This suggestion, which has horrified environmentalists, was made at a meeting
in Seville two weeks ago to decide how to deal with the pollution caused by the
seven million tonnes of toxic mud that spread along the Guadalquivir River last
year. Most of it has been mechanically removed, but heavy metals—including
lead, copper and arsenic—have leached into the soil to a depth of 20
centimetres.

Victor de Lorenzo of the National Centre of Biotechnology in Madrid wants to
remove the pollution with plants genetically engineered to suck more metals out
of the soil. However, environmentalists headed by Greenpeace say that no one yet
knows how bioengineered plants might affect the ecosystem.

Some scientists claim that natural plants and microorganisms would be just as
effective. Pilar Bernal of the Centre for Agronomic Research in Murcia suggested
that one hectare of the wild herb Alpine pennycress (Thlaspi
caerulescens) could dispose of about 130 kilograms of zinc a year. The snag
is that introducing foreign plants might also upset the local ecosystem. “You
have to be very careful how you select candidates, as we are talking about a
natural reserve,” says Bernal.

Researchers from the Doñana Biological Station in Seville took blood samples from 150 living and dead birds along the contaminated stretch of the river and inside the nearby Doñana National Park. In aquatic birds such as ducks and coots living about five kilometres from the dam, they found levels of copper and zinc of up to 300 parts per million, 20 times normal concentrations and 5 times maximum tolerable concentrations. Such levels can cause severe poisoning and death.

“Pollution is getting into the food chain, but until now we have not found any measurable effect in animal populations,” says Fernando Hiraldo, a director of the station and an expert in bird behaviour. Aquatic birds are most vulnerable, he says. “They get the toxic metals in three ways—from the animals they eat, the water and the toxic mud attached to the plants.” The team found no increase in blood-metal levels in birds inside the park.

Spain has less than 100 brown bears, Ursus arctos, compared with
several hundred last century, all living in the Cantábrica mountains in
the north. In 1847, a road was built through their habitat, and since then a
railway and a national highway have been added along the same route. A ski
resort has also been built in the vicinity. These developments split the bears
into a small eastern and a larger western population that are just 30 kilometres
away from each other but have had no contact for more than half a century.

żěè¶ĚĘÓƵs from the National Museum of Natural History in Madrid and the
Foundation for the Brown Bear in Santander examined DNA from the
excrement and hair of bears from both groups to check their genetic diversity.
They found a high variability among the 70 or so animals in the larger group.
But the 17 bears in the smaller group shared many of the same genes. Lack of
genetic diversity makes a population less adaptable and less likely to survive.
The eastern group has been declining in numbers for years, and the researchers
warn that it could soon become extinct.

“In less than one hundred years, the splitting of the original population
into two parts has produced serious genetic consequences for the smaller group,”
says Ignacio Doadrio, a biologist at the Madrid museum. He points out that a
male bear will mate with four or five females and then keep them away from other
males. Naturally, in a population with few individuals this leads to genetic
similarity.

Doadrio says the best way of increasing numbers is to boost the larger
population by removing the traps left by hunters to catch wild pigs, and then
ensure that these bears can reach the eastern population.

SPANISH scientists have solved a long-standing mystery about the fate of the
country’s endangered imperial eagles. By working out where the young eagles go
after leaving the nest, they have managed to dramatically reduce the number of
birds killed before they can breed.

Ornithologists knew that the young abandoned their nesting areas within a
year of being born, only to return several years later. But no one knew where
they went during that time—or why so many died before they could
breed.

A team of researchers led by Javier Oria, an engineer and expert on imperial
eagles (Aquila adalberti) based in Extremadura, has now completed a
two-year survey in which eight eagles were tracked from their nesting areas in
central Spain via satellite transmitters attached to the birds’ backs. In a
separate three-year survey, they tracked 40 birds using VHF radio
transmitters.

After leaving their nests, the team discovered, young eagles fly hundreds of
kilometres south to “dispersion zones”, where they stay until they reach sexual
maturity up to five years later. When they are ready to breed, the eagles return
to their nesting areas to look for a mate. “The young can travel up to 400
kilometres from the nests to live in these zones,” says Oria. In one case, the
satellite picked up an eagle in Senegal.

Many eagles did not return to breed, the team found, because they were killed
when they flew into or landed on electricity wires in the dispersion zones.
“Before the surveys, 60 per cent of the young eagles died after being
electrocuted,” says Oria. Now, however, the mortality rate has dropped to 10 per
cent after the local authorities moved the electricity wires out of the eagles’
flight paths.

In one area, outside the Doñana National Park, 84 per cent of young
eagles from the park were dying from electrocution. Wiring changes have reduced
this to a minimum.

YOU may never be able to pick up a sheet of music and play it like the great
alto saxophonist Charlie Parker. But your computer can—with a little help
from artificial intelligence researchers from Spain.

Researchers at the Artificial Intelligence Research Institute in Barcelona
(IIIA-CSIC) have developed a program called Saxex, which allows your computer to
play music like any of the great jazz saxophonists. The only input it needs is
the sheet music.

Musical notation is quite limited. A score provides information about which
notes to play and the sequence in which they are to be played, together with
some quite crude clues about their timing and loudness. Musicians have to make
their own interpretations—and they are all different.

Saxophonists are no exception. One note may be clipped short or smeared into
another. The players may increase their vibrato while playing a note, or even
bend the pitch of the note as they play it. They may have a different tone. It
is all part of musical expression. The tenor players Ben Webster and Lester
Young, for instance, never sounded the same even when they played the same
notes.

For the first time the Spanish researchers have been able to imitate the
nuances of these very human musical instruments. The software draws on a
repertoire of performances by real saxophonists that is stored in the computer’s
memory. “We are talking about several hundred notes in the memory for a program
of a few megabytes,” says Josep Lluis Arcos, one of the Barcelona researchers.

By analysing the sound spectrum of the notes in these performances, Saxex
looks at four expressive parameters from the samples stored in the computer’s
memory—how loud the notes are, their duration, vibrato and how staccato
they are played.

The researchers convert the score of a tune—say the jazz standard
Autumn Leaves— into the Musical Instrument Digital Interface (MIDI)
computer language, which is essentially a sophisticated version of a piano roll.
This provides the software with information about the notes of the melody, their
sequence and an idea of their timing. However, this MIDI file, and the sound
file that the computer generates, lack any expression.

Saxex uses the information in the MIDI file to divide the theme of Autumn
Leaves into musical phrases. It then searches the repertoire in its memory
for similar phrases and similar passages of music, and picks the best match.
When it has found a good match it attaches the expressive information from this
passage to each note before generating a final sound file.

“Saxex works by comparison, examining each note of the given melody with that
of the expressive recordings, and finds out what they share in common,” says
Ramon López de Mantarás, another member of the team that
created the program. “It decides how to transform the initial melody to produce
music that the audience perceives as more human.”

At present Saxex has only been programmed to transform relatively slow-tempo
themes, such as ballads . But the team plans to expand the program’s range and
widen the number of expressive parameters it can handle, to include
features such as the breathy tone of some saxophonists.