Science_-_7_February_2020_UserUpload.Net

(coco) #1
I

n a swift rise to prominence, Jonathan
Pruitt drew attention to the small
field of behavioral ecology with eye-
catching findings about contrasting
personalities—meek and aggressive—in
social spiders. But in just 2 weeks, the
field has turned on its young star. What be-
gan with questions about data in one paper
has flared into a scandal, with dozens of
papers based on his data on spiders and
other animals being scrutinized by scores
of co-authors, including his former students
and postdocs.
Already, two papers co-authored by
Pruitt, who in 2018 was given a prestigious,
well-funded Canada 150 Research Chair at
McMaster University, have been retracted
for data anomalies; a third journal is ex-
pected to expunge another soon. And the
more Pruitt’s co-authors look, the more po-
tential data problems they find in his pro-
lific output. Many additional retractions are
expected, perhaps even an unprecedented
number for behavioral ecology.
The furor has even earned a Twitter
hashtag—#PruittData—where former col-
laborators and others are discussing how
to analyze his results and debating the
implications for their field. They are also
voicing suspicions that the problems go be-
yond carelessness. “An important question
that Pruitt should answer is how these data
came about,” says behavioral ecologist Niels
Dingemanse of Ludwig Maximilian Univer-
sity of Munich. “We should see proof that
there is nothing at fault here, because the

patterns we see cannot easily be explained
based on natural data collection processes.”
As the storm rages, Pruitt is in the middle
of 4 months of fieldwork in Australia and
the South Pacific. He says he did not com-
mit fraud and that the data issues are all
mistakes. “These errors are not unheard
of in data management,” he told Science in
his first interview since the retractions. But
Pruitt acknowledges his career is likely over.
“If a scientist can’t be careful, that’s as big
an indictment as someone who goes around
and adjusts data.”
Spokespeople at McMaster and the Uni-
versity of California (UC), Santa Barbara,
where Pruitt had a previous faculty posi-
tion, acknowledged the allegations but
would not say whether investigations had
been launched. But Pruitt’s fellow behav-
ioral ecologists have not waited. They’ve
launched their own investigations, hoping
to control the damage to their field.
Some researchers have tweeted that
the affair reflects a lack of scientific rigor
in animal behavior studies, particularly
those documenting distinct personalities
in cognitively simple animals, but Pruitt’s
colleagues reject that. “It’s a gross over-
statement to say this is now the death of the
field,” Dingemanse says.
Last year, a young researcher not in
Pruitt’s lab came to behavioral ecologist Tom
Tregenza of the University of Exeter with
questions about a paper in The American
Naturalist co-authored by Pruitt. Tregenza
recruited Dingemanse and two other col-
leagues to probe the paper, which used social
spiders as a test case to explore how social

SCIENCE sciencemag.org 7 FEBRUARY 2020 • VOL 367 ISSUE 6478 613

PHOTO: PROVIDED BY JONATHAN PRUITT


data with the floats’ own current-borne tra-
jectories, investigators can reconstruct over-
all currents and their speed.
The data set, compiled by oceanographer
Alison Gray of the University of Washing-
ton, Seattle, covers only 6 years, from 2005
to 2010, but Hu found that it reveals an
even clearer global speedup than the re-
analysis models. “The evidence in the Argo
data is absolutely astonishing,” says Eleanor
Frajka-Williams, an oceanographer at the
United Kingdom’s National Oceanography
Centre, who was not part of the study.
Gray says she was startled by the magni-
tude of the acceleration. But she notes that
ocean winds, which drive most currents,
have steadily increased over the past 3 de-
cades. And Hu says there’s good evidence
that human activity has contributed to that
strengthening. For example, in the Southern
Hemisphere, ozone depletion and green-
house warming have altered atmospheric cir-
culation to push the Southern Ocean’s famed
westerly winds to the south, perhaps causing
a slight strengthening and spreading of the
Antarctic Circumpolar Current. Meanwhile,
heat from the warming tropical Atlantic has
goosed the Walker Circulation, an equatorial
pattern that drives the Pacific trade winds.
Still, natural fluctuations can’t be ruled
out, says Gerrit Lohmann, a climate scien-
tist at the Alfred Wegener Institute. Over the
past few decades, long-term cooling off west-
ern North America has caused Pacific winds
to pick up—and that cooling may reflect nat-
ural oscillations in the ocean’s state. Other
researchers doubt these cycles exist (Science,
31 May 2019, p. 814). Either way, Hu thinks
the oscillations could be responsible for at
most one-third of the wind speedup.
The ocean acceleration could have globe-
spanning effects. Stronger tropical currents
could carry more warm water to higher
latitudes, for example. Because carbon diox-
ide (CO 2 ) is less soluble in warm water, that
could slow the ocean’s uptake of CO 2 from the
atmosphere. The high-latitude warming may
also be shifting weather patterns. At the same
time, Hu adds that by reaching deep into the
ocean, the acceleration could boost the stor-
age of heat in the depths, helping slow the
warming on land. “This is the first global
study,” says Janet Sprintall, a co-author and
oceanographer at the Scripps Institution of
Oceanography. “There’s a lot of uncertainty.”
Oceanographers will likely fan out to test
the study’s findings. Perhaps the strongest
confirmation could come from updated data
from the Argo floats, due out later this year.
Still, it will probably take another decade
of observations to be sure the trend is real
and driven by global warming, Wijffels says.
“This paper does highlight how ill prepared
we are to truly diagnose what’s going on.” j

Prominent spider biologist spun


a web of questionable data


Two retractions spark scrutiny of dozens of papers


SCIENTIFIC INTEGRITY

By Elizabeth Pennisi

Jonathan Pruitt
studies personalities
of social spiders.

Published by AAAS
Free download pdf