THE NEW YORK TIMES, TUESDAY, JULY 30, 2019 N D3
greens, blue-greens and oranges tinged
with purple.
“Deep-sea creatures must find these col-
ored lights irresistible as they flicker and
flash faintly in the dark waters,” he wrote.
Speciation has produced a great diversity
of protruding lights and rods. Some angler-
fish have a long barbell extending from the
lower jaw as well as a rod above. One
species, Lasiognathus saccostoma, bears
not only a movable rod but extending from
it a line, a float, a lighted bait and three
hooks. The hooks, Dr. Idyll wrote, “are, alas,
not actually for catching prey” but simply
ornamental.
Anglerfish, he noted, are “rarely as large
as a man’s fist.” But one specimen, from a
depth of 2.2 miles off West Africa, was a foot
and a half long. It was also unusual in hav-
ing its glowing bait conveniently located in-
side its enormous mouth.
The largest known deep anglers are the
warty seadevils. The females typically run
about two and a half feet long, and free-
swimming males less than a half inch.
The examination of stomach contents has
revealed that anglers eat shrimplike ani-
mals, squids, worms and lanternfish, a com-
mon type of deep-sea fish with large eyes
and a highly developed visual system that
apparently can detect colors.
When an anglerfish suddenly opens its
giant mouth, Dr. Idyll wrote, the resulting
suction pulls in the luckless victim. After
the jaw slams shut, small teeth on the floor
of the mouth and throat deliver the meal to
the fish’s belly.
Videotaped Expeditions
The first undersea video recordings of the
creatures were made in 1999, and caught a
surprise. Scientists from the Woods Hole
Oceanographic Institution in Massachu-
setts had set up an undersea observatory in
the North Pacific between California and
Hawaii. It lay more than three miles down.
A seven-foot-long tethered robot named
Jason was lowered to survey the surround-
ing area. Soon, its operators were startled to
see a fish drifting in the bottom current up-
side-down, with its extremely long rod
hanging downward in a graceful, forward-
arching curve. Unexpectedly, they found
two other fish similarly upended.
Jon A. Moore, a fisheries biologist at Flor-
ida Atlantic University, identified the crea-
tures as whipnose anglerfish, although of
an unknown species. In a 2002 paper, he
wrote that they apparently were looking for
prey by trolling over the seabed with glow-
ing bait. Visible just below the fish, he noted,
were “numerous small burrows.”
Dr. Moore said the video represented “the
first time anyone had seen” any kind of
whipnose in its own dark habitat. He added
that, despite the intervening years, the
question of what the fish were pursuing on
the Pacific floor remained a mystery.
The Monterey research institute — in
Moss Landing, Calif., at the midpoint of the
bay shoreline — was established in 1987 by
David Packard, the billionaire co-founder of
Hewlett-Packard and a creator of Silicon
Valley. It has built generations of increas-
ingly smart, fast robots that probe the
nearby waters.
In 2005, nearly a mile deep in the waters
off Monterey, institute scientists were fly-
ing a tethered robot when they tracked an
angler for a record 24 minutes. The result-
ing paper, by Dr. Pietsch and another Uni-
versity of Washington scientist, detailed a
series of behaviors, from swimming bursts
to long bouts of drifting.
Over all, they wrote, their observations
supported the theory that “these animals
are lethargic, lie-and-wait predators.”
The range of known behaviors grew
larger when institute scientists investi-
gated seamount chains west of the Mon-
terey Canyon. Expeditions in 2002 and 2010
videotaped odd anglers with a bulbous
body, a shaggy lure and fins that the fish
used to walk along the rocky seabed. The
scientists speculated that walking disturbs
the seawater less than swimming does, re-
ducing the chances of startling nearby prey.
The newest video to go public was made
off the Azores by a research team from the
Rebikoff-Niggeler Foundation, based on the
island of Horta. In 2016, a half-mile down,
Kirsten and Joachim Jakobsen were return-
ing to the surface in their submersible when
they spotted a female angler “resplendent
with bioluminescent lights,” as Science
magazine described the fish. It was later
identified as a fanfin seadevil, a ghoul of the
deep with a bushy lure.
The team also videotaped a dwarf male
fused to her underside — a permanent
sperm donor. Males of that species hadnever before been seen by humans.Denizens of the Abyss
Young male anglerfish face the challenge of
finding a mate in the ocean’s vastness. They
have large olfactory organs, which suggests
that suitors follow a trail of pheromones. If
courtship is successful, the male fuses per-
manently to the female, and their tissues
and circulatory systems commingle.
In the case of the Azores discovery, “the
size of her belly indicates that she was grav-
id,” or full of offspring, Kirsten Jakobsen
said in an email.
The team was able to track the pair for 25
minutes; what mesmerized was not only
the procreative union but also the halo of fil-
aments that radiated from the female’s
body, shimmering with points of light.
Dr. Pietsch, of the University of Washing-
ton, said the rays contained nerves and
might act like sensory antennae, alerting
the angler to nearby prey. “We’ve hypothe-
sized that they pick up vibrations, like the
whiskers of a cat,” he said.
He and a colleague in Germany are trying
to determine whether the shimmering
lights in the rays are bioluminescent or
were merely reflecting light from the sub-
mersible. If the rays are glowing, he said, “it
would be really important.”
The new videos make clear — more so
than the old sketches and portraits — that
anglerfish look truly demonic. Why the
nightmarish appearance?
Dr. Robison noted that the exotic features
of anglerfish made perfect sense as evolu-
tionary adaptations to an icy, dark world in
which meals are few and survival depends
on cunning.
“Part of what appeals to us about other
fish is that they’re sleek and streamlined
and built for speed,” he said. “That’s attrac-
tive. But most anglerfish aren’t built for
speed. Their predatory approach is am-
bush. They draw things in. To aid that ap-
proach, they need to be stable in the water
column, to hold themselves in position.”
In the desert of the deep sea, he said,
“they have to take advantage of every prey
opportunity that comes by. That’s why they
have such huge mouths and distensible
stomachs: to take in a meal that might have
to last for months.”
“The big teeth may appeal to the 12-year-
old in all of us,” he added. “But those are re-
ally useful, too, in not only grasping prey
but trapping it in that maw.”
Most exciting, Dr. Robison said, is that
much about the realm of the anglerfish re-
mains ripe for discovery. Monterey Bay
may be “the best studied patch of ocean in
the world,” but it still produces surprises
about life in the abyss.
Water covers more than 70 percent of the
Earth’s surface and goes down miles; all
told, the global sea accounts for 99 percent
of the terrestrial biosphere.
“There’s a whole world of ocean out
there,” Dr. Robison said. “And most of it is
unexplored.”Ghouls of the Deep Caught on Tape!
A humpback anglerfish, fodder for nightmares.NORBERT WU/MINDEN PICTURESCONTINUED FROM PAGE D1
A few years ago, researchers at Harvard
and Kaiser Permanente Northern Califor-
nia had an inspired idea: Perhaps they
could use the wealth of personal data in
electronic health records to identify pa-
tients at high risk of getting infected with
H.I.V.
Doctors could use an algorithm to pin-
point these patients and then steer them to
a daily pill to prevent infection, a strategy
known as pre-exposure prophylaxis, or
PrEP.
Now the scientists have succeeded. Their
results, they say, show that it is possible to
correctly identify men at high risk by exam-
ining medical data already stored about
them.
But the researchers know they must
tread delicately in using the software they
developed. It’s one thing to have a computer
find a patient who is at risk for breast can-
cer. But to have software that suggests a pa-
tient is a person who too frequently has un-
safe sex and risks H.I.V. infection — how
should doctors use such a tool?
And if they do, can they initiate a conver-
sation about a patient’s sexual health with
understanding and delicacy?
“This certainly could be an aid for
providers,” said Damon L. Jacobs, a mar-
riage and family counselor in New York
who takes PrEP and educates others about
it.
But a lot depends a lot on the doctor: A
calculator that says a patient is at risk
“doesn’t mitigate the fact that providers are
often uncomfortable and clumsy talking
about sex,” he said.
And patients — especially those in minor-
ity communities who may mistrust doctors
— may well bristle. “Who’s looking over my
records? You think I’m a slut? You want me
to take an ‘anti-slut’ pill?” said Mr. Jacobs,
musing about how patients might react.
Clearly, doctors should not spring such a
result on patients, said Dr. Ellen Wright
Clayton, a professor of health policy at Van-
derbilt University. Instead, she said, they
should ask patients first if they want their
records reviewed by the software.
The drug Truvada, made by Gilead, was
approved seven years ago for PrEP. Taken
daily, it appears to be almost completely ef-
fective in protecting users against H.I.V. in-
fection.
But just 35 percent of the 1.1 million peo-
ple who could benefit from the pills use
them. And there are nearly 40,000 new
H.I.V. infections a year in the United States.
The problem is especially acute among
black gay and bisexual men: In their life-
time, one in two will be infected with H.I.V.,
said Dr. Julia Marcus of Kaiser Permanente
and a developer of the new algorithm.
Use of PrEP has lagged for a number of
reasons. Until recently, insurers did not al-
ways pay for the pills, which have a list
price of about $2,100 a month. And patients
do not always have a regular doctor who
knows them well enough to discuss H.I.V.
risk.
For the most part, the onus has been on
patients to ask for PrEP, said Dr. Douglas
Krakower, leader of the Harvard group de-
veloping an algorithm. “Clinicians are verybusy and have limited time and tools to
identify people who may be at risk,” he said.
Software might spare them the effort.
“We intuitively felt like there were many
data elements in the electronic health
record that could predict risk,” Dr. Marcus
said. The challenge was to identify which
ones worked best.
So the groups developed several different
models, using electronic records from 3.7
million uninfected patients at Kaiser Per-
manente and 1.1 million patients at two
Massachusetts medical centers.
Some models were simple: The software
looked at little more than sexual orientation
and history of sexually transmitted dis-
eases. But others were more complex.
The scientists tested these models by us-
ing them to review health records of people
who were free of H.I.V. infections and ask-
ing if they could identify who subsequently
became infected.
The final Kaiser Permanente model in-
cluded 44 predictors and factors like living
in an area with a high incidence of H.I.V.,
use of medications for erectile dysfunction,
number of positive tests for urethral gonor-
rhea and a positive urine test for
methadone.
The software flagged 2.2 percent of the
patient group, correctly identifying nearly
half of the men who later became infected.
The Harvard group had similar results with
its model.
Dr. Marcus said that there was a real need
to identify patients at risk. Half the patients
at Kaiser Permanente Northern California
who became infected said they had not
known about PrEP.
“Sexual health discussions may be awk-
ward, but they’re standard of care,” she
said.
She added that the group also would be
asking patients and providers about their
experience with the algorithm “to make
sure we implement our prediction tools in
the most practical and sensitive way.”
But the algorithm did not identify every-
one at risk, Dr. Clayton noted. So, after ask-
ing patients if they want the screening, doc-
tors will have to explain that the test is not
perfect and ask: “Are there other things in
your life that make you more at risk?”
Chris Tipton-King, a San Francisco film-
maker who takes PrEP, said the risk calcu-
lator could be a good opening for a dialogue
between a doctor and a patient.
“There are two sides to this equation,” Mr.
Tipton-King said. “One is educating pa-
tients, the other is educating doctors.”
A good friend of Mr. Tipton-King recently
asked his doctor to prescribe PrEP. Accord-
ing to Mr. Tipton-King, the doctor replied,
“Why don’t you just stop having so much
sex?”
Two weeks later, he said, his friend tested
positive for H.I.V.H.I.V. Predictor Requires Delicacy
JAMES STEINBERGSoftware reviews records to
identify at-risk patients who
need preventive measures.
By GINA KOLATAQ. What would the many chimney
swifts in my neighborhood do if
the chimneys all disappeared?
A.They would adapt by using
other high, protected hollow
spaces for roosting and nesting,
including trees.
North American chimney
swifts first adapted to the loss of
hollow trees by seeking out the
chimneys built by European
settlers. Reports of chimney nests
first appeared in the 17th century.
Most chimney swifts now share
nesting space in chimneys, but some still use trees. A 2003 review
article in the journal Aviation Conservation & Ecology identified the
most likely chimney alternatives, based on historical records from
1803 to 2013. The researchers listed 13 tree species, both deciduous
and coniferous, used by the birds in eastern North America.
The trees were all large, averaging 41 feet in height, and most were
in remote settings in unlogged forests. They included yellow birches,
maples, sycamores, cypresses, white pines and other pine trees. Some
had broken tops or other damage that created chimney-like openings.
Suitable chimneys are disappearing from neighborhoods, the re-
searchers said, and it was not clear whether there would be enough
appropriate trees to meet the future needs of the birds.
Birding organizations encourage leaving chimneys open during
nesting season, from spring through late summer. Researchers also
are experimenting with artificial nesting towers.
C. CLAIBORNE RAYQ&A Seeking Chimney, Suitable for Roosting
VICTORIA ROBERTSOn Sálvora Island, off the coast of
Spain, thousands of yellow-legged
gulls dot the grassy cliffs from
April to late July. It is a riot of
white wings and plaintive calls.
Occasionally, the chorus changes
as the seabirds engage in
courtship and chick-feeding. And
when the adults notice a predator,
such as a dusky-coated mink, the
chorus shifts again, to a charac-
teristic alarm call — ha-ha-ha.
These acoustic cues reach not
just young and adult gulls but
unhatched embryos, too. In 2018,
researchers found that when gull
eggs hatch, the ones that were
exposed to alarm calls were able
to crouch and hide from predators
a couple of seconds faster than
others.
A few other bird species, includ-
ing quails, fairywrens and zebra
finches, are known to relay simi-
lar cues about the environment to
their unhatched young, to prepare
hatchlings to fend for themselves.
But embryos aren’t receiving
wisdom only from their parents. A
new study, published in the jour-
nal Nature Ecology & Evolution,
suggests that they’re also receiv-
ing cues from nearby unhatched
siblings.
“Paying attention to cues from
the outside is important for sur-
vival,” said Jose C. Noguera, an
evolutionary ecologist at the
University of Vigo in Spain, who
led the study. Embryos that do so
develop traits that provide an
advantage in avoiding predators,
identifying other species of birds
or building their own nests in
warmer temperatures later in life,
he said.
To study how yellow-legged
gull embryos learned social cues,
Dr. Noguera and his colleague,
Alberto Velando, camped out on
tiny Sálvora during the birds’
egg-laying season.
Typically, the mother gulls lay
an egg a day until she has a
clutch of three. The researchers
waited and collected a total of 90
gull eggs over a period of three
days. The eggs were transferred
to incubators at a field station
nearby, and kept in groups equal
in size to wild clutches.
Each day, the researchers
removed two eggs from each
clutch and exposed them to audio
recordings of adult gull alarm
calls. Eggs from a control group
were handled similarly, except
they were not exposed to any
sounds. Thus manipulated, the
eggs were returned to their
clutches.
The researchers had created
two groups of gull embryos: Onein which two siblings had been
“informed” of possible predators,
with a third “naïve” egg, just as if
three had been laid a day or two
apart in the wild. In the other
group, all three gull siblings were
naïve.
All three embryos from the test
group soon exhibited the same
adaptations. The two “informed”
embryos made fewer peeping
noises while in the egg. But so did
the third embryo, even though it
hadn’t been exposed directly to
the gull calls.
And once the chicks hatched,
all three showed stunted growth,
higher levels of stress hormones
and faster reflexes than the con-
trol group, responding to alarms
by crouching and hiding more
quickly.
“That’s completely unexpect-
ed,” said Kevin McGowan, an
ornithologist at the Cornell Lab of
Ornithology, who was not in-
volved in the study. “It’s one thing
if everybody is on the same page
— they hear the same thing and
respond the same. But this study
suggests that birds that are a day
or two advanced can communi-
cate their wisdom to their sib-
lings.”
Vibrations from the two “in-
formed” eggs had transmitted
information about gull alarm calls
to the third egg. In the wild, the
small informational advantage
could mean the difference be-
tween being eaten or not.
“It’s a rough world out there,”
Dr. McGowan said. “Most young
birds do not survive to grow up to
be adults, and most adult birds do
not produce new offspring that
survive predation. Natural selec-
tion is a monstrous numbers
game.” If you can gain a leg up by
eavesdropping on your siblings,
why not do so?Bird Embryos Pay Heed
To Parental Alarm Bells
JOSE C. NOGUERA
Yellow-legged gull chicks are vulnerable
to carnivores such as minks.By KNVUL SHEIKH