ADVANCES
18 Scientific American, December 2019 Illustrations by Thomas FuchsE C O L O G YBiome Boost
Sewage-treatment changes
helped wildlife in an English riverRivers act as Earth’s arteries and veins,
providing sustenance and sweeping away
waste to keep terrestrial habitats in shape.
By that measure, England is unhealthy:
a startling 86 percent of its rivers do not
meet water-quality standards, posing a
risk to wildlife and human health.
A new study offers hope. Invertebrate
biodiversity in one Thames River tributary
has increased in the past 30 years, thanks to
an adjustment in wastewater treatment, sci-
entists at the U.K.’s Center for Ecology &
Hydrology have found. “It’s starting to reach
levels one might expect to find in a river
without any wastewater,” says environmen-
tal scientist Andrew Johnson, lead author on
the analysis, which appeared in August in
Environmental Toxicology & Chemistry.Invertebrates such as crustaceans, in
sects and worms are key players in aquatic
ecosystems. They shape their environment
by digging into riverbeds and filtering water,
and they are both predators and prey. These
animals also respond quickly to environmen-
tal changes, indicating an ecosystem’s health.
The researchers analyzed data collected
by the U.K. Environment Agency between
1977 and 2017 for a 12-kilometer stretch
of the River Ray downstream from a large
wastewater treatment plant in the south-
western town of Swindon. They found a
steady increase in the variety and numbers
of invertebrates since June 1991.
That timing coincides with the 1991 Euro-
pean Union Urban Wastewater Directive,
which pushed treatment plants to switch
from filtration to an activated sludge process
that uses microbes to break down contami-
nants. This dramatically cut the organic mat-
ter and toxic ammonia going into rivers—
and so invertebrate biodiversity slowly
improved, the team concluded. “You could
liken it to being given a diet of cheeseburgersfor 20 years and then switching to a healthy
diet,” Johnson says. “Recovery is not instant.”
John Sumpter, an ecotoxicologist at
Brunel University London, says this increase
most likely boosted diversity among larger
creatures, too, and that these results proba-
bly apply in other places. Still, published
studies showing such improvements are
rare. “A big problem is that very few coun-
tries have the long-term data sets to conduct
the analyses required,” he says.
The new work suggests that “urban
rivers in the U.K. are recovering from the
gross pollution problems of the industrial
era,” says Steve Ormerod, an ecologist at
Cardiff University in Wales. But full resto-
ration will require more work and tougher
regulations, he adds, noting the growing
problem of agricultural pollution: “The
basic story of British rivers is one of urban
improve ment but rural decline.”
Yet Johnson thinks the River Ray results
show a possible path forward. “And may-
be,” he says, “wildlife is more robust than
we’d thought.” — Prachi PatelARTIFICIAL INTELLIGENCE
Automating
History
Computers can tell what
will matter (slightly) better
than humans can
As 2019 draws to a close, prepare for
endless roundups of the year’s most
important news stories. But few of those
stories may be remembered by 2039: new
research shows the difficulty of predicting
which events will make the history books.
Philosopher Arthur Danto argued in
1965 that even the most informed person,
an “ideal chronicler,” cannot judge a recent
event’s ultimate significance because it
depends on chain reactions that have not
happened yet. Duncan Watts, a computa-
tional social scientist at the University of
Pennsylvania, had long wanted to test Dan-
to’s idea. He got his chance when Columbia
University historian Matthew Connelly sug-
gested analyzing a set of two million declas-
sified State Department cables sent between
1973 and 1979, along with a compendium of
the 0.1 percent of them that turned out to be
the most historically important (compiled by
historians decades after their transmission).
Connelly, Watts and their colleagues
first scored each cable’s “perceived con-
temporaneous importance” (PCI), based on
metadata such as how ur gent or secret it
had been rated. This score corresponded
only weakly with inclusion in the later com-
pendium, they reported in September in
Nature Human Behaviour: the highest-scor-
ing cables were only four percentage points
more likely to be included than the lowest-
scoring ones. The most common prediction
errors were false positives—cables that got
high scores but later proved unimportant.
“I do think there’s a kind of narcissism of
the present,” Connelly says. “I’ve been
struck by how many times sports fans say,
‘That’s one for the history books.’ ”
Next, Watts says, to approximate anideal chronicler, the scientists decided to
“build the beefiest, fanciest machinelearn-
ing model we could and throw everything
into it—all the metadata, all the text.” The
resulting AI algorithm significantly out
performed humans’ contemporaneous
judgment. In one statistical measure of its
ability to pick out cables later deemed sig-
nificant, where 1 denotes no incorrect
inclusions or exclusions, it scored 0.14,
whereas the PCI scored 0.05. Although the
algorithm’s performance was far from per-
fect, the re searchers suggest that such an
“artificial archivist” could help to narrow
the field of events to highlight for posterity.
When tuned for this purpose, their model
weeded out 96 percent of the cables while
retaining 80 percent of those that wound
up in the compendium.
Emily Erikson, a sociologist at Yale Uni-
versity, who was not involved in the new
research, says that despite its use of imper-
fect data—compendium inclusion was up to
the subjective judgment of a few historians,
for example—the study offers a practical
tool and addresses Danto’s hypothesis. “To
see a machine-learning empirical test of this
conceptual puzzle is really exciting,” she
says, “and just kind of fun to think through.”
— Matthew Hutson© 2019 Scientific American