body’s own source of energy, glucose,
into electricity. Since a small population
of yeast can sustain itself by replicating,
the GlucoCell™ battery is essentially
self-renewing, and should never need to
be replaced.
Second place went to Vsee Labs, a
company that uses proprietary software
to facilitate virtual classrooms. By
drawing on user feedback and a five-year
visual communication study by Vsee
founder Milton Chen, the company’s
design provides the most natural class-
room setting possible. Vsee has also
developed algorithms to monitor
students and to provide high-quality
video streams during critical times in the
classroom. While students are taking
notes or listening passively, their images
are updated at a low frame-rate, but
when resolution becomes important,
such as when students raise their hands,
BERKELEY
science 19
review
the image becomes crystal clear. This
allows the instructor to notice subtle
nuances in a student’s question, and
detect confusion or boredom, without
sacrificing bandwidth. The frame-rate
frugality allows Vsee Labs to use low-cost
PCs and still provide five times the
number of high-quality video streams as
their competitors.
Gastric Retention Technologies (GRT)
took the first-place prize of $50,000, in
spite of much maligning of their name
during the competition. Regardless of
what the judges may have thought
initially, GRT has nothing to do with gas,
constipation, or vomiting. Now renamed
BaroNova Therapeutics, Inc., the com-
pany is developing a polymer pill that
expands in the stomach to suppress
appetite. Based on polymer technology
patented by Kinam Park at Purdue
University, the pill hydrates quickly once
swallowed, causing it to swell and take up
space in the stomach for about a week
before rapidly degrading as it enters the
small intestine. Because the polymer
never enters the bloodstream, it is
 Baytech beat
classified as a medical device rather than a
drug, and is therefore able to bypass the
costly clinical trials required by the FDA.
Unlike other more radical weight-loss
options like gastric balloons or gastric
bypass surgery, BaroNova aims to provide
an easy-to-use, non-invasive alternative
for losing weight.
Each of these winners paired a good
product with a shrewd business plan.
Teams are judged on market opportuni-
ties, their competition, and the qualifica-
tions of team members, at least one of
whom must be a Berkeley student or
alum. As Ilya Entin, chair of the student
organizing committee put it, judges are
“not looking for a cool gadget, they are
looking for a product they can sell.”
To help teams sell their product, the
competition organizes and manages
several supplementary programs. A
series of workshops gives students advice
on running a successful business, from
the basics of writing a business plan to
managing intellectual property or
estimating profitability. The
Entrepreneur’s Exchange acts as an
online dating service for entrepreneurs
seeking the right product, or a business
idea looking for the right combination of
technical, marketing, financial, and legal
expertise. The competition is organized
to provide constructive feedback at many
different stages. Semifinalists are
assigned mentors to guide them through
the process. Senior venture capitalists
judge the business presentations and
provide feedback after each round of
competition, allowing competitors to
refine and improve their pitch as the
competition goes on.
Several teams from previous competi-
tions have founded successful businesses.
Timbre Technologies, the winner of the
first annual competition, sold three years
later for $138 million. SkyFlow handles
high-volume calls at contact centers for
companies such as Apple Computer and
Bravanta. ZipRealty sold over $1 billion
in real estate in the last year via their
online realty services. Win or lose,
“There’s no such thing as a bad entrepre-
neurial experience,” says Nick Sturiale,
founder of Timbre Technologies. He
started two companies that failed before
striking gold with Timbre. According to
Entin, a main goal of the competition is
to help competitors make connections
and learn skills that will “maybe start
something big ten years down the road.”
Preparations for the sixth annual Berkeley
Business Plan Competition begin this fall.
Keep your eyes open—you could be part
of the Next Big Thing.
Teresa Ho was a graduate student in the
Department of Molecular and Cell Biology.

Read
Write
Advertise
Berkeley Science Review
sciencereview.berkeley.edu
BERKELEY
science 20
review
 Book Review
MY LIFE IN SCIENCE
BY SYDNEY
BRENNER
Reviewed By:
Giovanna Guerrero
ost people think
M discovery is the
sole motive behind
a scientist’s work. The exhila-
ration of “eureka!” aside, the
real thrill of science lies in the
chase; asking the right questions and find-
ing the most clever and convincing means
to answer them. “The right question” is the
main focus of Sydney Brenner’s memoir,
My Life in Science.
Sydney Brenner has had a remarkable life in
science. During the groundbreaking days of
molecular biology, Brenner made seminal
contributions that we take for granted today.
He demonstrated the role of messenger RNA
and helped to elucidate the nature of the ge-
netic code. Interested in developmental biol-
ogy since childhood, he then switched fields
and established the nematode C. elegans as an
important model organism. This “lowly
worm” has been vital to studies ranging from
neurobiology to cancer. For this effort, Sydney
Brenner was awarded the Nobel Prize in
Medicine in 2002. Yet, in spite of being so
respected and accomplished, Brenner remains
the same excited scientist who arrived in Ox-
ford fifty years ago. He has always been more
concerned with solving life’s puzzles than with
creating a reputation.
Like Brenner, My Life in Science is unpre-
tentious. It is the transcript of a 15-hour
videotaped conver sation between
Brenner and fellow scientist Lewis
Wolpert, with short commentaries by
editor Errol Friedberg inserted to add
structure to Brenner’s sometimes wan-
dering recollections. From his humble
beginnings in South Africa to his collabo-
rations with the likes of Francis Crick and
Seymour Benzer, My Life in
Science recounts the personal
journey of a scientific non-
conformist.
Image: Biomed Central
Those who have read Sydney
Brenner’s monthly commen-
tary in the journal Current Bi-
ology know him to be humor-
ous, irreverent, and, at times,
brutally honest. My Life in Sci-
ence displays this same candor.
Brenner is liberal with his
opinions no matter what the topic––he ad-
mits to stealing a book from the library as
a poor boy in South Africa, and quickly sug-
gests thieving if it’s the only means to an
education. However, what resonates
through the text is Brenner’s love for the
scientific process. His fervor for discovery
is particularly evident in his account of the
experiment that revealed mRNA’s central
role in protein synthesis. In order to dem-
onstrate that protein synthesis required not
just DNA and ribosomes, but also an in-
termediate messenger, Brenner and Jacob
needed to show that new radioactively la-
beled mRNA was required for protein
translation to occur: “… we got delirious
because the radioactivity curve began to
rise and I said in French, ‘Ascendez,
ascendez. It’s rising, it’s rising!’ Then it
went on rising. And then we realized it was
time for the radioactivity to drop if the ex-
periment was correct. So we were both
shouting at this machine ‘Go down, go
down, down, down,’ and the next tube
went up a bit but the increase was less and
I said, ‘It’s less, it’s less!’ And we were ac-
tually striving to bend the numbers. Then
the numbers came down and it was abso-
lutely convincing.”
It’s equally enjoyable to read how Brenner
went about choosing an uncharted research
direction. After participating in the first ma-
jor thrust of molecular biology—its “heroic
BERKELEY
science 21
review
period,” according to Günter Stent—the
subsequent years seemed to Brenner just
“sorting out the details.” He wanted to use
the knowledge he had gained from genes
to tackle more complex questions of de-
velopment and behavior.To do so he needed
to find an organism with a simplified body
plan that was amenable to genetic research.
This hunt led Brenner to the
“pervertebrates,” nature’s oddities, which
to him represented exaggerations of gen-
eral biological paradigms. Although many
of the organisms he tested turned out not
to be representative of general developmen-
tal principles, the trek through zoology’s
outer reaches finally led Brenner to C.
elegans. Due to its rapid growth, small num-
ber of cells, and straightforward genetics,
the nematode quickly became an important
tool for the discovery of genes affecting
development and neuronal function. From
a historical perspective it is fascinating to
read about the origins of C. elegans research.
But it is Brenner’s philosophical reflections
on biology and on the means needed to
approach the complex workings of life that
really drive his narration.
Scientific wisdom aside, My Life in Science’s
narrative is occasionally jumpy, and the re-
counting of experiments, hypotheses, and
conclusions would benefit at times from
more detailed explanations. In spite of these
flaws, and perhaps because of them,
Brenner’s story is engaging. The book reads
more like an intimate conversation than a
showy memoir or a textbook lecture. Fans
of Sydney Brenner and those eager to read a
personal account of the exciting days after
Watson and Crick’s discovery will enjoy this
book. However, what you’ll read is more
than just the autobiography of a remarkable
individual.What becomes evident is the uni-
versal lure of science, the thrill of asking the
right question.
Giovanna Guerrero is a graduate student in the
Department of Molecular and Cell Biology.
 Feature
GOODBYE, CYCLOTRON ROAD?
Feds may
deep-six Cal’s
nuclear pride
Chris Weber

f you walk from the Berkeley

I campus uphill along Cyclo-

tron Road to the Lawrence
Berkeley Laboratory (LBL), past
the security gates and into a eu-
calyptus grove, you will find
yourself at the 88" Cyclotron. LBL's Building 88, which houses the 88" Cyclotron. (Photo: LBNL)
The “FAQ” sheet in the lobby an-
ticipates the visitor’s question, “Have [scientists] ever discov- Nuclear science in the United States has set its sights on con-
ered anything really cool at this cyclotron?” and answers, “Well, struction of the next big operation, the Rare Isotope Accelerator
no discoveries honored by a Nobel Prize originated here. Not (RIA). With the field’s federal funding stagnant, nuclear science
yet.” Now time is running out for the cyclotron’s Nobel aspira- can barely afford research and development for RIA, much less
tions. The nuclear science facility, scheduled to close in Novem- RIA’s projected cost of $1 billion. In November of 2001, the US
ber of this year, has already reduced its operating schedule from Department of Energy (DOE), which provides 85% of the funds
seven days a week to four and a half. for nuclear science in the United States, recommended closure of
the 88" Cyclotron if budgets were to become tight.
The 88" Cyclotron’s very name places it in a gradually disappear-
ing scientific niche. The goal of a cyclotron is to accelerate ions— The committee that prepared the DOE report did not want the
atoms missing one or more electrons—to high energies and then 88" Cyclotron closed and called the possible closure “a significant
to collide them with stationary atoms. This cyclotron’s 88 inches loss to the nuclear physics community.” Freedman points out that
measure the diameter of its magnet, a number that determines “these reports are meant to be used. One way to use it is for getting
the maximum energy to which it can accelerate ions. The name is money: to say, ‘look at all the good research that will be lost if the
anachronistic in an era when science is largely driven by the push nuclear science budget is cut.’” The nuclear science budget did, in-
to higher energies and bigger facilities, since the 88" was not the deed, increase in 2002, but the strategy failed in 2003: In February
largest cyclotron—by more than two times—even in 1961, when the DOE announced that push had indeed come to shove, and that
it was built. To be sure, the 88" cyclotron has at various times the 88" cyclotron would have to close by November.
been the nation’s or the world’s best, by one measure or another.
But as Stuart Freedman, professor of physics at UC Berkeley and Putting new ions in an old accelerator
senior scientist at the cyclotron, points out, with an annual oper-
ating budget of five to six million dollars, “this isn’t big science; Nuclear scientists like those at the 88", their title notwithstanding,
this is a small operation.” do not design nuclear bombs or nuclear reactors. They leave those

BERKELEY
science 22
review
tasks to engineers and a few “applied” nuclear scientists, while they
explore the fundamental properties of atomic nuclei. Their wares
are the protons and neutrons (collectively “nucleons”) that make up
the nucleus, the quarks that compose nucleons themselves, and the
particles that carry forces between them. From these particles
nuclear scientists learn about the laws that govern nature, on scales
from subatomic to stellar.

Freedman’s own research at the 88", in collaboration with staff

scientist Paul Vetter, uses low-energy ions accelerated by the cy-
clotron to study the so-called weak interaction. One of the four
“fundamental” forces of nature and a linchpin of modern particle
physics, the weak interaction helps to regulate nuclear reactions
in stars, and creates the nearly-undetectable neutrinos that make Construction of the cyclotron's 88" magnet, 1960. (Photo: LBNL)

up between 10––20% of the universe’s mass. Freedman collides

electric fields to accelerate ions, the force it can exert depends on the
ion’s charge––that is, on the number of electrons removed. Heavier
...the small size of the 88" offers ions, being more sluggish, must have more electrons removed than
“grad-student-sized projects” that lighter ones in order to reach the same energies.

a single student can complete
the accelerated ions with atoms to create radioactive nuclei, which
then stream directly into his “atom trap.” By monitoring the ra-
dioactive decay of the nuclei held in this laser-beam trap, Freed-
man and Vetter hope to discover new information about the weak
interaction.
When the 88" Cyclotron first opened, available ion sources only
allowed it to accelerate ions containing a few nucleons, like bare
protons or the nuclei of helium atoms. In the late 1960s, scien-
tists at Berkeley and elsewhere invented what were then consid-
ered “heavy-ion” sources: devices that could remove up to six elec-
trons from elements as heavy as neon, with 20 nucleons. Posi-
tioned on a platform above the cyclotron, these sources delivered

Research at the 88" represents a branch of nuclear physics, in which

relatively low energies are desirable—though the energies are still
high enough to require a cyclotron. At these energies, nuclear in-
teractions lead to forms of “collective” order, in which groups of
nucleons take on strange new properties. The nucleons can form
pairs that become insensitive to the interactions around them, stay-
ing put while the unpaired portion of the nucleus rotates, or the
whole nucleus can assume a “super-deformed” state, elongated like
a football. Low-energy ions may also be collided with target mate-
rials in hopes of creating new elements—an enterprise at which
earlier Berkeley accelerators excelled, beginning with the 1940 dis-
covery of plutonium by Edwin McMillan, Philip Abelson, Glenn
Seaborg, and Emilio Segrè using a 37" cyclotron on the Cal cam-
The Advanced ECR ion source de-
pus.The 88" Cyclotron is now the only US facility that searches for livers ions through a beamline into
such heavy elements. the cyclotron's center, where they
are accelerated to high energies.
Inset: A cartoon from the series 'The
None of the research in this branch of nuclear science requires that cyclotron as seen by ... (Photo and
the cyclotron accelerate ions to especially high energies. Instead, image: LBNL)
nuclear scientists seek to accelerate ions of ever-heavier elements to
about the same modest energy per nucleon. Because a cyclotron uses

BERKELEY
science 23
review
 Feature
The nuclear line of stability. Black squares show the number of protons and
neutrons in stable isotopes; the yellow region shows known radioactive
nuclei. RIA is intended to explore the neutron-rich region marked "terra
incognita". (Image: Witold Nazarewicz, University and Tennessee and Oak
Ridge National Laboratory)
ions through an evacuated tube—a “beamline”—down into the
cyclotron’s center. The cyclotron itself, which Vetter says looks
like a “big tuna fish can,” accelerates the ions in
ever-larger circles until, at sufficiently high en-
ergies, they escape from the edge and stream into
one of the facility’s experiments. Operation of
Getting bigger
the 88" Cyclotron consists, in essence, of this may just mean
progress of ions along beamlines from source to
cyclotron to experiment. leaving good
Because new ion sources can be added to the science behind.
outside of an existing cyclotron, and can even share
a beamline with an older source, cyclotrons
themselves tend not to change, instead simply adding ever better
ion sources. Each new source, allowing the acceleration of yet heavier
elements to energies useful to nuclear scientists, opens potentially
new scientific territory. In 1982 Claude Lyneis, now the head of
the 88" Cyclotron, was sent to France to learn how to build a new
type of source that used “electron cyclotron resonance” (ECR), and
that could remove thirteen electrons from argon, an atom with 40
nucleons. He recalls that in late 1984, when the ECR source he was
building at the 88" passed its first test-run, the scientific demand
for the new ions was immediate and ceaseless.
Terra incognita
One of the most common uses for the accelerated ions leaving the
cyclotron is to collide them with a stationary target of a known
material. The energy of the collision can allow the ions’ nuclei to
BERKELEY
science
review
fuse with those of the target or can set the target nuclei spinning. As
the nucleus or the fused nuclear complex recovers from the colli-
sion, it emits particles—gamma rays, electrons, or nucleons—that
reveal its properties. For instance, scientists at the 88" have used
the Gammasphere, the world’s preeminent gamma-ray detector,
to learn about nuclear sizes and shapes. The device proved invalu-
able for investigating the collective low-energy behavior of nucle-
ons in super-deformed nuclei. In fact, it “nailed” the problem, ac-
cording to I-Yang Lee, the cyclotron’s scientific head, who explains
that though nuclear scientists had already seen evidence for the ex-
istence of super-deformed states, the Gammasphere allowed them
to determine for what combinations of nucleons the states could
form and how long they lasted.
Nuclear scientists’ ability to probe the unknown depends largely
on the creation of new, fused nuclear complexes, but they are lim-
ited by the range of isotopes available as parent nuclei. Elements
are defined by their number of protons, while isotopes of a given
element have different numbers of neutrons. Most isotopes are
unstable, shedding or transforming nucleons via radioactive decay,
with half-lives ranging from milliseconds to millennia.When nuclear
scientists chart the number of neutrons against the number of pro-
tons in the various nuclear isotopes, they find that
the stable nuclei lie on a nearly straight line. Ra-
dioactive isotopes typically live just off of this line
of stability, having slightly more or fewer neu-
trons than their stable counterparts.
Radioactive nuclei are typically created by
colliding and combining two smaller nuclei at
low energy. Most of the smaller parent nuclei
are stable, and fusion yields nuclei that are either
stable or slightly deficient in neutrons.
Substituting radioactive parent nuclei for stable ones can allow
fusion-reactions to produce neutron-rich nuclear complexes, but
radioactive parents are difficult to inject into a cyclotron’s ion
source. A few laboratories, like Oak Ridge National Laboratory
in Tennessee, can accelerate radioactive ions, but nuclear
compositions with significantly more neutrons than stable atoms
remain, according to Argonne National Laboratory near Chicago,
“terra incognita.”
Exploring this range will require the expensive Rare Isotope
Accelerator, proposed to be built either at Michigan State University
or at Argonne. Instead of creating radioactive ions by fusion of light
nuclei, RIA, if built, will break uranium—a heavy nucleus unusually
rich in neutrons—into nuclear fragments. It will be more
complicated than a cyclotron; one component will produce a beam;
of ions to break up nuclei in a uranium target, while a second will
24
accelerate nuclear fragments escaping the target to energies
comparable to those used at the 88" Cyclotron. The fragments will
be radioactive, neutron-rich nuclei as heavy as silver, gold, or lead—
isotopes that so far have remained out of reach.
Construction of RIA still awaits federal funding, however. Nuclear
science’s money-saving measures, such as the closure of the 88"
Cyclotron, are intended only to help the field to prepare for RIA.
In addition to modest increases in RIA’s research and development,
the DOE report suggested increased funding for the accelerators at
Argonne and Oak Ridge. The experiments at these facilities bear
closely on the nuclear science to be done at RIA, so they can help to
establish a base of scientific users for RIA. Argonne is also a potential
site for RIA. Berkeley, on the other hand, could not host the project;
the city’s politics would not accommodate an accelerator of
radioactive ions. Michael Beaudrow, a former operator of the 88",
sees a further connection between RIA and the cyclotron’s closure:
“When you’re asking for as much money as RIA will take, you have
to give something up to show that you’re serious about it. The 88"
was the sacrificial lamb.”
Bigger and better?
Many scientists at the 88" suspect, with varying degrees of
trepidation, that nuclear science might become as centralized as
particle physics, the original big science. According to Freedman,
the push toward ever larger accelerators has left that field with only
two facilities in the United States: Fermilab, in Chicago, and the
Stanford Linear Accelerator Center. He points out that until 1972
nearly every university with a program in nuclear science had a
cyclotron or a small “tandem” accelerator. Now two accelerators—
at Brookhaven National Laboratory in NewYork and at the Jefferson
Laboratory in Virginia—consume half the budget of nuclear science.
RIA will signal the consolidation of the majority of the field’s funding
into just three facilities.
As nuclear science prepares to consolidate its budget and personnel
into the new facility, however, some scientists believe the move will
detract from the field. To justify its funding, a big facility such as
RIA must direct itself toward some specific mission and consequently
tends to cut off research in avenues that don’t bear directly on that
mission. Moreover, big projects, like little ones, can fail—but at
higher stakes.
Not everybody is convinced that centralization has damaged par-
ticle physics. But progress in particle physics always means working
at higher energy; and high energy, Vetter notes, is a moving target:
“It just means the highest energies that you can currently reach. It’s
like trying to define ‘modern.’” Higher energy means larger accel-
erators, which in turn means fewer accelerators. For nuclear sci-
ence, on the other hand, progress is not necessarily linked to growth.
Getting bigger may just mean leaving good science behind. In fact,
nuclear science already comprises problems that particle physics
left behind unsolved.
Gathering dust
Row upon row of lights, switches, and dials cover the walls of the
88" Cyclotron’s control-room, from which the cyclotron’s opera-
tors run the apparatus and monitor the ion beam. The room must
have looked futuristic in 1961; now it looks gaudy by comparison
with the banks of high-resolution screens controlling modern ac-
celerators. The report that recommended “termination of DOE
support for operations” at the 88" did not, in fact, give a reason for
targeting the facility, but speculations abound. Possibly, Vetter says,
the 88" was targeted because of its age and accompanying “decrepi-
tude.” It takes several full-time
technicians—no small frac-
tion of the machine’s operat-
ing costs—to maintain the
vacuum and electrical systems
of the old and fussy beamlines.
Vetter gestures to the top of
the cyclotron and says, “that’s
40-year-old dust on there.”

Did the 88" Cyclotron sim-

Operators Mike Johnson and Diane Riley-
Cole in the the 88" Cyclotron's control
room. On the right cartoon from the se-
ries 'The cyclotron as seen by ... (Photo:
Chris Weber; image: LBNL).
ply miss its date with destiny
by not working with radioac-
tive ions? Actually, radioac-
tive ions have been intro-
duced into the 88" in a lim-

BERKELEY
science 25
review
 Feature
Paul Vetter and postdoc Reina Maruyama with Freedman's atom-trap. The
researchers use the atom-trap to investigate the weak interaction.
(Photo: Chris Weber)
ited way.The ions enter the 88" after traveling for twelve seconds
as a gas along a capillary tube from a medical cyclotron elsewhere
on the LBL site. Medical cyclotrons cost about a million dollars
each and require only a single person to operate them. Hundreds
of hospitals around the nation use them to
“The 88" was the
produce radioactive isotopes for medical
imaging. Medical cyclotrons have a limited
use for nuclear research, however, since
only certain ions can be piped along the
capillary tube before they decay. More-
sacrificial lamb.”
over, Lee says, the cyclotron’s radioactive beams simply have not
come on line fast enough to contribute to preparations for RIA.
Beaudrow explains that the quantity of radioactive ions that the
project could deliver, though constantly increasing, was still too
small.
Bernard Harvey, the 88" Cyclotron’s first director, wonders what
role element 118 might have played in the DOE’s decision. In
1999, amid great fanfare, scientists from the cyclotron’s “gas-filled
separator” announced the creation of an element with 118 pro-
tons, the largest ever made. But in 2001, to their great embar-
rassment, they retracted the claim, having traced the result to
fabricated data that one scientist had inserted among the experi-
mental results. Most scientists at the 88" seem to doubt that the
scandal of element 118 influenced the DOE report. On the other
hand, Vetter speculates that perhaps without element 118, LBL
and UC Berkeley would have opposed the closure more vocally—
though possibly still unsuccessfully.
BERKELEY
science
review
Have experiment, will travel
The money that the DOE provides to LBL’s Nuclear Science Divi-
sion supports both the operation of the 88" Cyclotron and the re-
search of the division’s many scientists. Many of these scientist work
at the cyclotron, but since the early 1990s some have done their
experiments at Brookhaven, where they can accelerate ions to higher
energies.They travel for a few months each year and analyze data at
LBL in the remaining months. After the cyclotron’s closure the DOE
will continue to support LBL’s nuclear scientists. But those now at
the 88", like their colleagues, will have to pack their valises and
vacuum-pumps and move their experiments to a different cyclo-
tron—as will the 40 to 50 scientists from other institutions who
now travel to the 88" Cyclotron each year.
Freedman’s investigations of the weak interaction will certainly suf-
fer from the move to another cyclotron.Vieregg explains that their
atom-trapping apparatus, which covers several tables with mirrors
and lenses, will take several years to move and reassemble. It is a
shame,Vetter says, to stop the project now, since the atom-trap has
just begun to yield high-quality data. Most graduate students in
Freedman’s group will be able to finish their experiments by the
time the cyclotron closes, but future students won’t have the op-
tion of doing research at the 88". Graduate students Wes Winter
and Nick Scielzo say that the cyclotron’s
ready availability figured prominently in
their decisions to study nuclear science;
neither student arrived at Berkeley with
the field in mind. Freedman explains that
the small size of the 88" offers “grad-stu-
dent-sized projects” that a single student can complete. He notes
that particle physics, more centralized than nuclear science and with
larger projects, finds graduate students scarce.
Because Freedman and Vetter’s experiments ask some of the same
questions as particle physics, Vetter calls them “particle physics on
the cheap.” Certainly these little experiments strike a stark contrast
with that field’s giant accelerators and transcontinental collabora-
tions. Even on the scale of the 88" Vetter’s experiment is small; far
from needing the unusual ions that RIA will provide, this project
doesn’t even use the full capabilities of the 88" Cyclotron’s ion
beams. Both Freedman and Vetter mention the possibility of pur-
chasing a medical cyclotron in order to carry on some portion of
their research at the 88". The scheme may work for this experi-
ment, but in either case Vetter detects an absurdity in the situation
of the Nuclear Science Division once the 88" closes. “If we’re a
nuclear science lab without a cyclotron, then what the heck are we
doing?” he says.
26
The fates of some researchers at the 88" are currently up in the air. mand for beam time. Because most industrial users at the 88" are
Winter and Scielzo are graduating.Vieregg may forsake nuclear sci- ultimately contracted to DOD, the Air Force sent letters to DOE in
ence for atomic physics.Vetter could also switch to atomic physics, July expressing concern that the cyclotron’s closure could impede
but has other options. He could remain at LBL and travel to other national security. Later in the month, a concrete proposal emerged:
cyclotrons to do his experiments, or he could move to Argonne or DOD would put $2 million yearly toward continuous operation of
the cyclotron at the University of Washington. the cyclotron, and DOE would put in the remaining $3 million.
The beam time allotted to chip-testing would rise from 25% to
Unlike the cyclotron’s sci- 50%. The DOD has good rea-
entists, its electricians, son to fund the cyclotron: if
vacuum technicians, and the 88" shuts down, only the
operators will have no cyclotron at Texas A&M Uni-
choice but to leave the 88". versity will remain suitable for
Beaudrow is lucky; he’s al- these tests. Many scientists at
ready switched from oper- the 88" estimate that the
ating the 88" to operating money DOD saves by testing
LBL’s Advanced Light a satellite’s chips pre-launch is
Source, an accelerator that far more than the cost of op-
uses high-energy electrons erating the cyclotron round-
to produce X-rays for a va- the-clock, seven days a week.
riety of experiments in
chemistry, biology, and ma- The success or failure of the
terials science. Many of the DOD proposal could mean
88" Cyclotron’s technical life or death to the cyclotron.
staff will have to change “The cyclotron as seen by the government funding agency.” Part of the 1967 car- Moreover, Freedman points
their field more radically; toon series “The cyclotron as seen by . . . “, by Dave Judd and Ronn MacKenzie. out, “the program of testing
See the rest of the series at http://imglib.lbl.gov/ImgLib/photo-archive.html
one former operator has image: LBNL.) chips for satellites has real
chosen to switch to LBL’s benefits for everybody.”
radiation safety program. Vieregg is concerned, though,
Others could lose their jobs, says Peggy McMahan, the cyclotron’s that if the 88" Cyclotron depends on the military for funding, “its
research coordinator, if LBL’s demand for technical staff drops. Ac- long-term future may not be secure. It will always be competing
cording to Freedman, the presence of a technical staff helps draw for money with military bases in senators’ home states.” As the
scientists to big facilities and even to middle-sized facilities like the BSR goes to press, nobody knows the eventual outcome of these
88". Few universities can offer individual researchers such direct negotiations. But with the closure scheduled for November, there’s
access to trained technical people. little time left for discussions. McMahan says that if the deadline
gets much closer without a decision, “even more staff will leave
Science’s wealthy Uncle the cyclotron.”

Despite the DOE’s report, the 88” may not be absolutely doomed.
For some twenty years, the cyclotron has sold beam time to indus-
trial users interested in testing how microchips intended for use in
satellites stand up to ionized nuclei streaming from the sun. These
users make chips for aerospace companies such as Lockheed-Mar-
tin that are themselves contracted to NASA or to the military. Manu-
facturers pay about $800 per hour to use the cyclotron’s beams of
accelerated ions to simulate the solar bombardment.
Since the 88" Cyclotron began testing chips, transistors have be-
come smaller, and chips have incorporated more of them.The need
for testing has steadily increased and so has industrial users’ de-
Most DOE programs are directed toward specific missions, such as
developing renewable energy technologies. The funds for nuclear
science, however, come from a mission-exempt category.The policy,
Freedman says, has historically been to treat nuclear science “like a
park—something worth preserving even though it has no practical
value.” The 88" Cyclotron is poised to break that pattern; its prac-
tical value to aerospace is all that can preserve it now. Even if the
cyclotron is saved, the clouds of consolidation are gathering over
nuclear science. It’s anybody’s guess how much room there will be
for small facilities under that particular umbrella.
ChrisWeber is a graduate student in the Department of Physics

BERKELEY
science 27
review
 Feature
AN EVENING IN THE GARDEN
Touring Berkeley’s botanical treasure

Letty Brown
© 2003 Katie Standke Photography.

Photographs by Katie Standke Pacific chorus frog, Pseudoacris regilla, on the bird's nest
fern, Asplenium nidus, which is native to tropical Asia.

arrive at the UC Botanical Garden on a midsummer evening

I just in time for the twilight tour. Already, the group of 25

botanists, gardeners, orchid lovers, and people just out for a stroll
crowds excitedly around the guide. He is Jerry Parsons, horticultur-
ist for the Australasian section, a tan man in his forties whose eyes
light up as he speaks. “At kilometer 73 from Xingshan Xian, hang a
left on a gravel road, continue for 1.3 kilometers …” He is reading
the acquisition record for Rosa helenae, a plant in the Asian section
that was collected as seed on a 1980 expedition to China. “Fagus-
Quercus-Betula forest on steep slopes. Southeast exposure. Rocky soil.
Shaded forest margins.” For every plant brought to the garden, expe-
dition members take detailed notes on its original location and envi-
ronment. These notes are one reason that the UC Botanical Garden
is often cited as the premiere university botanical garden in the na-
tion. “This is far more than just a collection of pretty flowers,” ex-
plains Parsons. “Each plant has valuable scientific data associated with
it.”The crowd gathers closer and starts to ask questions. Parsons leads
on and the tour begins.

© 2003 Katie Standke Photography.
Giant horsetail, Equisetum telmateia var. braunii, native to California.
One of the first things I notice is that the garden is divided into
geographic regions. The plantings are organized according to their
place of origin, in settings resembling their native habitat. As Paul
Licht, the garden’s new director and former dean of the College of
Letters and Science, puts it: “Where else can you visit South Africa,
Australia, or Chile in December, and some of the Mediterranean,
California, or Central America, all on the same day? It’s a rather
unique setting in that sense.”

BERKELEY
science 28
review
Worlds apart Echinopsis terscheckii, a saguaro-like cactus from the Andes of Argen-
tina that stands 30 feet tall. Suddenly, within a matter of feet, we
At our first stop, we come to a small pond outside the fern leave the desert and enter a lush forest that centers around Straw-
and carnivorous plants house. This display is maintained by berry Creek.The climate literally shifts around us. Before this area,
Parsons, and he lists some of its aquatic plants: “Cannas, cat- known as the Asian section, was established, visiting researcher and
tails, sedges, watercress, Cruciferae, waterlilies, this one’s in famous adventurer and botanist Joseph Rock recognized that its
the Onagraceae, water fern, papyrus.” The garden grows pa- microclimate matched the river gorges of western China. At the
pyrus for Professor Donald time Rock was en
Mastronarde of the Depart- route to this region,
ment of Classics and Greek and he later brought
Literature. He teaches a fresh- back many of the
man seminar in which students section’s earliest
use it to make paper accord- plants.
ing to ancient methods from
Egypt and Mesopotamia. We wander on past

© 2003 Katie Standke Photography.

the Japanese Pool, an
The garden is a teaching tool for important breeding
many departments on campus, habitat for the Califor-
and the class that brings the most nia newt (Taricha
students up the hill by far is Inte- torosa), and then file
grative Biology’s Biology 1B. Each along a shaded stone
semester, 1,000 students pour path to stop at a large
into the garden for two different Rhododendron tree.
Bio 1B labs.As we walk to the next Parsons recounts see-
section and stand between the Beautyberry, Callicarpa sp., from Asia.
Southern African Section and the
New World Desert, Parsons illustrates how the garden is used to
teach an important evolutionary concept. He points to the aloe plants
that dot South Africa Hill, and then to the agaves of the New World
Desert, plants with spiky bluish-colored whorls of leaves. “Conver-
gent evolution means that different types of plants in similar envi-
ronments can evolve the same types of structures. These look really
similar, but are in completely different families. The aloes are in the
Liliaceae, the agaves in the Agavaceae.” The plants have completely
divergent flower morphologies and aloe leaves have a gelatinous in-
terior, whereas agave leaves are fibrous. However, they evolved the
same leaf morphology to survive climates with long periods of
drought. Many other plants evolved this rosette leaf shape indepen-
dently, including plants in the Asteraceae, Crassulaceae, and
Bromeliaceae families. “Another example of convergent evolution is
the cactus family and the euphorbia family,” says Parsons. He points
to a South African plant that looks distinctly cactus-like, but, he tells
us, is a Euphorbia. “When we go to see movies we can tell if they’ve
really gone to Africa or not, because if you see anything that has a
cactus in it and [is] supposedly in Africa, that’s not right,” explains
Parsons. “The Cactaceae, with the exception of one genus of epi-
phytic cactus on the African continent, is only New World.” © 2003 Katie Standke Photography.
A tropical pitcher plant, Nepenthes ventricosa, an endangered
Moving on past the desert section, we stop to gape at the massive species from the Philippines.

BERKELEY
science 29
review
 Feature
© 2003 Katie Standke Photography.
Borage, Borago officinalis, in the Crops of the World Garden.
© 2003 Katie Standke Photography.
Dudleya sp., a succulent in the stonecrop family.
BERKELEY
science
review
ing archival photographs from Rock’s 1932 expedition to China, black-
and-white images of carts and mule trains transporting tree rhododen-
drons such as this one. Indeed, many of the garden’s 13,000 taxa were
acquired during expeditions of this sort in the mid 1930s through 1960s.
Now, however, that era has passed. “Those types of expeditions are un-
likely to happen in the same way they did in the thirties and forties
because of bioprospecting,” explains Chris Carmichael, manager of col-
lections and horticulture for the garden, “The regulatory climate has
changed and [they] would no longer be allowed.” Today, most of the
garden’s plants come from wild seed, either from seed-collecting trips
or from international seed exchanges.
Two things make the garden’s collection of particular value to science. First,
each plant comes from the wild; most come from seed, and some come
from cuttings or a seedling. Second, as Parsons demonstrated, each plant is
accompanied by very specific data on location and environment. This is of
use to scientists who study the evolutionary relationships of plants. Parsons
explains: “If someone is doing a treatment of the genus Rosa, they don’t have
to go to China to collect it. If they are doing, say, genetic research on it, we
can send them a sample of that material with exactly where it was col-
lected.” This is no small achievement. According to curator Holly Forbes,
the garden has supplied hundreds of researchers with living plant material
collected from all over the world, and demand is expanding as botanical
research increasingly requires fresh material for molecular techniques. Some-
times this work makes life harder for the garden. Recently the lily family,
Liliaceae, was broken up into many different families.This change will entail
a great deal of relabeling in the garden.
The garden is also premier among botanical gardens because of the variety
of ecosystems it encompasses, which is a result of the Bay Area climate. Di-
rector Paul Licht explains: “If you’re a gardener, this is paradise. We live in a
gardening zone that is the envy of most of the country. It doesn’t get too hot
or too cold, there are rarely prolonged frosts. What this allows us to do is to
grow more different kinds of plants than other places. Missouri Botanical
Garden can be under snow six months out of the year, and it gets above 100
degrees in the summer. I would guess you couldn’t grow 80 percent of what
we grow here outside.”
Golden State treasures
Continuing along, we enter the California section, where much of the garden’s
conservation efforts are focused. We pass a pygmy forest (native to the
Mendocino coast), a vernal pool, and an alpine fell-field, host to plants found
at elevations of 8,000 to 13,000 feet in the Sierra Nevada. We come to one of
the garden’s rarest acquisitions, the Presidio manzanita (Arctostaphylus hookeri
ssp. ravenii), a low-growing evergreen shrub with reddish bark. It is sometimes
called Raven’s manzanita after Peter Raven, the famous botanist, Berkeley alum-
nus, and director of the Missouri Botanical garden.The plant historically ranged
30
 © 2003 Katie Standke Photography.
Agave sp.

BERKELEY
science 31
review
 Feature
been using the garden's spiral aloe, Aloe polyphylla, to obtain seeds
that will be grown for future garden plant sales. The plant is an
obligate outcrosser (meaning it cannot self-fertilize), so pollen
has to be obtained from a different individual. The closest indi-
viduals of Aloe polyphylla that were in flower at the same time
were in the backyard of a Santa Cruz resident, who collected
seed in Lesotho as a Peace Corps volunteer many years ago. This
year when the spiral aloe bloomed, Barany brought valuable
pollen from plants in Santa Cruz back to the garden, painted it
on the flowers with a small brush, then watched closely as they
produced seed.

We cross Centennial Drive to visit the garden’s tallest plants,

the coast redwoods (Sequoia sempervirens) in the five-acre Mather
Redwood Grove. As we walk into the grove, Nathan Smith tells
me a little California trivia. “Did you know that California houses
the world’s oldest plant [a creosote bush in the Mojave], most
massive plant [a giant sequoia], and the tallest plant [a coast
redwood]?”he asks. We gaze up gaze up an example of the world's
tallest species. The Mather Redwood Grove is used by Professor
Todd Dawson of Integrative Biology who studies questions re-
lated to coastal redwood distribution, which is now vastly smaller
than its historic range [see the Labscopes section to learn more
about Dawson’s research]. He uses the garden to test sensors
that measure variables like light, temperature, wind speed, and rela-
© 2003 Katie Standke Photography.

Fern fiddleneck

in scattered serpentine soils along six miles of the San Francisco

peninsula. Primarily because of development, only a single plant
remains in the wild, though a few individuals survive from cuttings
of the parent plant in botanical gardens such as this one.The garden
is researching options for propagation and in the future may assist
in adding new individuals to the wild. Nathan Smith, the California
section horticulturist, tells me, “We serve as a repository.This plant
is here should the original plant be lost in the wild.” This reposi-
tory function is also critical to the garden’s scientific efforts. As
Carmichael points out, “This is a scientific collection—if you lose
one plant, you can’t just go down to the local nursery and replace
it. Some of them are truly irreplaceable.”

In total, the garden houses near 600 threatened or endangered

species (one-third of which are from California), each identi-
fied by a red dot on its tag. The garden works with a number of
conservation programs for these species, including the national
Center for Plant Conservation, of which the garden is a partici-
pating institution, and various state and federal agencies. Some
© 2003 Katie Standke Photography.
of these species are especially prized in the horticultural trade.
Bill Barany, the horticulturist for Southern African plants, has Knobby clubrush, Isolepis nodosa from Australia and New Zealand.

BERKELEY
science 32
review
tive humidity. He and his stu- invasive. Garza and other
dents climb some of the garden staff work with the
grove’s 70-meter trees to in- California Exotic Pest Plant
stall the small wireless sen- Council to come up with a
sors. “[The proximity of the list of alternatives to invasive
garden] allows me to test new species that people are still
technology, to explore ques- planting in their gardens.
tions here close to campus as They also keep close tabs on
opposed to in remote sites— garden plants that are poten-
it saves us an enormous tially problematic, “weedy,”
amount of resources,” or as Garza calls them, “spa-
Dawson tells me one after- tially inquisitive.”
noon while we watch his
students rig ropes high in The garden is well aware of
the trees. the threat posed by invasive
© 2003 Katie Standke Photography.
species, and the role botani-
The garden’s research focus cal gardens have played in
Echeveria sp., a succulent in the stonecrop family. past introductions. Now bo-
stems from its origins as a re-
search garden for botanists tanical gardens follow many
on campus. Founded in 1890, it was originally located where rules when importing material. The vast majority of acquisitions
Haviland Hall and the Memorial Glade are today. Development on brought in by the garden are in the form of seed, which are much
campus and a desire to be above the atmospheric inversion layer less likely to contain pathogens than plants or cuttings. In the rare
inspired its move up the hill in the 1920s to its current location on case that cuttings or seedlings are transported, all soil is removed
34 acres of Strawberry Canyon. Back then, its principal function and each plant is inspected carefully by the USDA. Parsons, who
was to support botany instructors and researchers on campus. It has collected for the garden in Borneo, Costa Rica, and Ecuador,
was only in the 1960s that the garden launched a program of out- explains that the USDA has lists of prohibited plant materials that
reach to the wider community. Now it is the only one of the five are potential hosts for disease pathogens or are known to be
Berkeley natural history museums to be open to the public, except “weedy.” For instance, Parsons explains, “plants in the Persea (or
for a small arm of the Hearst Anthropology Museum. The public avocado) genus are prohibited because of a weevil found in wild
has been responsive; more than 67,000 people visit the garden an- avocado seed. If introduced, it could wipe out California’s entire
nually. The outreach program is nationally recognized, funded by avocado crop.”
NSF and other major granting agencies, and teaches about 4,000
elementary and middle school students annually. Additionally, the The twilight tour is wrapping up.The sun has long since set and the
garden has a team of 200 volunteers who give about 20,000 hours tour members, fatigued from their walk across five continents, have
annually, including a corps of volunteer plant propagators who earn ceased to bombard Parsons with quite so many questions. Parsons
more than $40,000 a year for the garden with their plant sales.The closes the gate behind us, no doubt eager to get home, as he has to
garden has also earned some romantic distinctions: it was voted be back early tomorrow to help chase wayward deer out of the
best place for a first date and best place for an outdoor wedding in garden’s boundaries. We walk to our cars, leaving behind us the
Northern California, according to the East Bay Express. spectacle of desert, chaparral, cloud forest, and redwood grove, all
interwoven into one 34-acre plot of land.
A garden on guard
Letty Brown is a graduate student in the Department
As Parsons’s tour continues, we walk through Australasia, an area of Environmental Science, Policy & Management.
encompassing Australia, New Zealand, Indonesia, and the South
Pacific islands. Here we pass a plant called Gahnia aspera, a tall sedge
that bears an incredible likeness to pampas grass, the Argentinian Want to know more?
plant considered one of the worst invaders in the state. Anthony
Garza, supervisor of the garden’s horticulturists explains that al- Visit http://botanicalgarden.berkeley.edu/
though Gahnia looks like pampas, the garden has found that it is not

BERKELEY
science 33
review
Feature
NATURAL SOLUTIONS
Organic Answers to Toxic Questions
Carol Hunter

J ust after midnight on March 24, 1989, the giant oil tanker Exxon Valdez ran aground on Bligh Reef
off the coast of Alaska, spilling almost 21 million gallons of crude oil into the sea. The disaster
caused an oil slick that covered 3,000 square miles and contaminated 1,090 miles of coastline along
the pristine Prince William Sound, killing thousands of marine creatures and costing Exxon $1.28 bil-
lion in clean-up costs.

While it was an environmental nightmare, the ExxonValdez clean-up was also a large—and very public—
field test of a technology known as bioremediation. When skimmers and other clean-up machinery
could not handle the job, scientists at the Environmental Protection Agency added fertilizer to the beaches
to stimulate the growth of indigenous microorganisms and speed the natural breakdown of the toxins,
cutting the time to degrade the oil on Alaska’s shoreline from five to 10 years to an estimated three to
five years.

The high-profile clean-up efforts captured headlines in the NewYork Times, Washington Post, Newsweek, and
other major media outlets. But bioremediation technology had been used long before Exxon Valdez ran
aground for more insidious environmental disasters—gasoline leaking into groundwater from under-
ground storage tanks or toxic chemicals from contaminated military sites seeping into lakes and streams.

The cost of clean-

ing up existing en-
vironmental con-
tamination in the
United States
could be as much
as $1 trillion.
(Photo: the Envi-
ronmental Mining
Council of British
Columbia)

BERKELEY
science 34
review
Scientists from many UC Berkeley departments and the Lawrence
Berkeley National Laboratory have made significant contributions
to the field of bioremediation. Three of these scientists, Professor
John Coates, Dr. Hoi-Ying Holman, and Professor Norman Terry,
demonstrate the wide range of research that can be put to task in
the field, from new strains of bacteria and innovative monitoring
techniques to genetically modified plants.

Putting Nature to work

Put simply, bioremediation is the use of biological organisms to

return the environment to its natural state. Bioremediation in the
form of decomposition has been around as long as life has existed,
and human civilizations have been engineering systems such as com-
post piles and kitchen middens to take advantage of this natural
process since at least 6,000 BCE.

These days, the term bioremediation usually describes more com-
plex, scientifically engineered systems that use organisms such as
bacteria, fungi, and plants to process toxic chemicals into less-toxic
or non-toxic forms. Bioremediation has been used to clean up pe-
troleum waste since the 1950s. Because petroleum is a naturally
occurring, organic substance, the environment is full of microor-
ganisms that can degrade it and often only require additional nutri-
ents or an oxygen source to accelerate the process. More recent
research has focused on other toxins: chemical solvents, heavy met-
als, and radioactive elements.
In 1980, the EPA started the Superfund Program to deal with these
toxins by identifying and cleaning up sites that were so polluted
they posed significant risk to the environment and public health.
These sites are filled with contaminants such as vinyl chloride, chlo-
roform, benzene, polychlorinated biphenyl (PCBs), mercury, lead,
and others, many known to cause cancer, birth defects, and liver
damage. The program has investigated almost 45,000 sites across
the country and placed 1,500 sites on its National Priority List.The
Bay Area is one of the toxic hotspots in the country, with 27 sites on
the Superfund priority list, including the Hunter’s Point Naval ship-
yard in San Francisco, the Alameda Naval Air Station, Lawrence
Livermore Lab, and over a dozen semiconductor manufacturers in
Santa Clara County.
Dechloromonas strain RCB bacteria can break down two toxins, both per-
chlorate and benzene, in anaerobic environments. (Image: John Bozzola
and Steven Schmitt, SIUC IMAGE Facility)
fers deep underground. A 1986 survey of groundwater found that
36% of more than 5,000 community water sources in the United
States had organic contaminant concentrations above the maximum
contaminant levels allowed for drinking water. Once a contami-
nant has infiltrated a large body of water, it is almost impossible to
treat chemically. For these cases, bioremediation is the only afford-
able alternative.
Bacteria to the rescue
Dr. John Coates, a new professor in UC Berkeley’s Department of
Plant and Microbial Biology, has been researching the bioremediation
of one of these groundwater contaminants, a chemical known as
perchlorate. Used for solid rocket fuel as well as in explosives,
bleaching agents, and defoliants, perchlorate interferes with iodine
uptake into the thyroid gland and can cause fatal bone marrow dis-
ease and thyroid gland tumors. Because the thyroid plays an impor-
tant role in development, perchlorate is especially dangerous for
expectant mothers, since the fetus could suffer delayed develop-
ment and learning disabilities.

The cost of removing these contaminants chemically or mechani-
cally is prohibitive.The Superfund program alone has spent around
$40 billion to clean up just over half of the sites on the National
Priority List, and, according to the US Geological Survey, cleaning
up existing environmental contamination in the United States could
cost as much as $1 trillion. Many of toxins have leached away from
their original source, contaminating rivers and lakes as well as aqui-
Unlike petroleum contaminants, perchlorate is not naturally found
in most environments. According to Coates, it’s only known to
occur naturally in the remote Atacama Desert in Chile. Scientists
didn’t expect many microorganisms to have evolved a way to break
it down. “As of six years ago, there were only two known bacteria
species that could process perchlorate,” says Coates. “It was very
poorly studied.”

BERKELEY
science 35
review
 Feature
John Coates with graduate student Romy Chakraborty at Southern Illinois
University. Chakraborty was the first person to isolate a bacterium that could
break down benzene in oxygen-depleted environments. (Photo: J. Coates)
But once it was found in California drinking water in the late 1990s
and recognized as a threat to public health, interest in the
bioremediation of perchlorate increased dramatically. In 1997,
Coates, then at Southern Illinois University, along with colleague
Dr. Laurie Achenbach, began searching for more microbes that could
break down the contaminant. Coates first developed a selective
medium that would screen out everything but perchlorate-loving
microbes. He then conducted a worldwide search for these bacte-
ria in almost every type of environment, including contaminated
and pristine soils, wetlands, aquifers, lake sediments, river sedi-
ments—even swine waste. “You name it, we probably looked there,”
says Coates.
To their surprise, they found that these microbes were not rare, but
were ubiquitous in the environment. “We found these bacteria in
every site, even in Antarctica,” Coates says. He eventually isolated
about 40 different species of the microorganism, all belonging to
the phylum Proteobacteria. “The Proteobacteria is very broad, with
five subdivisions,” says Coates. “This metabolic pathway was found
in four of the five subdivisions. It is very phylogenetically diverse.”
From an academic standpoint, these bacteria are intriguing due to
their diversity and distribution in the environment. Some species
can flourish in very harsh conditions, from acidic to basic, as well
high salinity. But their ability to break down perchlorate makes these
organisms interesting from a practical standpoint as well.
Unlike many bacteria previously used for bioremediation, these
proteobacteria eat away at the perchlorate in an anaerobic envi-
BERKELEY
science
review
ronment. In fact, they require an anaerobic environment before
they will start processing the chemical. Just as animals need to
breathe air that contains oxygen to survive, many bacteria must
absorb oxygen from their environment in order to complete their
metabolic process and get energy from their food. But these
proteobacteria will use perchlorate as an oxygen substitute when
oxygen is not present.
In the field, engineers add a nutrient source like lactic acid to the
contaminated area, stimulating the growth of all kinds of bacteria
that quickly “breathe” all the available oxygen. Once the oxygen
has been depleted, then the naturally occurring proteobacteria will
start “breathing” perchlorate. But creating the right conditions to
get things working can be tricky. If too little lactic acid is added,
the bacterial growth will be limited by available food, oxygen will
still be available, and the perchlorate-utilizing bacteria won’t pro-
cess the perchlorate. But if too much is added, bacteria growth will
be over-stimulated, meaning the microbial populations might munch
through all the perchlorate and turn to other oxygen substitutes,
altering the natural geochemistry of the environment. According
to Coates, they would first use ferric iron, which can cause a bad
taste in the water as well as the accumulation of rust in pipes, then
they would turn sulfate to sulfide, creating a nasty “rotten egg” smell,
and finally they would produce methane, a greenhouse gas.
Once this delicate balance is achieved, however, the results are
extremely positive. “For the field trials that have been done, per-
chlorate has been completely removed to all intents and pur-
poses,” Coates says. “It was immeasurable or undetectable after
stimulation in the field.” Coates is currently involved in two field
trials, where his lab is following the remediative process more
closely and carefully studying the microorganisms involved.
Coates also discovered that these bacteria can remediate more
than perchlorate. His lab has identified a particular strain of
proteobacteria, Dechloromonas strain RCB, that can use the pe-
troleum contaminant benzene as its food source at the same time
it is using perchlorate as an oxygen substitute. Benzene is a dan-
gerous, carcinogenic chemical often found deep in aquifers or
soils where oxygen is not present. Not only do these particular
Dechloromonas bacteria take care of two contaminants at the same
time, they are also the first organisms discovered that can break
down benzene in an anaerobic environment.
Spying on cells
One difficulty with bioremediation is that scientists don’t always
understand what is going on in the field. As Coates puts it, there is
a lot of “wait and pray” in the bioremediation business—you add
bacteria or nutrients to an environment and then hope that the con-
36
centration of toxins goes down. Because of the complex systems
involved, it can be difficult to determine what exactly is causing a
change. A team at Lawrence Berkeley National Laboratory has dis-
covered a method that will take some of the guesswork out of
bioremediation. Using a technique known as Infrared
Spectromicroscopy, Hoi-Ying Holman and her colleagues Michael
Martin and Wayne McKinney are able to watch the inner workings
of cellular metabolism in real time.

On most days at LBL, you can find Holman hurrying up the

249 steps from her office to the experimental station inside
the Advanced Light Source (ALS)—her daily workout routine,
she jokes. She rushes past other stations with long thin metal
tubes wrapped in tin foil transporting X-rays from the syn-
chrotron in ultra-high vacuum, and finally arrives at her own
small space. Specially designed optics capture the infrared
beam—extremely bright but low photon energy—from the
ALS, where it is bounced through a series of 20 mirrors and
beamed through a combination optical/infrared microscope
at the microscope stage incubator containing her experimen-
tal microbes. The bacteria inside, magnified about 300 times,
appear on a computer monitor on the table nearby, along with
masses of data showing the infrared measurements. Holman
says she’s eager every day to see what her microbes are up to.
“We really get attached to our bacteria,” she admits. “I always
feel bad when we have to kill them.”
Hoi-Ying Holman (center) with colleagues Wayne McKinney (left) and Michael
Martin (right) at their experimental station inside LBL’s Advanced Light Source.
(Photo: LBNL)
But the amazing thing about Holman’s technique is that she doesn’t
have to kill them at all. Most other methods of monitoring cellular
mechanisms involve extracting the contents of cells, destroying them
in the process, or using dyes or other agents that can affect the cell
chemistry. Holman’s infrared beam allows her and her colleagues
to observe molecular reactions occurring inside cells in real-time,
like a live movie. By analyzing the detailed characteristics of the
infrared absorption bands produced by different compounds,

Unique spectral signatures show the spatial distribution of the biological protein amide II, indicative of the indigenous microorganisms (left), at the same
location as reduced chromium, Cr(III) (right) after a four month period. (Image: LBNL)

BERKELEY
science 37
review
 Feature
Professor Norman Terry in the greenhouse. Terry is genetically modifying
plants to remove selenium from the environment more efficiently. (Photo:
Martin Sundberg)
Holman can tell exactly what chemicals are present and what reac-
tions are taking place. Initial tests with human kidney cells demon-
strated that the low energy of the infrared light does not affect cel-
lular functions. The brightness of the ALS beam provides a spatial
resolution of 2 to 12 microns, allowing the team to focus on a small
colony of bacteria.
Holman started working on infrared spectromicroscopy in 1997.
By early 1999, she had used her new method to follow the re-
duction of toxic metals among natural communities of bacteria
taken from the Department of Energy’s Idaho National Engi-
neering and Environmental Laboratory’s Radioactive Waste Man-
agement Complex. The team monitored and studied the bio-
geochemical transformation of hexavalent chromium, Cr(VI),
into trivalent chromium, Cr(III), as it was occurring inside the
living system. “We were the first on the planet to watch the
bacteria in action as they detoxified chromium, like a play on
stage,” says Holman. “It was very exciting.”
The bioremediation of a toxic metal like chromium is very differ-
ent from the bioremediation of organic pollutants such as petro-
leum products or organic solvents. While these are complex or-
ganic molecules that can be broken down into harmless pieces, heavy
BERKELEY
science
review
metals are atomic structures that can not be broken down further.
Instead, bioremediation uses organisms to stabilize and immobilize
heavy metals by changing their species–––by adding or removing
electrons to change the metal’s valence state. Different species of
heavy metals have different chemical, physical, and biological prop-
erties that have different impacts on public health.
Hexavalent chromium, a common industrial chemical used for
chrome plating, dyes, leather tanning, and wood preserving, is
found at two-thirds of the EPA’s National Priority List sites. It is
carcinogenic, mutagenic, highly soluble, and biologically active,
easily crossing cell membranes and disrupting DNA replication
inside cells. Trivalent chromium, on the other hand, is considered
less dangerous because it is much less soluble and can not cross
cell membranes.
At the Idaho waste site, high-level radioactive waste has been
stored for more than 40 years, creating a toxic soup of inor-
ganic metallic ions like hexavalent chromium, other inorganic
ions, and radionuclides, as well as petroleum hydrocarbons and
other volatile organic compounds. Over time, these toxins have
seeped deep into the porous volcanic rock beneath the site. At
the time of Holman’s research, it was well known that the
amounts of hexavalent chromium and other toxic metals were
being reduced by natural processes at contaminated sites, but
no one knew for certain whether this was due to a microbio-
logical process or a geochemical reaction with the rocks.
To find out, Holman and her colleagues compared a sterilized sample
rock under controlled conditions with another sample harboring a
living community of Arthrobacter oxydans—bacteria that effectively
reduce hexavalent chromium. Using the ALS beam, Holman found
that, after five days, no reduction was taking place on the sterile
rock sample and only small changes were measured in the sample
with A. oxydans. But when the researchers added a weak solution of
toluene, a petroleum chemical also found at the waste site, to the
colonized sample, the infrared spectromicroscopy showed evidence
of the reduction of hexavalent chromium, as well as degradation of
the toluene where the bacteria were located.
To make sure the results accurately reflected what would occur in
the field, Holman brought thin slices of basalt rock taken from 75
meters beneath the waste site, complete with their native micro-
bial communities, into the lab. She exposed them to a hexavalent
chromium solution and toluene vapor and watched them carefully
under the ALS beam. After four months, spectromicroscopic graphic
images showed colonies of bacteria at the same location as new triva-
lent chromium, where hexavalent chromium had vanished. The re-
searchers also saw two new peaks that they believe were caused by
38
pentavalent chromium, Cr(V), an unstable, intermediate species of
the metal. “Some of the intermediate compounds we saw can be
more toxic,” Holman explains. “But in the past there was no way to
see them because they are only stable in live systems.”
Holman has also used her new technique to examine the role mi-
croorganisms play in the detoxification of other carcinogens, the
degradation of polyaromatic hydrocarbons, and the reduction of
uranium. Her work has generated a great deal of interest in infra-
red spectromicroscopy, both nationally and internationally. Holman
has even debunked some overstated bioremediation claims made
on behalf of certain microorganisms. On one occasion she was given
a sample of microbes that were supposed to be remediating con-
taminated soil, but didn’t seem to be having any effect. Holman put
them under the infrared beam and confirmed that, indeed, they
were not doing anything. “They were just sitting there,” she says.
“They weren’t even dividing.” She later determined that the organ-
isms were meant to work in aquatic environments and were inap-
propriate for their intended project, a discovery that saved a great
deal of time and money.
Solutions from the greenhouse
Bacteria aren’t the only ones doing the dirty work.
Phytoremediation—the use of plants to reduce or remove toxic
chemicals from the environment—also holds great promise.
Norman Terry, professor of environmental plant physiology at UC
Berkeley’s Department of Plant and Microbial Biology, has
focused his research on the phytoremediation of selenium
for almost 15 years. He has discovered that using wetland
plants such as cord grass, salt marsh bulrush, and rabbitfoot
brush can be an effective, low-cost way to remove selenium
from oil-refinery wastewater and agricultural drainage wa-
ter. The key to his research is a mechanism in some plants
and microorganisms that can turn selenium salts into the gas
dimethyl selenide—a process known as volatilization.
sulfate, which plants need to form vital sulfur-containing proteins.
“The problem is that selenium compounds are chemically analo-
gous to sulfur compounds,” said Terry, “and they tend to go through
the same sulfate assimilation pathway in kind of a competitive way.”
Because it mimics sulfur, selenium is also incorporated into sulfur-
containing proteins, resulting in toxicity to plants and to the ani-
mals that ingest them. But not all of these elements are stored in
the plant. Fortuitously, the same metabolic pathway that changes
sulfur into dimethyl sulfide gas also enables plants to change sele-
nium into dimethyl selenide gas.
This chance volatilization of selenium is of tremendous benefit when
it comes to phytoremediation. Volatilization takes selenium out of
the contaminated environment—the sediments, the water, and the
biomatter—and puts it into the atmosphere. Eventually the sele-
nium will come back down to earth, but since it is only harmful at
high concentrations, this is generally not a problem. “Volatilization
of selenium is an excellent way of cleaning it out of the system
because it takes it completely out of a local area and puts it in an-
other area where probably it’s going to be helpful rather than hurt-
ful,” says Terry. “Toxicity spots are very, very localized. In the case
of California, if it comes down somewhere else, chances are it will
come down in an area that is suffering from selenium deficiency.”
Terry got his first field test of selenium volatilization in 1995 on an
experimental wetland in Richmond run by the Chevron Corpora-
tion.The oil company had determined that their 35-hectacre marsh

Selenium is a naturally occurring metalloid that is abundant

in the soils of California’s central valley. The detrimental ef-
fects of selenium made headlines in 1983, when scientists
discovered that high concentrations of selenium at the
Kesterson Reservoir in the San Joaquin Valley were causing
dead and deformed waterfowl. But in low concentrations,
selenium is harmless and in trace amounts it is actually a
vital nutrient.

Terry started studying how volatilization occurs in plants in This experimental wetland in front of Chevron’s Richmond oil refinery can reduce the
amount of selenium in the refinery wastewater by about three quarters. (Photo: Paul
1989 and found that, although the plants do not use sele- Kagawa/Chevron)
nium, the structure of selenium salt is very similar to that of

BERKELEY
science 39
review
 Feature
was successfully reducing the amount of selenium in its refinery Terry sees enormous potential in using engineered wetlands to re-
wastewater by about three quarters, but the accumulation of con- duce the selenium concentration in agricultural drainage water. But
centrations in the sediments and plant biomatter could only ac- how would these experimental marshes avoid turning into another
count for about 70% of the missing environmental disaster like
selenium.They called in Terry to see Kesterson? One major difference is
if volatilization could account for the that the filtering marsh would be a
missing metalloid. Using small flow-through system instead of a
Plexiglas chambers on meter-square closed system like the Kesterson res-
test plots in the wetland, Terry and ervoir. “In Kesterson, you were just
his colleagues were able to capture pouring selenium in and giving it no-
dimethyl selenide being released where to go,”Terry explains, “whereas
from the plants and show that the with the flow-through systems you’re
plants and microbes were indeed trying to filter it out, but you’ve got a
volatilizing at a healthy rate. The re- constant flow of water through there.
search found that as much as 10– That stops the selenium from build-
30% of the selenium could be vola- ing up to super-high levels.” As part
tilized from the marsh. of a bioremediation project, the marsh
Terry thought that if wetlands could would also be carefully monitored.
help remediate oil refinery water, Once the sediments and the plants
they could probably help out with ag- became saturated with selenium, the
ricultural water in California’s Cen- marsh would have to be dried out, the
tral Valley as well, where heavy irri- selenium-filled plants mowed down,
gation causes selenium to leach out and the sediments either dug up and
of the soil and into the drainage wa- hauled away as toxic waste or
ter. After the Kesterson disaster, remediated further with more sele-
farmers were no longer allowed to nium-tolerant plants.
flush water into drains and instead
were forced to use huge evaporation Terry is trying to develop new su-
ponds, which can be as large as 300 per-selenium-loving plants using ge-
hectares. As the water evaporates, it netic engineering. “The problem is
leaves behind highly concentrated that plants at present work kind of
toxic salts. Marsh plants like this rabbit foot grass (Polypogon monspeliensis) slowly. It would take quite a few
are very successful at removing selenium from the surrounding water.
In one summer month during their research project in Corcoran,
years for them to really significantly
In 1996,Terry’s team joined the UC CA, Terry’s team found that the test wetland cell filled with rabbit draw down the selenium pollution
Salinity/Drainage Program—a joint foot grass volatilized nearly half the selenium entering it. (Photo: in the soil,” he says. “What you want
Elkhorn Slough Foundation)
research project involving scientists is to genetically engineer plants that
from UC Berkeley, Davis, and Riv- will rapidly speed up this process, so
erside as well as the Tulare Lake Drainage District in Corcoran, instead of taking 10 years or 50 years or 100 years to do it, you
CA. The researchers set up 10 quarter-acre wetland cells in the want them to do it in 2 or 3 years.”
Central Valley and planted them with species such as cord grass,
salt marsh bulrush, and rabbitfoot grass. Terry’s research showed Terry has spent the last eight years genetically engineering plants
that the miniature wetlands removed an average of 69% of the sele- such as Indian mustard (Brassica juncea) to tolerate, accumulate, and
nium from the inflow, with most of that being stored in the sedi- volatilize high levels of selenium, using genes from E. coli bacteria
ments. But volatilization was also occurring, especially among cer- or other plants such as Arabidopsis. He is currently conducting the
tain plant species. The team found that in one summer month, the first field trials in the United States of plants genetically modified
test wetland cell filled with rabbitfoot grass volatilized nearly half for phytoremediation, using genetically enhanced Indian mustard
the selenium entering it. to remediate highly contaminated agricultural drainage sediments
in a joint project with the US Department of Agriculture in Parlier.

BERKELEY
science 40
review
Terry has also been working with a plant called Astragalus bisulcatus
or the two-grooved poison vetch—a small, slow-growing plant from
the North American prairies that tolerates and accumulates ex-
tremely high levels of selenium. His team has identified one of the
genes that give Astragalus bisulcatus its high tolerance for selenium
and has transplanted this gene into Arabidopsis and the fast-grow-
ing, high biomass Indian mustard.
“What we will eventually do,” says Terry, “is to use the genes that
we got from Astragalus bisulcatus and combine them with the genes
that we’ve gotten from these other places and pop them into the
same plant, trying to create kind of a super plant for
phytoremediation. That’s the goal.” He hopes that, using genetic
engineering, he will be able to create plants that can take up 10 to
100 times as much selenium as natural plants, greatly enhancing
the chance for making plant-based remediation systems function
quickly and economically.
The future of bioremediation
Over the last 10 years, significant scientific advances have been made
in the field of bioremediation thanks in part to scientists like Coates,
Holman, Terry, and others at UC Berkeley and Lawrence Berkeley
Labs. Chemicals that were previously believed to be persistent in
the environment can now be broken down or contained through
proven bioremediation techniques. Bioremediation has moved far
beyond its initial application to petroleum contaminants to the
remediation of organic and inorganic solvents, heavy metals, and
radionuclides. Scientists hope that bioremediation will soon be suc-
cessfully applied to the complex mix of toxics commonly found at
waste disposal sites and Superfund National Priority List sites
throughout the United States.
There are still many questions to be answered in the field of
bioremediation—whether to use genetically modified organisms,
when to add organisms to an area at all, and when regulators can
simply sit back and monitor, letting nature take its course. But
each year of research brings new discoveries and a better under-
standing of the hydrology and biogeochemistry of contaminated
areas, of new species of microorganisms with unique metabolic
pathways to break down toxins, and of the various environmental
conditions such as oxygen level, pH, and available nutrients re-
quired to make these microorganisms function. Bioremediation
is no silver bullet, but it is an important tool, offering a rapid,
cost-effective, and environmentally friendly way to clean up con-
taminated environments.
Carol Hunter is a second-year graduate student at
the UC Berkeley School of Journalism.

Want to know more?

Environmental Factors That Control Microbial Perchlorate Reduction, K Swades et al., Applied Environmen-
tal Microbiology (2002); Vol. 68, pages 4425–4430.

Anaerobic benzene oxidation coupled to nitrate reduction in pure culture by two strains of
Dechloromonas, JD Coates et al., Nature (2001); Vol. 411, pages 1039–1043.

Microbial Reduction of Hexavalent Chromium:

http://www-als.lbl.gov/als/science/sci_archive/bioremed.html.

Selenium Removal by Constructed Wetlands: Quantitative Importance of Biological Volatilization in the

Treatment of Selenium-Laden Agricultural Drainage Water, Z-Q Lin and N Terry, Environmental Science &
Technology (2003); Vol. 37, pages 606–615.

BERKELEY
science 41
review
 University

FAULTY TOWERS
Shoring up the foundations of excellence
Jessica H. Marshall

A shear wall partially

covers the end of
Latimer Hall. The
building’s face carries
a new frame: a few
feet of reinforced con-
crete that follows the
building’s original
rectangular-grid struc-
ture. (Photo: Karen
Levy)

efore I left civil engineering professor Jack Moehle’s office,

B he had broken a piece of chalk, bent a paper clip back and

forth repeatedly, slowly torn a piece of paper, squished an
eraser between his fingers, and jiggled a disconnected computer
power cord. While it might sound like he has a fidgeting problem,
each of these acts illustrated a seismic design principle.

UC Berkeley’s proximity to the Hayward fault—which runs through

section KK of Memorial Stadium—gives campus buildings a po-
tential “fidgeting problem.” Since 1997, when the university allo-
cated $1 billion over 20 years for seismic safety projects, the cam-
pus community has been acutely aware of efforts to protect campus Professor Moehle points out a damper in the bracing on
buildings from earthquake damage; the jack hammering and chain Hildebrand Hall. The damper allows the brace to change length
link fences are hard to ignore. during an earthquake without buckling. (Photo: Karen Levy)

So, what’s the matter with these buildings? Many of the campus’s
original buildings are made of stone or brick. These notoriously
brittle materials perform so poorly in earthquakes that California
banned un-reinforced masonry construction in 1933. Campus build-
ings built in the 1950s and 60s are also seismic offenders. “The
engineering aesthetic in the 1960s was beautiful minimalism,”
says architecture professor Stephen Tobriner, a historian of seismic
construction. “However, in earthquakes, what you really want is
not the slightest, lightest, most minimal structure, but something
that has redundancies.”
According to Moehle, who also directs the Pacific Earthquake En-
gineering Research Center, engineers in the 1960s “didn’t really
appreciate how big the forces during an earthquake could be. They
thought the design load was maybe one-tenth of what we now know
it to be.” Most of the 1960s-era buildings are made of reinforced
concrete. They bear the load of gravity well, but the reinforced-
concrete designs of the time were not sufficiently reinforced to re-
sist earthquakes. Moehle demonstrates with a piece of chalk: while
the chalk can withstand significant vertical pressure, it snaps easily
under a horizontal force.

BERKELEY
science 42
review

“The last picture we
want on the news is the
Campanile collapsed.”
Not surprisingly, many of these minimalist, reinforced-concrete
structures were rated as “very poor” when the university evaluated
the seismic safety of its buildings in 1997. The evaluation showed
that 27% of the square footage on campus needed retrofitting. At
the time of the report, buildings including South Hall, Wheeler,
Dwinelle, Doe Library, and the Unit 1 and 2 Residence Halls had
already been retrofitted. Since then, 2.4 million square feet have
been retrofitted including Barker, Barrows, Hildebrand, Latimer,
and Wurster Halls. These retrofits cost $700 million and represent
only half of the area identified in the 1997 evaluation.
An additional 950,000 square feet and $301 million are slated for
retrofitting in the next five years, including this past spring’s demo-
lition of Stanley Hall and the scheduled demolition of the north
side of Davis Hall. Both buildings will be rebuilt. The final 1.4 mil-
lion square feet of retrofitting may take a while, since it is distrib-
uted over 70 buildings on central campus and at outlying sites.
The cost of retrofitting is high, but so is the potential cost of seis-
mic damage. According to architecture professor Mary Comerio’s
2000 report, “The Economic Benefits of a Disaster-Resistant Uni-
versity,” the campus could sustain losses from $600 million to $3.5
billion depending upon the size of an earthquake. The higher cost
would result from an earthquake of the size expected to occur near
Berkeley once every 1000 years.
In the face of such low probability but such high stakes, the univer-
sity must decide what retrofitting standard is good enough. Retro-
fitting designs are tested against two hypothetical earthquakes. The
building is designed not to collapse in the case of the larger earth-
quake, of a size expected to occur every 2000 years. For an earth-
quake recurring with a 500-year period—equivalent to 7.0 on the
Richter scale centered on the Hayward fault—the retrofit is also
designed to protect “life safety.” For instance, secured objects should
remain secured, and not fall on people.
The retrofitting projects completed since 1997 have undoubtedly
mitigated the potential damage, but the cost of direct physical
damage may be the least of the university’s concerns in the event
of a large seismic event. Comerio’s report also noted that 75% of
UC Berkeley’s research funds from corporations and the federal
government are spent in just seventeen buildings, and that half to
BERKELEY
science
review
three-quarters of the space in those buildings is “likely to be sig-
nificantly damaged or closed after a major seismic event.” The
report continues that such damage would “seriously disrupt ex-
isting research and limit the capacity to take on new research for
a long period.” Would faculty leave under such circumstances?
Moving and shaking
While much retrofitting design is done using computer
models, “real-life tests” are done here on campus. Struc-
tures are tested on the second floor of Davis Hall, situated
on a “strong floor”—the top of a very rigid box, one
story tall, built down into the floor below. Structures
(beams, columns, etc.) can be bolted to this floor. Giant
hydraulic actuators attached to the strong floor can push
a structure with forces of the same magnitude found in an
earthquake. The floor’s stiffness minimizes its absorption
of energy, so the element being tested receives the full
input energy of the earthquake. Through these tests, re-
searchers can find out how the structure breaks and what
makes it tougher.
At the Richmond field station, a 20 by 20 foot shaking
table is used to simulate earthquakes in “real time,” and
to test the performance of structures in the simulated earth-
quakes. A computer tells actuators underneath the table
to push the table in a particular pattern to mimic a pre-
scribed earthquake’s motion. The table can move up to
eight inches (though a real earthquake can move farther)
and is surrounded by a rubber bladder. The size of the
table means that test structures are not built to full scale.
Since a building’s response depends on the axial load on
columns, and since length and volume scale differently,
weight must be added to the tops of structures so the ratio
of size to mass matches that of a real building. The shake
table has been used to compare the performance of a
three-story wood-framed apartment building with and
without seismic retrofitting braces. It has also been used
to test reinforced-concrete columns that support a six-story
office building. To determine how the contents of a build-
ing will behave during an earthquake, a wet lab was
built on the shake table, complete with shelves, benches,
refrigerators, and equipment, including the typical an-
choring devices used in labs.
While it might sound fun to stand on the table and expe-
rience an earthquake in a controlled setting, it’s not al-
lowed. The shake table is capable of generating a shock
that could break a person’s legs!
http://peer.berkeley.edu/~elwood/research/shake_table.htm
43
 University
Would students? The potential impact on Cal’s academic standing contend with. On the other hand, it can be cumbersome to imple-
is difficult to quantify. ment, and because it covers the building’s surface, it is often inap-
propriate for historical buildings.
Retrofitting 1A
Another way to improve a building’s performance is to reduce its
A number of retrofitting tools have been used to ensure that cam- mass by removing stories. Naturally, story removal is not popular
pus buildings can stand up to earthquakes. Steel braces strengthen since square footage is typically scarce, especially at Berkeley. Even
Cal’s residence halls, and reinforced concrete shear walls stiffen temporary loss of square footage disrupts campus operations—a
the ends of Barrows, Latimer, and Wurster Halls. An entirely new fact that plays a major role in the choice of a retrofit design. Inte-
frame of reinforced concrete, like a strong exoskeleton, supports rior retrofitting plans often require that occupants relocate during
the north and south faces of Latimer Hall. Much of the interior of construction, which can be costly—state retrofitting funds do not
South Hall was removed and con- cover these costs—and logisti-
crete walls and steel anchors were Hearst Mining Building is designed cally difficult. Barrows and
built on the interior. Comerio sum- Latimer Halls have very high oc-
marizes campus-retrofitting strate- to move up to 30 inches. cupancy, and for both buildings
gies this way: “You can put a build- an exterior retrofit plan was
ing in a cage, as in the case of University Hall.You can put a cage chosen, allowing the building to remain fully, if not happily, occu-
inside the building, as in the case of South Hall. In the case of pied throughout the retrofit project.
Wurster and Barrows, they grabbed the building on either end
and stiffened the ends a lot.” The high density of campus buildings also constrains retrofitting
operations. Lay-down space for construction materials is scarce,
All of these retrofitting approaches are designed to stiffen the build- and it can be difficult to bring in large equipment. When digging
ing, reducing the amount that it sways in an earthquake.While stiff- holes for pilings to support Wurster Hall, engineers hit rock. The
ening a building makes it able to resist larger forces, it makes for a rock-boring equipment would not fit on site, so one of the holes
rougher ride in an earthquake: the contents of the building, often had to be dug by hand. “There was one guy with a bucket and shovel,”
including expensive equipment or hazardous materials, will shake says Comerio, “He literally dug and sent the dirt up in a bucket,
harder. Stiffening also reduces the building’s natural period of vi- while shoring the sides as he went. It took about two months.”
bration, typically bringing it close to half of a second. Coinciden-
tally, the largest forces of an earthquake tend to occur at this pe- Hearst Mining Building: A masterpiece of architecture—
riod. So, while a stiffer building is stronger, it experiences larger and of seismic retrofitting
forces than a “softer” building. The goal, then, is to stiffen a build-
ing, but not too much. In the 1970s, Berkeley civil engineering professor James Kelly pio-
neered base isolation, a major innovation in retrofitting design.This
Brittle building elements, like the insufficiently designed reinforced approach to protecting a building makes it less stiff: it is set atop
concrete columns of the 1960s, can be protected by jacketing them “base isolators,” which are essentially giant pencil erasers that allow
with “shotcrete” (a sprayable concrete), steel, or fiberglass. Jackets the entire building to slide back and forth. The base isolators in-
provide tough skins; even if a crack develops internally, the struc- crease the period of the building’s oscillation to much longer than
ture will be held together by the jacket. Jacketing does not change the half-second typical of an earthquake’s largest motions.The Hearst
the stiffness of the building; it just makes it better able to withstand Mining Building, the only base-isolated structure on campus, has a
the damage that an earthquake doles out. Fiberglass sheets—simi- period of around 2.75 seconds. Hal Davis, an engineer at Ruther-
lar to casting material—were tightly wrapped around columns in ford and Chekene, the engineering company responsible for retro-
the interior of Wurster Hall. fitting the Hearst Mining Building, says that in an earthquake the
building “moves very slowly and deliberately back and forth.” But it
The choice of a retrofitting design balances cost, ease of installa- must go a long way to move so slowly: Hearst Mining Building is
tion, aesthetics, and historical importance. Steel braces leave win- designed to move up to 30 inches.
dows relatively unobstructed, but are very noticeable. Reinforced
concrete walls and new frames can be designed to fit into a struc- The Hearst Mining Building was designed by John Galen Howard,
ture well, but they often block light. Jacketing doesn’t change the the architect responsible for many early campus buildings, and is
floor plan of a building since there are no new walls or beams to listed as a national landmark by the American Institute of Archi-

BERKELEY
science 44
review
Top Left: The ceiling of the main hall of the Hearst
Mining Building, an American Institute of Architects'
national landmark.(Photo: William Porter)

Top Right: Installation of steel braces and dampers

to stiffen Hildebrand Hall. (Photo: Jack Moehle)

Bottom Left: One of 134 rubber base isolators on

which the Hearst Mining Building rests. (Photo: Jack
Moehle)

Bottom Right: The "moat" surrounding the Hearst

Mining Building allows space for the building to
move up to 30 inches.(Photo: Karen Levy)

BERKELEY
science 45
review
 University
tects. Its masonry construction made it particularly susceptible to lowing the brace to change its length during an earthquake, ab-
seismic damage, and according to Davis, there was no way to stiffen sorbing earthquake energy without buckling.
the building, inside or out, without affecting its architecture. The
Hearst Mining Building’s historical importance led the University Founded safe
to spend $90.6 million to retrofit it.
If retrofitting the Hearst Mining Building is worth $90.6 million,
To install the base isolators, workers first “dug out a big bathtub what about Cal’s signature structure, the Campanile? The 1997 seis-
underneath the building, and shored up the building using pipe mic evaluation rated it “poor” and estimated that it would cost $4
columns,” says Moehle. Then they built a new,
heavily reinforced foundation on the bottom of
the “bathtub.” They installed 134 rubber isola-
Oh, where does all the money go?
tors on the foundation, and removed the pipe
columns, transferring the building’s weight onto
Seismic damage could pose a greater danger to Cal’s research than to its
the isolators.
students. Comerio’s report, “The Economic Benefits of a Disaster-Resistant
University,” found that the funds with which corporations and the federal
Moehle turns a white rubber eraser on end to government sponsor UC Berkeley’s research are concentrated in just a few
demonstrate how the flexibility of the rubber al- buildings. Fully three quarters of these funds go to the seventeen buildings
lows the bottom of the base isolator to move in- below. The top seven buildings house half of sponsored research.
dependently of the top. “You can see how this
eraser can wiggle back and forth in the horizon- Building Sponsored Research*1 1997 Seismic Rating*2
tal direction. However,” he adds, “if you squeeze (Million $)
on it in the vertical direction, it tends to bulge.” Cory 31.12 Fair
Because these isolators support the entire weight LSB Addition 21.53 Good
Koshland 17.06 Good
of the building, whether it’s shaking or not, the
Soda 12.73 Good
isolators were built in layers, alternating steel Hildebrand 7.78 Very Poor
3

plates with rubber slabs to prevent bulging, and Warren 7.74 Poor
encapsulated in rubber.The resulting isolators are Barker 7.42 Retrofit in progress
each 40 inches across, 18 inches high, and weigh Etcheverry 7.26 Good
3
4,500 pounds. Latimer 6.87 Poor
University 6.02 Already Retrofitted
Because the Hearst Mining Building is so close Davis (old, new) 5.36 Fair, Poor
to the fault, engineers at Rutherford and Stanley 4.16 Poor
Chekene feared that the building might displace Donner Lab (old, new) 4.14 Good, Poor
VLSB 3.86 Good
more than the allowable 30 inches in a large
Mclaughlin 3.30 Poor
earthquake.Workers installed big, horizontal hy- Morgan 3.01 Good
draulic pistons, like the shock absorbers on a Tolman 2.88 Poor
car, connected at one end to the “bathtub foun-
dation” and at the other to the building. In an * For 1994–99, in 1999 dollars.
earthquake, part of the foundation’s lateral en-
ergy will be dissipated into the fluid in the pis- 1 Mary C. Comerio, “The Economic Benefits of a Disaster Resistant Univer-
ton, reducing the displacement of the building. sity: Earthquake Loss Estimation for UC Berkeley” (April 1, 2000). Institute
The dampers dissipate this energy as heat, and of Urban & Regional Development. IURD Working Paper Series. Paper
could reach 120–150oC. According to Davis, the WP-2000–02. http://repositories.cdlib.org/iurd/wps/WP-2000-02.
use of dampers in a variety of situations is a real
2 Seismic Action Plan for Facilities Enhancement and Renewal. University
improvement in seismic engineering over the of California, Berkeley (1997).http://www.berkeley.edu/SAFER/ .
last several years. For instance, in Hildebrand
Hall dampers were installed in steel braces, al- 3 Retrofitted since 1997.

BERKELEY
science 46
review
million to retrofit. Its low occupancy, though, puts it low on the list
of retrofitting priorities. Looking out his Davis Hall window at the
Campanile, Moehle says, “I think it’s pretty safe. It has a masonry
structure, but it’s got a steel frame.” Even so, “it’s such a symbol,”
says Comerio. “The last picture we want on the news is the Campa-
nile collapsed.”

The flurry of “seismic activity” on campus might lead one to think

that until recently, the University ignored its shaky location just a
few hundred feet from the Hayward fault. Actually, earthquakes were
a concern during the construction of Cal’s first building, South Hall,
a fact that came to light as the seismic retrofitting of South Hall
proceeded in the 1980s. Engineers discovered that the brick struc-
ture was heavily reinforced with iron. According to Tobriner, “This
kind of reinforcement was exceptional and very peculiar. It revealed
beyond a doubt that the original architect, David Farquharson, had
attempted to build a seismically resistant brick structure in 1870.”
The first regents insisted on such a building after seeing the damage
caused by the 1868 earthquake. “One of our first cares,” they wrote,
“should be to make our buildings as safe as possible for the youths
who may be confided in our charge.” Armed with modern under-
standing and new technology, the University may now be making
good on the regents’ directive.
Jessica H. Marshall is a graduate student in the
.
Department of Chemical Engineering

BERKELEY
science 47
review
Quanta (heard on campus)
"Most of your experiences are miserable. You wouldn't want to put them
into long-term memory. Like the boring lectures you hear—who wants to
remember this tomorrow?"
[Commenting on why organisms evolved mechanisms inhibiting the
transcription of experiences into long-term memory.]

Eric R. Kandel, 2000 Nobel Laureate, Physiology or Medicine

April 23, 2003
"The Long and Short of Long-Term Memory: Memory and Its Disorders"

Lecture sponsored by Helen Wills Neuroscience Institute

"My God, you're disrespectful! That's good."

Ignacio Chapela, Department of Environmental Science, Policy & Management

August 20, 2003
Graduate student orientation keynote address

"If you never want to be criticized and never want to make

a mistake, never do anything in science."

Norman Borlaug, 1970 Nobel Laureate, Peace Prize

July 10, 2003
"The Story of Norman Borlaug: 60 Years Fighting Hunger"
Lecture sponsored by College of Natural Resources

"We cannot repeat the early universe due to severe

funding limitations."

Boris Kayser, Fermi National Accelerator Laboratory

April 14, 2003
"The Neutrino World: Present and Future"
Department of Physics Colloquium

BERKELEY
science 48
review
 Natha
n
worki Bramall i

The Backpage
ng w s
Green ith Bu a gradua
land ford te stu
pretty an Pri dent
busy, d on the ce. They in p
but he Antar do th hysics at
still ha ctic ic eir re Cal,
s time e s sea
Dear Mum, to wri heet. It ke rch in
te hom ep
e to m s him
om.
Guess what? Ryan Bay and I are at Summit Camp on top of Greenland’s ice sheet (72° 34’N 38° 29’W)!
We’re here to deploy our “dust logger” to detect dust and volcanic ash within the ice.We lower it three kilometers
down a six-inch-wide borehole and it emits light into the ice. A phototube detects how much of that light is
scattered back by the dust and ash, telling us about the earth’s past climate. The ice is layered like sediment, with
the oldest climate record near the bottom of the hole.

It’s a bit warmer here than you might think, but it can get cold! Taking data the other night was the worst,
as the temperature was -35°C with a sustained 20–25 knot wind and ice fog. Brrr.We had to stand outside on
scaffolding all night and the next morning. Afterwards, I was glad to collapse in my Arctic Oven, sort of a sturdy 4-
season backpacking tent. It may sound uncomfortable, but I’ve been sleeping like a baby on top of these three
kilometers of snow and ice. It’s so peaceful and quiet!Well … except for the other night when a little storm came
up and buried my tent—excavating snow in the midst of a blow during the wee morning hours isn’t much fun.

Summit Camp houses seven staff and about a dozen scientists, who are all very cool and a bit quirky.The
camp cook was trained as a pastry chef so she makes sure that we always have at least one dessert for every meal!
Despite the cold and hard work, I’m actually packing on a few pounds.To help out with chores, I’ve become the
“water boy,” which means I’m responsible for the camp’s water supply.Waste heat from the camp’s diesel generator
melts snow into potable water. I shovel mounds of snow into a hopper, let it melt, and transport the resulting water
to the main holding tank using a snow machine attached to a sled-mounted, 200-gallon container.

“Ryan deploying
the dust logger.”
Of course it hasn’t all been hard work! The other day we had a fourth of July
BBQ, complete with volleyball and booze! I can’t remember who won the game, but I
vaguely recollect a few Bloody Marys …
Love,
Nathan
“Me with the Tucker SnoCat.”

(Photo: Nathan Bramall)

“The New York Air National Guard bringing us in.”